Case Study: Hradec Králové, Czech Republic, Muzeum Východních Čech, Knihovna, MS Hr-7 (II A 7), (manuscript)
By Ian Rumbold
Hradec Králové, Muzeum Východních Čech, Knihovna, MS II A 7 (CS-HKm II A 7) is commonly known as the ‘Speciálník Codex’ after a term used in a seventeenth-century inscription, now badly damaged and only partly legible, on the inside front cover of the manuscript, which was reconstructed by Dobroslav Orel (Orel 1914, 11; Orel 1922a, 4) as follows:
Letha panie 1611 po bozim Tiele
w pondeli Michl muratt Hospodarz
wlazni pod auwozem Tento Specialnik od [ewzdal]
k Sw[ate]mu Petru.....ch* czest a chwa[lu]
dala Panu Bohu wsse[mo]h[au] czimu.
Beatus populus qui scit Jubilationem
Mox [recte: Non] moriar sed vivam et narrabo [o]pera Domini.
* v Okovech? (in Vinculis)
In the Year of Our Lord 1611, on Monday
after Corpus Christi, Michal Muratt, landowner
at ‘Lazni pod Auwozem’, gave this Specialnik
to (the Church of) St Peter (?in Chains). Honour and Praise
be given to the Lord God Almighty.
Happy is the people who know jubilation.
I shall not die, but live and tell the works of the Lord.
The term ‘Speciálník’ – found also in relation to books in a number of wills from the period (Zilynskij 1987) – relates to the special nature of the manuscript: it is not a regular liturgical book such as a Gradual or Antiphoner, and it contains primarily polyphonic music (though some monophonic music is also present, mostly in a single gathering at the end of the manuscript).
Speciálník was compiled between c.1485 and c.1500 not in Hradec Králové (in Eastern Bohemia) itself – it was purchased for the library there in 1896 (Orel 1914, 12) or 1901 (Orel 1922a) – but probably in Prague. Nothing is known of Michal Muratt, the original donor of the manuscript, beyond the information provided in the inscription quoted above that he owned land at Lazni pod Auwozem (the baths below Ujezdem, one of the principal streets in Prague); the church of St Peter referred to also remains unidentified. Mass movements found on ff. 10v–11r, 78v–79r and 168v–169r are described in the main body of the manuscript or in the original index as ‘Patrem mostske’, ‘Patrem malostranske’ and ‘Sanctus berúnské’, apparently referring respectively to the Stone Bridge (Charles Bridge), Malá Strana (Lesser Town, one of the four cities of medieval Prague) and to Beroun (a city in central Bohemia close to Prague) (Mračková 2009, 39, who, however, cautions that ‘Patrem mostske’ and ‘Sanctus berúnské’ may alternatively refer to the city of Most in northern Bohemia and the Italian city of Verona respectively). The songs Bud bouhu and Wzdaymez chwalu (ff. 35v–36r and 44v) also both indicate connections with Prague. The Patrem on ff. 216v–218r, described in the original index as ‘Patrem Motyčkovo’, may be somehow connected with ‘Wenceslaus Motyczka, faber’, documented as a member of Prague city council between 1484 and 1519, a smith and a resident of the Malá Strana, where he bought a house in 1477 (Mračková 2009, 40).
Speciálník was first studied in detail by Dobroslav Orel in his Ph.D. dissertation (Orel 1914) and elsewhere (Orel 1922a, 1922b), and the older repertory it contains (i.e. the part of the manuscript that was copied in full notation; see further below) was the subject of a further dissertation more than 40 years later (Gierschik 1965). The manuscript figured in Jaromír Černý’s catalogue of the musical sources at Hradec Králové (Černý 1966, 40–1), and Černý went on to study some of the music from Speciálník – principally the older repertory again – in the course of his work on medieval polyphony from Bohemia and on the composer Petrus Wilhelmi (Černý 1971, 1972, 1973, 1975, 1984, 1989, 1990, 1993). Dagmar Vanišová focused in her dissertation on the secular music in the manuscript (Vanišová 1981), and published an edition of some of the later repertory (Vanišová 1990). Jitka Petrusová provided an inventory of the later works (those copied in void notation) and explored their concordances in Western European sources; by this and other means, including an investigation of the watermarks, she was able to correct earlier mistaken assumptions about the dating of the manuscript, and to place it in the last fifteen years or so of the fifteenth century (Petrusová 1994). (The manuscript had previously been believed to be of much later date, and thus to be uncommonly retrospective. Such errors were based on dates found on the binding itself (1540, 1546), and even in a donor’s inscription (1611: Orel 1914, 11; Reese 1959, 728–41; Hughes/Abraham 1960, 300).) Petrusová then went on, in a second MA dissertation, to investigate the contrafacta in the manuscript (Petrusová 1996). Lenka Mračková studied the Mass music (Mráčková 1998, 2004b, 2004c, 2006), and Paweł Gancarczyk investigated the internal chronology of the manuscript (Gancarczyk 2001). Mračková has published studies of the manuscript and its context in international journals that have brought it to the attention of a wider readership (Mračková 2003, 2009a).
The genesis of the manuscript has traditionally been seen in the context of the Utraquist Church, with its practice of communion in both kinds (sub utraque specie; Orel 1922a, 22), and of the literary confraternities that developed in Bohemia (and particularly in Prague) in the 1490s, following the treaty between Catholics and Utraquists known as the Peace of Kuttenberg (Kutná Hora, 1485). (On the development of these confraternities, see Pátková 2000.) However, its compilation arguably predates the widespread establishment of these confraternities, and it differs from sources more securely identified as having been associated with them (such as the Franus Codex, housed in the same library: MS II A 6) even in such basic features as its external dimensions (which are smaller) and in the material on which it is written (paper, rather than parchment). More recently it has been suggested that it may have been compiled for a Latin school rather than one of these confraternities (Horyna 2007).
The Speciálník Codex is a paper manuscript that now contains both polyphonic music (some notated in void notation, some in full) and monophonic, and consists of 305 folios, which measure about 375 x 275 mm. The partial loss of some marginal text indicates that substantial trimming took place prior to binding, but no music was lost. The present binding, which dates from the mid-sixteenth century (the date 1540 appears on the binding itself), measures 395 x 290 mm, and is of wood covered with white blind-tooled pig’s leather strengthened with metalwork: both covers have a central square of metal with a boss, fittings for two metal clasps on the outer edge (of which the lower is missing from the front cover, the upper from the back cover) and metal bands along the central portion of the top, outer and bottom edges, and originally had four pointed corner-pieces, all of which are now lost. The gathering containing chant hymns (ff. 292–303) has been thought to be of independent origin from the rest of the manuscript, but probably emanates from the same circle as the rest of the manuscript, since one of the paper-types found in the core of the manuscript also appears in the chant gathering, and since the original index of the manuscript (see below) begins on the last page of this gathering. Note that, except where earlier folio/page numbers that physically appear in the manuscript itself are specifically under discussion, the foliation referred to in this case study is that assigned in the online facsimile of the manuscript at www.manuscriptorium.com, since that is, and will presumably remain, the primary source of access to the manuscript for most present and future users.
The core of the manuscript consists of 23 gatherings, the folios having originally been numbered (in the upper right-hand corner of each recto) a1–a20, b1–b20, etc., to n1–n20, and o1–o13. Gathering I seems to have been a quintern, from which the outer bifolio (ff. a1 and a10) is now missing. Prior to the binding of the manuscript, this gathering must have become dismembered. The separated folios, whose edges have been repaired with adhesive tape which is now itself flaking, were reassembled as artificial bifolios. An additional single folio (manuscriptorium f. I, which may or may not originally have belonged to the manuscript) was incorporated as a substitute for the lost f. a1, so that the gathering now contains an odd number of folios. The isolated folios were pasted onto opposite edges of strips of new paper to form bifolios in ways that do not consistently reflect the original structure of the gathering. The fact that the paper of this first gathering has no watermark (see below) is a further hindrance to confirming the original and present structure of the gathering. The new (artificial) structure of the gathering cannot be ascertained with certainty because of the tightness of the binding, but stitching can be observed between original ff. a3 and a4 (manuscriptorium ff. 2 and 3) and between original ff. a8 and a9 (manuscriptorium ff. 7 and 8). The new folio (manuscriptorium f. I) is probably conjoined with f. a6, f. a2 with f. a5, f. a3 with f. a4, and f. a8 with f. a9; f. a7 seems to have been pasted on to f. a6 (see Figure 1). Furthermore, the first surviving folio containing music (original f. a2; manuscriptorium f. 1) was bound in the wrong way round, with the original recto now following the original verso. The following 22 gatherings are all sexterns, and all are intact, except that one folio is missing (and was apparently so at the time the original foliation was inserted, presumably prior to binding) between ff. 72 and 73 (original ff. d14 and d15) in gathering VII: only the stub of this folio remains. (No music is missing at this point.) Gathering XXIII contains an extra half-sheet, not bound in, following f. 267 (original f. o9; not reproduced in manuscriptorium).
This core section of the manuscript was probably begun in the 1480s and completed in the mid-1490s, when an index was prepared of the music that had been copied by that stage. (Only some of the later additions to the manuscript were subsequently added to the index.) To these 23 gatherings were added a further two (both quinterns), numbered p1–p14, q1–q6, and a further gathering of chant hymns, which were not originally foliated. The alphabetical index, as already mentioned, begins on the last verso of the chant gathering (f. 303v), and it continues on both sides of the first folio (f. 304r–v) of an additional bifolium. (The index must therefore postdate the addition of gathering XXVI.) One further polyphonic item was inserted on the recto of the remaining folio (f. 305r). In addition to the early foliation already mentioned, the pages of the manuscript are numbered in pencil at the centre of the foot of each page. A collation of the gathering structure, the original foliation, the later pagination, and the modern editorial foliation used here is presented in Table 1. Eight different paper-types can be distinguished (Petrusová 1994, 11–26; Gancarczyk 2001, 112–13, 119–21), and an attempt has been made to trace the sequence of the compilation of the manuscript by grouping the ascribable contents into four layers representing five-year periods (Gancarczyk 2001, 121–4).
Although the origins of Speciálník are firmly Central European, it differs from contemporary Bohemian Cantionals – such as the Franus Codex, already mentioned – in that it contains a substantial repertory of contemporary music, mostly copied in void notation, that was either imported from Western Europe or composed locally under the influence of such music, as well as a more retrospective repertory of specifically Bohemian motets and cantiones, mostly copied in full notation. Its significance rests partly in its substantiality – it contains a total of 228 polyphonic compositions, of which many are unica – but, importantly, also in the evidence it provides both that Western European music was well known and (presumably) performed in Bohemia in the late fifteenth century, and that the composition of polyphonic music was regularly practised in the area at that time.
Of the 228 polyphonic items in Speciálník (see Table 2), about two thirds (153 works; nos. 1–132, 134–7, 167, 169–83, 209–11, 213, 218, 220, 222, 225–8) were copied in void mensural notation, the remainder in full mensural or Gothic chant notation. In addition, the first section of the Sanctus on f. 9v–10r (no. 6), which is double-texted ‘Sanctus’ and ‘Pleni’, is copied in full notation, though the following ‘Osanna’ section is in void notation; and the tenor of the following Patrem (no. 7, on ff. 10v–11v) is in full notation, though the other parts are in void. Among the full-notation material, Gothic chant notation is used for the Sanctus settings nos. 142 and 144–8, and for the Gloria trope no. 214. Elsewhere, it is also used for the tenor only of nos. 88 and 90–2 (the Proper items of the Mass cycle on ff. 164r–168v; see below, Example 4). Unpublished inventories of the works copied in void and full notation are to be found in Petrusová 1994 and Gierschik 1965 respectively. These inventories do not assign separate numbers to the movements of Mass cycles, and therefore count only 199 items. An (unnumbered) basic list of the items may be found online at
The composers of about half of the void-notation works can be identified through ascriptions either in the manuscript itself, or in concordant sources. 31 composers are represented in this part of the repertory (the inscription ‘Gontrassek’, found on f. 36r
'Gontrassek', found on f. 36r
Close , is not, as has sometimes been stated, the name of a composer, but means ‘contratenor’). Full notation is used in gatherings XVIII–XXV (on ff. 210v–211r, 214v–235v, 236v–237v, 248r–266r, 268v–269r, 271r–273r, 277r, 279v–280r and 283r–285v) primarily for an older, mostly anonymous repertory of Bohemian cantiones and motets. Among a sequence of such works, however, is a more modern Sanctus setting by Johannes Pullois (no. 141, ff. 218v–220r), whose first two sections (‘Sanctus’ and ‘Osanna’) were copied in full notation, though the following ‘Benedictus’ is in void. The opening words of the texts of a few of these works provide (through an acrostic) the name of their author – and, presumably, the composer of the music to which they are set – namely Petrus Wilhelmi of Grudencz (see Černý 1975 and 1993). Details of the identifiable composers and their contributions to this manuscript are provided in Table 3. For the ascriptions ‘Yacten’ and ‘Elezanger’ see Mráčková 2006. Inscriptions such as ‘sociorum’ and ‘ad equales’, which occur fairly frequently both in the main part of the manuscript and in the index, indicate works that are written for adult male voices only (i.e. do not include an upper voice-part for boys’ voices). This practice has been associated in the literature with the Bohemian confraternities which, however, as already mentioned, appear not to have flourished until some time after Speciálník had been compiled.
2. Description and Analysis
a. Preparation of the page
In some gatherings (III–V, IX, XIII, XV–XVII, XX–XXI and XXIII–XXV; i.e. ff. 21–56, 92–103, 140–151, 164–139, 224–247, 260–291), single bounding lines were drawn in the left and right margins as frames for the staves. (There is no obvious rationale that would explain why such lines were used in some gatherings but not others.) Prickings can often, though not always, be seen in the corners of the pages as guides for these borders. The lines were drawn in lead point, and on the rectos only, though they can normally be seen sufficiently clearly also on the versos to have served as stave-frames there also. The staves were drawn in brown ink using a single rastrum. The width of the rastrum or rastra used in the majority of the manuscript was 19 mm, though one or more tools of 21 mm width were employed throughout gatherings XIX–XX (ff. 212–235) and XXII–XXIII (ff. 248–271). The width of the staves varies between 211 mm and 232 mm, and the distance between the top line of the top stave and the bottom line of the bottom stave between 305 mm and 325 mm. The spaces between staves vary between 16 mm and 22 mm. The four-line staves in gathering XXVI were ruled with one or more 11 mm rastra, while the spaces between the staves were ruled for text at 10 mm intervals.
The first eighteen gatherings of the manuscript and gathering XXI were all ruled with nine music staves per page, but gatherings XIX–XX and XXII–XXV have only eight. In gatherings I–II, VI–VII and X–XI, some staves were indented in order to leave space for initials and/or voice names. Most frequently, both the first and either the fifth or the sixth of the nine staves were indented, evidently with the typical layout of a four-voice composition in mind (see below): four or five staves were thus made available for each voice-part on each opening, and when stave 6 was indented, frequently stave 5 was left empty as a divider between the parts. On some pages within these gatherings the stave-ruler omitted to indent one or both of the two staves (see, for example, ff. 70r–71r and 78v); and on two occasions (ff. 121r and 122v) he indented both the fifth and sixth staves, as well as the first. Only rarely is the potential of the stave-ruler’s design realized: only a few initials were inserted in the spaces left (see ff. 3r, 72v–73r and 79r), all of the remainder being added in the left margin or superimposed over the first few centimetres of the stave, or both (see further below, under ‘Decoration’). Although the names of lower voices were sometimes inserted in the lower space of a verso and either or both spaces of a recto (see ff. 9r, 13r
'the names of lower voice were sometimes inserted in the spaces left by stave-ruler'
Close , 14r–16r, 17r, 18r–v, 19v–20r, 104r–105r, 106r, 107r, 108r–109r, 110r, 111r, 112r, 113r, 114r, 115r, 116r, 122v, 123v, 124v), these names were almost as often written outside the spaces, even when the spaces fall at appropriate places in the layout (see ff. 5r
'these names were almost as often written outside the spaces, even when the spaces fall at appropriate places in the layout'
Close , 7r, 8r–v, 11r–v, 12v, 57r, 58r, 59v–60r, 61v–62r, 63r, 106v, 111v, 116v, 116v, 122v, 123r, 124r, 126r). Very frequently, the indentations left by the stave-ruler do not fall conveniently at the beginning of a new voice-part, but in the middle of one, and the scribes’ usual response to this was to fill in the space at the beginning of the stave by extending the stave backwards, usually freehand (see below, under ‘Copying of text and music’).
b. Copying of text and music
Only one page of the Speciálník Codex was left completely blank: f. 291v, the last page of gathering XXV, and the last page of the manuscript before the chant gathering. Within the manuscript, some 24 pages (ff. 36v–37r, 55v–56r, 58v, 69v–70r, 76r–78r, 79v, 99r, 130r, 163v, 186r, 269v, 282v, 283v–285r, 286r) contain only empty staves; a further two rectos (ff. 183r, 280r) contain only a little music on the last stave, run over from a voice-part begun on the facing verso; and one verso (f. 276v) contains only two lines of text in the bottom margin.
A large number of scribes – probably about 25 – contributed to Speciálník, and there are significant differences between their work with regard to many aspects of mise-en-page. A detailed study by Ian Rumbold of the characteristics of their hands and their contributions to this manuscript is in progress; the following, however, incorporates a brief and provisional summary of his findings. The precise number of scribes whose hand appears in Speciálník is impossible to determine, but in any case the majority of them made only isolated contributions of one or two pieces. Most of the manuscript was copied by only five hands, of whom one (here referred to as scribe A) predominates. He was responsible for copying the whole of five of the gatherings (gatherings XVII, XIX–XX and XXII–XXIII; i.e. ff. 188–199, 212–235, 248–271), and his hand is present in a further nine (gatherings III, VII–IX, XII, XV–XVI, XVIII and XXI; i.e. ff. 21–32, 69–103, 128–139, 164–187, 200–211 and 236–247). He copied, in fact, most of the full-notation material in this manuscript, and was also the largest single contributor to the void-notation material. There is considerable variation in his hand within the manuscript, and his full-notation work looks quite different from his void-notation work, owing to the different equipment and techniques involved.
Two further scribes, here referred to as B and C respectively, also made important contributions throughout the manuscript. Scribe B copied the whole of one gathering (gathering X; i.e. ff. 104–115), and made contributions to a further eleven (gatherings II–VI, IX, XI, XIV, XXI, XXIV–XXV; i.e. ff. 9–68, 92–103, 152–163, 152–163, 236–247 and 272–291). Scribe C contributed to 8 gatherings (gatherings III–V, XII–XV, XXI; i.e. ff. 21–56, 128–175 and 236–247). Two more scribes, D and E, made significant but smaller contributions which, furthermore, are restricted to a few gatherings of the manuscript. Scribe D, using void notation, filled all but the last page of gathering I, and also copied Ghiselin-Verbonnet’s Patrem La belle se siet in gathering VII, on ff. 70v–74r. Scribe E, by contrast, used full notation only, and contributed to gatherings XXIV and XXV (i.e. ff. 272–291).
Within the void-notation section of the manuscript, each new voice-part normally begins on a new stave. Where the conclusion of a voice-part requires only a fragment of a stave, the last notes may be written at the beginning of the last stave allocated to that part, leaving the end of the stave empty, but are often written rather at the end of it, so that the beginning of the stave is left empty. The disposition of the voice-parts on the page or opening is affected by a number of factors, not least the numbers of parts for which the music in question was composed. Most of the void-notation repertory in Speciálník is in three or four voice-parts. Only one piece (no. 37, an anonymous Patrem on f. 80r
Close ) is in two parts throughout. (A reduced two-part texture is additionally used for internal sections of some works – mostly Sanctus settings – as a means of contrast to the fuller texture of surrounding sections.) No. 37 is a simple setting of the Credo notated without clefs and occupying only five staves on f. 80r
Close . Neither voice is labelled, but the second clearly functions as tenor to the discantus above it. Identified in the original index as ‘Patrem plihalowo’, the setting consists of only two short sections, each provided in both voices with multiple text incipits to show which section is to be used for which phrases of the text. (The phrase ‘Et in spiritum … venturi seculi’ is omitted.) Only three works in the void-notation repertory are for five voice-parts: nos. 22 (Anon., Miserere nostri domine, ff. 39v–41r), 121 (Bedingham, Festo isti, ff. 201v and 202v–203r, a contrafact version of the composer’s three-voice song Gentil madonna, with two additional voices uniquely preserved here in Speciálník) and 178 (Anon., Omnis mundus iocundetur, ff. 242v–243r). (Additionally, the ‘Benedictus’ section of the Sanctus from the Missa Dieu quel mariage, no. 4, on ff. 5v and 6v–9r, is in five parts, a ‘secundus discantus’ being added to the three-part texture of the remainder of the work, and the tenor being performed both forwards and backwards simultaneously: ‘Tenor secundus cancrisat’.) The layout of these works is further discussed below.
The remainder of this void-notation repertory is in three or four parts. In both the three-part and the four-part works that occupy a single page or less, or where a longer work has been divided into sections of which each is presented on a single page, the upper voice (discantus) is invariably written at the top of the page. as in, for example, the three-part Wzdaymez chwalu, no. 26, on f. 44v and the four-part Kyrie, no. 8, on ff. 12r–13r. In this Kyrie, the scribe has opted to devote a page to each of the three sections of the setting, and to keep the four voices of each together on the same page, when he could, alternatively, have combined two sections across an opening, and placed the other section on a third page. The name of the third voice-part of this Kyrie also changes during the course of the piece, from ‘contratenor’ on f. 12v to ‘bassus’ on f. 13r; or perhaps, despite their separation on the opening, one is supposed to construct the two names as the two halves of a single name (‘contratenor bassus’). (The part is not named at all on f. 12r.
In three-part works the discantus is usually followed by the tenor, then the remaining part, normally called ‘contratenor’. ‘Altus’ and ‘bassus’ are alternative names for this third part, and indeed the name does not always remain consistent within a piece. The third part of the Sanctus for men’s voices on ff. 5v–8r (no. 4), for example, is labelled ‘contra’ on f. 7r, but ‘bassus’ on f. 7v and ‘contratenor’ on f. 8r. (Elsewhere in the piece it is not named at all.) This work is mostly in three parts, but, as already mentioned, a fourth part, labelled ‘secundus discantus’, joins in for the ‘Benedictus’ on f. 8r, where the canonic tenor part also divides into two, one performing the part as written, the other in retrograde, producing a temporary five-part texture. For some reason the ‘Pleni’ and the (completely untexted, two-part) first ‘Osanna’ were written out (on ff. 5v and 6r respectively) before the rest of the piece; the scribe made the note ‘Pleni ante sanctus’ on f. 7r, following the opening ‘Sanctus’ section. Four-part works that occupy only a single page generally display the voices in the order discantus, altus, tenor, bassus (e.g. the three sections of the Kyrie Petite camusette, no. 8, on ff. 12r–13r, and Qui te aspectamus / Da nobis, no. 43, on f. 86r) or discantus, tenor, altus, bassus (e.g. nos. 170–5, on ff. 238v–241r), but the sequence discantus, tenor, bassus, altus is also found (e.g. Berbigant’s Pfobenscwanz, no. 116, on f. 195r, and Frye’s Ave regina, no. 126, on ff. 204v–205r).
Where, as is usually the case, a three-part work requires more than a single page, it is normally displayed across one or more openings of the manuscript. The discantus invariably occupies the upper staves of the verso, but there is considerable variety in the arrangement of the other voices on the opening. The tenor may be placed beneath the discantus, with the contratenor (sometimes alternatively labelled ‘bassus’) on the facing recto, as in the anonymous three-part setting of Ave Maria, no. 24, on ff. 42v–43r, where the upper half of the recto folio has been left blank, and the bassus has been placed in the lower half, i.e. opposite the tenor rather than the discantus. (The voices are untexted apart from a two-word incipit in the discantus – just enough to enable the singers to know which text they should supply, presumably from memory.) Much less commonly, the contratenor is placed beneath the discantus, with the tenor on the facing recto, as in, for example, O lucis alme sator, no. 120, on ff. 200v–201r, and Imperatrix virgo gloriosa, no. 176, on ff. 241v–242r. In the first of these, none of the voices is named, but the musical characteristics of the lower voices define that on the lower portion of 200v as a contratenor, that on the upper portion of f. 201r as a tenor; on f. 201r an omission from the contratenor has been inserted on the recto opposite. In Imperatrix virgo gloriosa the voice at the top of f. 242r is labelled ‘tenor’, so that at the foot of f. 241v must be the contratenor. (The inscription at the beginning of the latter voice – ‘Faux perzě naquinternu’? – has not been satisfactorily interpreted.) The lower portion of f. 242r has been used for another short piece, Nobis est natus (no. 177).
Occasionally in three-part works occupying one or more openings, the tenor and contratenor appear one above the other on the facing recto. The tenor is normally placed above the contratenor in such cases, but they may be reversed, and indeed the order may not remain constant even within a single piece that occupies more than one opening. In the Patrem on ff. 3v–5r (no. 3), for example, the contratenor appears above the tenor on the first opening, but beneath it on the second. This may be related to the fact that the contratenor is silent in the section (‘Qui propter nos homines’) that immediately follows the page-turn, which is a duet between the discantus and tenor. On other occasions, especially where it would be difficult to fit both of the lower parts onto a single page, one of them may be placed on the upper staves of the recto, while the other begins beneath the discantus on the verso and continues on the lower staves of the recto. In such cases, the voice-part that is split between the verso and the recto may be either the tenor or the contratenor. Where a part is divided between facing pages in this way, some method of visual leading from the end of the material copied on the verso to the beginning of that copied on the recto is usually found in addition to the usual custos. Most often – and throughout scribe B’s work, for example – this is merely a straight horizontal or diagonal line: see ff. 10v–11r, 17v–18r, 22v–23r, 86v–87r, 93v–94r, 95v–96r, 123v–124r, 137v–138r, 175v–176r, 177v–178r, 178v–179r, 187v–188r, 188v–189r, 192v–193r, 194v–195r, 200v–201r, 215v–216r, 245v–246r, 273v–274r, 274v–275r, 275v–276r, 286v–287r, 287v–288r. On two occasions, this diagonal line is an extension of the custos itself: see ff. 23v–24r and 116v–117r. Sometimes, especially in the work of scribes A and C, a pair of identical or very similar signs is used, one placed at the end of the music on the verso, the other at the beginning of that on the recto. Occasionally an ‘x’ cross made up of between one and three diagonal lines crossed by one or more similar lines in the other direction, and sometimes with a pair of dots to left and right or top and bottom is used: see ff. 28v–29r, 29v–30r, 100v–101r and 195v–196r. On ff. 28v–29r, 29v–30r and 195v–196r this sign appears in conjunction with a longer diagonal line leading from one point to the other. On f. 196r a pointing finger (the only one in the manuscript) also occurs, indicating the point at which the voice-part resumes on the facing recto, and on ff. 29v–30r the link is made even plainer by an additional pair of signs, consisting of four dots arranged as a ‘+’ cross, with curlicues leading from the dots themselves and from the spaces between them. Similar signs, always consisting four dots or rhombuses arranged as a cross, with or without curlicues, are used for the same purpose on ff. 1r/3r, 35v–36r, 150v–151r, 244v–245r. On f. 288v a pair of signs of this type is used to lead to the continuation of a voice-part not on the facing recto but higher up on the same verso. And a related pair of signs, consisting of a ‘+’ cross with a trio of dots at the end of each arm, appears across the opening on ff. 99v–100r, in the work of a scribe who copied only a pair of works on ff. 99v–102r.
In some cases where a voice-part is divided part across the bottom of an opening, space is left between the discantus and one lower part at the top of the two pages and the other, divided, lower part at the foot, and the quantity of music at the foot of the verso and at the foot of the recto is balanced. (See, for example, the Patrem labelled ‘Patrem Sociorum’, no. 12, on ff. 21v–23r, where, however, the empty stave above the beginning of the (unlabelled) bassus part at the foot of f. 22v was subsequently used to insert a passage that had been omitted from the discantus part, the point of insertion being marked by a sign in the second stave on that page.) For an example in which the tenor is split in this way, see Bud bouhu chwala, no. 20, on ff. 35v–36r. In this case, no empty stave was left above the tenor on f. 35v, even though the last stave on that page is unused, but two were left on f. 36r, which were filled by a large initial letter (see under ‘Decoration’ below). Elsewhere, the divided lower part begins immediately after the discantus, as in Bud bouhu chwala, but (unlike there) as much of the lower part as possible is inserted at the foot of the verso, and the foot of the recto is used only as necessary for the remainder of the part; in such cases more of the divided part may be found at the foot of the verso than at the foot of the recto.
Only rarely in this manuscript does a voice-part which begins at the foot of a verso continue at the top of the facing recto. However, in no. 95 (Anon., Regina celi, ff. 169v–170r), for example, copied by a scribe whose hand does not appear elsewhere in the manuscript, the discantus is followed by the tenor, which begins on staves 6–9 of the verso and continues on staves 1–2 of the recto; the bassus follows on staves 3–8 of the recto. (See also the detailed discussion of nos. 88–93 in Example 4 below.) In one case in which a tenor begins at the foot of a verso but cannot be completed there, the conclusion of the voice-part is written both at the top and at the foot of the recto (see the first opening, ff. 86v–87r, of Agricola’s Patrem, no. 45). And in one case in which a bass part is displayed across the bottom of an opening (ff. 123v–124r, the first opening of Thomek’s Patrem J’abandonne (no. 66)), some of the staves on the facing pages are treated as continuous: the music begins on stave 6 of the verso, and continues on stave 7, then proceeds to stave 7 of the facing recto before crossing back to stave 8 of the verso and running on to stave 8 of the recto; stave 9 of the verso is left blank, and the music concludes on stave 9 of the recto.
Often details of the layout reflect the nature of the music and the relative space required by the different parts. See, for example, the three-part Patrem by Tourout, no. 49, on ff. 92v–94r. The section ‘Patrem … descendit de celis’ appears on the first opening, with the discantus in the top left position, the tenor top right, and the contratenor bottom right (concluding on an additional fragmentary stave in the bottom right corner of f. 93r, drawn freehand). The following section, ‘Et incarnatus … sepultus est’, is a contrasting duo between the discantus and the tenor, and the two parts are written out, in that order, at the foot of f. 92v. The discantus can therefore read the relevant voice-part continuously, since the second section of his part continues directly on from the first. The tenor, by contrast, has to leap at the end of the first section from the middle of f. 93r (where the scribe has inserted the comment ‘Tenor Et incarnatus’) to the penultimate stave of the facing verso, where the two voice-parts of the duo are copied consecutively. The remainder of the work (‘Et resurrexit … Amen’) is presented as a single section on the second opening. The discantus and contratenor parts occupy the upper positions on the verso and recto respectively, while the tenor fills in the lower staves of both pages. However, somewhat unusually, the contratenor voice-part of this third section requires more space than the discantus (seven staves, as opposed to five), and these two parts face each other across the upper staves of the opening. The tenor begins directly after the discantus and continues on the last two staves that remain after the contratenor on the recto, to which he is led by a diagonal line and the marking ‘finis tenoris’. The differences in the layout of the two openings onto which this piece was copied thus reflects the nature of the music itself.
In four-part works, the third and fourth voices are normally called ‘contratenor altus’ and ‘contratenor bassus’, or simpy ‘altus’ and bassus’, respectively. The discantus again appears in the upper part of the verso. The most common arrangement is then for the bassus to be placed beneath the discantus, and the altus and tenor to be placed at the top and bottom respectively of the facing recto. (The two highest-pitched voice-parts are thus aligned in the upper part of the opening, the two lowest in the lower part.) In a few cases in which the discantus (usually the part requiring most space) and the bassus cannot be fitted entirely onto the verso but there is spare space on the recto, the conclusion of the bassus appears at the foot of the recto (as on, for example, ff. 175v–176r, the first opening of Finck’s Misericors / Miserator, no. 99/99a). Occasionally the tenor is placed beneath the discantus on the verso, in which case the altus and bassus normally appear on the facing recto, usually in that order (as, for example, in no. 10, the anonymous Paradisus trinitatis, on ff. 34v–35r, and nos. 96–8, Obrecht’s Precantibus diva virgo [= Wat willen], Josquin’s Virgo prudentissima and the anonymous Summe trinitati , on ff. 170v–175r); and occasionally the altus appears beneath the discantus, with the tenor and bassus on the facing recto (as in no. 40, the anonymous Sanctus on ff. 89v–90r). The arrangement does not always remain consistent throughout a work: the first opening of the Ghiselin–Verbonnet Patrem La belle se siet (no. 34, ff. 70v–74r), for example, places the tenor beneath the discantus on the verso, with the altus and bassus in the upper and lower positions on the facing recto respectively, while the remaining openings have the bassus on the verso, the altus and tenor on the recto. In shorter works, three voice-parts may be placed on the verso, the fourth alone on the recto (as in the anonymous Michael prepositus, no. 109, on ff. 190v–191r). Rarely, the altus or bassus may begin beneath the tenor, but conclude at the foot of the facing recto. The order of the altus and bassus may be reversed, and again need not remain consistent throughout a piece which occupies more than one opening (as in no. 107, Balduinus Tectis’s Gaude dei genetrix, on ff. 187v–189r).
Of the five-voice works, one, the anonymous Miserere nostri domine (no. 22, ff. 39v–41r), occupies two openings, on both of which the discantus and tenor are placed on the verso, the remaining three parts (labelled altus, discurrens and tenor) on the facing recto. Bedingham’s Festo isti (no. 121) is divided between a single page (f. 201v; the facing recto is used for an independent composition) and an opening (ff. 202v–203r). On the single page, the two upper voice-parts (one, fully texted, with clef C3, the other, with only a text incipit, with clef C1) are placed one after the other at the top of the page, followed by the (untexted) contratenor [altus], tenor, and contratenor [bassus]. In the second section of the work, the two upper parts are placed at the top of the verso, followed this time by the tenor, then the contratenor [altus], leaving only the contratenor [bassus] to appear at the top of the facing recto; the lower part of that recto is then used for an independent composition. The only other five-voice work, Omnis mundus iocundetur (no. 178, ff. 242v–243r) places the primus discantus and the bassus on the verso, and the altus, the secundus discantus and the tenor (in that order) on the recto.
The presentation of the mostly older, full-notation repertory in Speciálník – which consists of Mass movements, motets and cantiones – is governed by conventions that differ significantly from those found in the void-notation sections of the manuscript. Some of the full-notation works – mostly the Mass movements and the cantiones – adopt the same nomenclature for the voices as the void-notation repertory: the upper voice is usually unnamed, and the lower voices are labelled ‘tenor’ and ‘contratenor’. (Not infrequently, especially when the third voice-part is in a higher range than the tenor, the latter is called ‘medium’, rather than ‘contratenor’.) Some works (principally the motets), by contrast, generally eschew this nomenclature, and instead attach a Latin ordinal number to each voice-part to indicate the pitch of that part in relation to the others: the voice-part which starts on the lowest note is described as ‘1a’ (‘prima’); any part which begins a fifth, an octave, or (more rarely) a third, above that note is described as ‘5a’ (‘quinta’), ‘8a’ (‘octava’) or ‘3a’ (‘tercia’) respectively (see below, Example 5). Other intervals, such as ‘4a’ – ‘quarta’ – may be called upon where the beginning of a voice-part is delayed by rests. Where two or more parts begin on the same note, the same ordinal number is attached to both. A few of the cantiones employ both systems of nomenclature (see, for example, no. 166, Anon., In natali domini, on f. 235v, and no. 185, Anon., In hoc anni circulo, on f. 248v).
Although the extent of the eight-stave ruling pattern in this manuscript does not coincide exactly with that of the full-notation material, there is nevertheless a strong association between them. In other words, the full-notation material is mostly written on eight staves, the void-notation material on nine. A further significant tendency in the full-notation material is that the lower voice-parts, rather than beginning on a fresh stave, often start immediately after the end of the previously written voice-part. The beginning of the texts of these second and subsequent voice-parts, especially of polytextual motets, is often marked by a small initial, which serves also to mark the beginning of the voice-part – an especially useful feature if (presumably in order to save space) the part begins mid-stave. Often, in fact, it is only these lower parts which have initials, the spaces that had been left for the initial at the head of the upper part being left empty ; presumably the intention was that the upper-voice initials would be more elaborate than the lower-voice ones, but in the event they were never executed (see, for example, no. 188, Anon., Terra tremuit, on f. 250r). The second section of a cantio, usually labelled ‘repeticio’, is usually separated from the first in each voice-part by means of a vertical rule through the stave, but does not begin with an initial. Where a polytextual motet that begins on one opening continues on the next, the voice-parts are frequently identified on the second opening not by the ordinal numbers used on the first openings (which could lead to confusion, since the harmonic configuration of the voice-parts may not be the same at the beginning of the music that appears on the second opening as it was at the beginning of the work), but by the incipits of the texts with which they began on the first opening (see, for example, no. 192, Petrus Wilhelmi, Pneuma eucharistiarum / Veni vere illustrator / Dator eya graciarum / Paraclito tripudia, on ff. 252v–254r; see also below, Example 5). Some motets have a cue at the end of each voice-part to the opening of the work, indicating that the part should be repeated, potentially ad infinitum (see again below, Example 5). Motets of this type were described in theoretical sources as rote or rotule (‘wheels’).
Throughout the manuscript, additional staves were added at the foot of a few folios (ff. 109v–110r, 118v, 119v, 126r, 130v, 158r and 159v), making ten, not nine, staves; and in gatherings XXIV–XXV additional staves were added on ff. 273v–276r, 277v and 286v–287r, making nine, not eight, staves. Sometimes these additional staves extend well into the left and right margins of the page. Often the same rastrum appears to have been used for the additional staves as for the others on those folios, but minor variations from a consistent horizontal may bear witness to the relative difficulty of adding a stave in what was intended to be the bottom margin. Fragmentary staves were also added in the bottom margin of ff. 30r
Fragmentary stave in bottom margin.
Close , 93r, 107v, 109r, 111r, 127r–v, 131r, 144r, 178r, 209v, 222r, 234r, 247r, 251v, 278v, 281v, 282r, 288r, 291r. Usually these were drawn freehand, but sometimes a ruler, or even on one occasion (on f. 222r) a rastrum, was used. In some places in gatherings I–II, VII and X–XI where a stave had been indented to allow for an initial but the indentation was not required, the stave was subsequently extended backwards across all or part of the indentation, usually freehand: see ff. 3v (stave 6)
Indentation for initial. Initial not needed so stave extended backwards freehand.
Close , 4r (6), 6v (6), 7r (6), 13v (5), 16v (5), 17r (6), 17v (5), 19r (5), 74v (5), 75r (6), 75v (6), 109r (5), 109v (5), 110r (5), 118v (5), 119r (5), 119v (5), 124r (5), 125r (5), 125v (5), 126v (5) and 127v (5); indeed, on a few pages where the left margin was relatively wide, a few other staves without indentations were extended backwards into the margin: see ff. 3v (staves 5, 7 and 8)
stave without indentation extended backwards into margin
staves without indentation extended backwards into margin
Close , 7v (4–6) and 32r (7). Fairly frequently, one or more lines of the stave were extended into the right margin of the page in order to accommodate notes, rests or custodes: see ff. 4r (stave 9), 7r (1–2), 8r (3, 5), 8v (8), 13r (5), 14v (7), 17r (6–7), 18r (8), 18v (5), 21r (3), 21v (5), 23v (6), 24r (2), 30r (3, 8)
stave extended into margin
stave extended into margin
Close , 30v (1), 40r (6), 41r (7), 67v (1), 79r (2), 89r (6), 93r (8–9), 95v (9), 103v (4), 109v (5), 110r (5), 118r (2–3, 6), 118v (10), 119r (8), 119v (10), 120r (5, 8), 122v (9), 125v (6), 128r (6), 131r (9), 135v (3), 136r (1), 158v (3), 160r (5), 164v (7), 165r (4), 165v (1, 5), 168r (7), 173r (9), 178r (9), 188v (3), 192r (7), 193r (9), 202r (1), 205r (8), 213r (8), 216r (2, 4), 219r (4), 220r (6), 222r (2–4, 7), 223v (2), 234r (7), 236r (5), 237v (3), 243r (2), 247r (6–9), 271v (7–8), 274r (7), 274v (2, 6), 276r (6), 279r (1, 6), 281r (4) and 287r (1, 4).
Practices with regard to texting and voice-names vary considerably from one scribe to the next. Scribe B, for example, frequently underlaid the sung text in all voice-parts, and provided the voice-names – usually unabbreviated or with only light abbreviation – in the indentations left by the stave-ruler or in the left margin of the page. (Sometimes they were placed so far into the margin that they were subsequently partially trimmed off by the binder.) Scribe C also normally texted all voice-parts, and only rarely wrote in the voice-names; where he did so, they appear beneath the clef and/or the first few notes of the music. Scribe A, by contrast, more often provided underlay for the discantus only, and inserted incipits for the lower voice-parts or left them untexted; sometimes in his work both the discantus and the tenor are texted, but the other lower part or parts have incipits or are untexted. He normally used more heavily abbreviated forms of the voice-names than scribe B, and typically placed them beneath the clef and first notes of the lower parts (before the incipit, where there is one). One of the minor scribes, whose hand appears only on ff. 181v–183r, omitted the first letters both of the text underlay and (in the lower parts) of the voice-names; initials for the voice-names and ordinary capitals for the text underlay were provided for the secunda pars of the tenor and bassus parts only.
The density of the textual underlay is largely dependent on the nature of the music being copied: textually longer Mass movements such as the Gloria or Credo, for example, tend to have fewer notes per syllable of text than shorter movements such as the Kyrie and Sanctus, and so appear more densely texted. In the void-notation material there is frequently some ambiguity about exactly to which notes a syllable of text should be applied. In the full-notation material, by contrast, especially the motets, the texting is often almost syllabic, and frequently pairs or small groups of shorter note values are grouped together over the syllable to which they are to be sung, preventing any significant ambiguity with regard to the music–text relationship.
A number of pages in the second part of the manuscript contain additional text, not underlaid to the music: see ff. 192r, 223v, 225r, 231r, 232r, 233r, 235v–236r, 238v–240v, 248r–v, 251v, 257r, 261r, 266v–267r, 268r, 276v–277r, 283r and 295v. The text concerned consists in all cases of further verses for strophic items, including Bohemian cantiones, where the additional text is for the verse section, while the second section of music (repeticio) is repeated with the same text after each verse, functioning as a refrain. The additional text is normally written onto staves left completely or partially empty after the music had been copied, often using the lines of the staves as guidelines, or in empty space left at the foot of the page. In one case (f. 266v
Lines drawn to accommodate text.
Close ) additional lines were drawn in at the foot of the page as guidelines for the extra verses of text; similarly, a half-line was drawn between two staves on f. 268r for this purpose.
Most of the scribes of Speciálník use on occasion a somewhat larger and heavier script for the voice-names, and sometimes also for text incipits, which distinguishes these items from the textual underlay. This feature is particularly striking in the work of a minor scribe whose hand appears only on ff. 170v–171r. Composer ascriptions are provided relatively infrequently in this manuscript, and are normally placed in the top centre of the page, often again using a heavier script. On one occasion (on f. 43v) a rebus is used for the syllable ‘-la’ in the name ‘Agricola’. Red ink was used on one opening (ff. 116v–117r) to distinguish the voice-names and the rubric ‘Verte 10 fol’, and on f. 172r for the voice-name ‘[B]assus’, to which the ‘B’ was added as a (brown) initial. Elsewhere, it was used to distinguish the alternative text (‘Sacerdotes incensum domini’) of Tourout’s Recordare virgo mater (no. 118) on ff. 196v–198r, and for all of the (multiple) text underlay and voice-names of Nobis natus hodie (no. 177) on the lower staves of f. 242r.
The clef forms used in conjunction with full notation are relatively simple forms of the letters C and F, but in the void-notation section of the manuscript are more varied, and not always consistent within the work of a single scribe. Mensuration signs are usually provided, and may follow the clef, or appear above or below the stave, or even in the left margin. Custodes take the form in void notation of a wavy or zigzag line terminated by an upward diagonal stroke, and in full notation of a narrow oblong with a down-stroke on the right side, usually longer than that used for longae. Signa congruencie have the standard form throughout of three dots arranged in the form of a pyramid, with an upward curlicue rising from the upper dot; the sign is sometimes inverted, especially when applied to a note in the upper part of the stave. Pauses adopt the form of the top half of a semicircle when placed above the note, the bottom half when placed beneath, though some scribes characteristically place the sign both above and below the note. One or more dots may be placed within the semicircle. More elaborate forms of the pause may occasionally be observed (see, for example, those of the minor scribe who copied the music and text on ff. 80v–83r). A further feature of the music notation in Speciálník is the occasional use of numerals above some notes, indicating their duration as a multiple of a lower note value. These are restricted in this source to lower voice-parts, and are found in the context of both void notation (see ff. 27v, 28v, 29r, 30v, 31r, 32r, 90v–91r, 110r, 132r, 135r, 136r–v, 137v, 158r, 199r, 205v and 207r) and full (see ff. 211r, 218v, 253r). On one occasion (f. 15v) pairs of dots were used to indicate semibreves that should be altered and breves that should be imperfected, and a group of three dots to indicate a breve that should not be imperfected, in order to clarify the rhythmic structure of two ligatures in a discantus part.
The manuscript is not heavily corrected, and it appears that most of the corrections were carried out by the scribe who made the error in the first place: where insertions were made or substitute passages copied in (as, for example, on the last stave on f. 53v), they are normally in the hand of the scribe that copied the remainder of the work. (An exception may be the substitute passage inserted at the end of stave 5 on f. 175v.) A short passage erased on f. 6v (stave 1) is additionally annotated ‘vacat’; three notes omitted on f. 68r (stave 5) were inserted on a small stave drawn over the main one; a long erased passage on f. 123v (stave 7) was overwritten by the scribe who copied the erroneous passage; the first few notes of the tenor on f. 125v (stave 4) were erased and the revised version written to the left of the clef; and so on. Elsewhere, isolated superfluous notes and stems have been lined out or erased (or both), and notes that were copied as full notes but should have been void are marked with an inverted ‘v’. A few annotations were made by a later (nineteenth- or twentieth-century?) hand on ff. 218v–223r.
The various scribes who contributed to Speciálník frequently reserved spaces for decorated initials to be filled in throughout the manuscript. They did so either by indenting the staves or by indenting the music notation and the text underlay, leaving the beginning of the staves empty. Often – especially in the full-notation sections of the manuscript -- they omitted the first letter of the underlay or of the voice-name, as appropriate, leaving it to be supplied by the initial. (In the void-notation sections, the first letter of the underlay was omitted more frequently at the beginning of Patrem and Sanctus settings than in those of other texts, and the first letter is omitted more often in the discantus than in the lower parts.) Elsewhere, however, the first word of the text underlay or voice-name was written in full by the text scribes, despite the fact that they also allowed space for a decorated initial. There is no strong relationship between omission of the first letter of the text or voice-name and the provision of an initial: initials were frequently added where the first letter of the text was already present (indeed the only initial in the first gathering – on f. 3r – occurs where the first letter of the discantus underlay is already present), and they were frequently omitted where both a space had been left and the first letter of the text omitted. So although many spaces have been left empty without any further decoration, others were filled with initials that are more conspicuous than the ordinary capitals which they replace or duplicate
Three types of penned initials in brown ink have been employed to decorate the codex (in addition to those found in the last section, ff. 292–303, which contains only chant and received plain initials, alternately in blue and red). The initials are in general one stave high, but occasionally occupy two (see ff. 45r, 46v, 52v
Initial occupies two staves.
Close , 139v) or even three staves (f. 140r).
1. The first type of decorated initial appears sporadically, mainly towards the beginning of the manuscript (ff. 3r, 32v–33r, 34v–36r, 39v–40r, 43v–46v, 49v–50r, 51v–54r, 55r, 72v–73r,139v–140r, 143v–144v, 145v, 146v–147r, 148v–149r, 171v (lower initial only), 172r and 241v–242r). These initials consist of large penned letters which are sharply drawn in brown. Void areas contrast with dark areas, and the letters seem to float in front of the page although they appear closely attached to the staves. The decorative vocabulary refers to traditional calligraphic techniques insofar as the letters are built up of lines that gradually increase or decrease in width. Instead of building patterns determined by the broad, flat nib of a pen, however, the body of the initial is constructed three-dimensionally: void areas are contrasted with filled shapes of equal breadth that function as shadows, so that the letters appear as if they were lit from the left. The organically entwined forms seem inspired by sprouting branches that create not only a means of distinguishing between the different letters of the alphabet but also for providing a wide variety of designs for each letter. The initials appear quite individual and they impress through their inventiveness in terms of construction and variety.
The distinctive character of this type of decoration allows us to distinguish three separate hands at work. The main tranche of work, at the beginning of the manuscript, was carried out by a hand that was arguably the most creative of the three in terms of exploring various shapes. A second hand executed the type 1 initials on ff. 72v and 73r and reinforced the association with tree branches by rendering the lines enclosing the white shapes with foliate forms and wood knots. The third hand, responsible for the rest of the type 1 initials after f. 73r, is close to the main decorator, but the anatomy of the letters often appears simpler, and the spatial effects of the black and white contrasts remain rather flat.
This unusual type of decoration is also used for a ‘space-filler’ (f. 36r
Close ). Line-fillers are absent in the manuscript. The empty space at the end of the contratenor (indicated in Czech as ‘Gontrassek’) of the three-part Bud bouhu chwala is, however, filled with abstract decoration that is in keeping with the style of the decorated initials.
2. The second type of penned initial appears mainly in the second half of the manuscript (ff. 183v, 184v, 186v–187v, 189v, 190v, 191v, 195v, 196v, 204r, 210v–211r, 214v, 215v, 216v, 219r, 220v–221v, 222v–223r, 224r, 225v–226r, 227r–228r, 229r–230r, 231v–232v, 234r–235r, 238r–241r, 245v
Second type of penned initial.
Close , 249r–251r, 252r–253r, 254r–256r, 257r–259v, 261v–262v, 264v–267v, 268v, 271r–273r, 277r, 279v, 283r and 285v). The bold letters are dark and solid in appearance. The traditional use of a pen with a broad, flat nib creates the distinctive ‘thick-and-thin’ effect when moved over the page while being held at the same angle. Because of their typographical aspect, this type of letter survived in printed books. (The initials on ff. 78v and 79r are closely related, but try to imitate loosely the black-and-white contrast of the first type of initial.)
3. The third type of initial is similar in its technique to the second, and is employed mainly in the last two thirds of the manuscript. The letters are, however, thinly drawn and lack mostly the graphic character of the second type. At the beginning of the page (and sometimes also for voice names) they are often rendered in double lines to give them a stronger appearance (the most elaborate ones occur on ff. 89v, 92r–v, 99v–102r, 164v, 165v–166r, 167r
third type of initial
third type of initial
Close , 168r–169v, 170v–171r, 171v (upper initial only) , 176v, 182v, 193v, 200v–201r, 202r, 203r, 205v, 206v, 207v, 209v, 211v, 215r and 257r). They are also often used in a subordinate form, in which they are smaller and drawn more simply with a single line. On a few occasions, these also appear in combination with the previous type of initial, but at a hierarchically lower position.
The Speciálník Codex is a substantial collection of polyphony compiled, probably in Prague, over a period of some 15 years at the end of the fifteenth century. It contains both music that belongs to a relatively old, local tradition, that continued by convention, even at this late date and considerably beyond, to be copied in full notation, as well as a more modern repertory that was partly imported from Western European, partly composed locally (to some extent under the influence of musical developments in Western Europe), and was copied in void notation. The manuscript was decorated only sporadically, and many reserved places have been left empty. To this material a gathering of chant hymns was added at the back of the manuscript. An alphabetical index covering most of the polyphonic music was compiled, starting on the last page of the chant gathering.
The fact that relatively few paper-types are found in this manuscript is one indication that it is not a composite source, put together from gatherings copied in different places, but rather the result of a coherent, sustained and cumulative process of growth. Although a large number of scribes, probably about 25, contributed to the manuscript, one in particular predominates in both the full-notation and the void-notation sections of the manuscript. Four other scribes made significant contributions, but the remainder copied only isolated works or small groups of works.
The full-notation section of the manuscript, whose repertory consists mostly of Mass movements, cantiones and polytextual motets, not only looks very different from the void-notation section, it also follows a different set of conventions with regard to mise-en-page, especially in the motets, where all voice-parts are normally fully texted, and where, typically, the voice-parts are labelled with an ordinal number which expresses the intervallic relationship between its first note and that of the lowest-sounding note at the beginning of the motet. Normally, one voice begins immediately after the previous one ends, rather than beginning a new stave, and there is less vacant space on the page. The full-notation Mass movements and cantiones, on the other hand, follow more closely the conventions of the void-notation repertory, with distinct tenor and contratenor parts that may begin on a new stave and are frequently left untexted or with only textual incipits; additional strophes of text for the versus of the cantio are, however, often written out at the foot of the page, or in spaces left after one or more of the voice-parts.
While only two scribes contributed to the full-notation material in this manuscript, many more worked on the void-notation material, and many elements of mise-en-page in the latter are differently inflected by individual scribes. The clef forms used in conjunction with full notation are relatively simple, and other peripheral signs used within the music (i.e. signs other than notes or rests) are relatively few. Within the void-notation section of the manuscript, by contrast, the clef forms are more elaborate and more varied, and other signs (signa congruencie, signs that lead to the continuation of a voice-part, correction signs, etc.) are more frequently encountered. The disposition of the voice-parts on the opening is more varied in the void-notation material, as is the extent of texting (the discantus is normally fully texted, but lower voices may be untexted, have incipits only, or also be fully texted).
Example 1: No. 181. Petrus Wilhelmi, ff. 245v–246v
No. 181, Petrus Wilhelmi’s Presidiorum erogatrix (ff. 245v–246v), illustrates two of the ways in which three-part music is typically laid out in the void-notation sections of this manuscript. Copied here by scribe A, the work is in three distinct partes, each of which concludes with a clear cadence on a long in each voice-part. The cadence at the end of the first section is further marked in the discantus by a vertical line through the stave
The cadence at the end of the first section is marked in the discantus by a vertical line through the stave.
Close . The first two sections, ‘Presidiorum erogatrix’ and ‘Tota pulcra es’, are displayed on the first opening, the third, ‘O virga Yesse’, on the following verso; a new piece begins on the facing recto. On the first opening, the discantus voice for the first two sections
The discantus voice for the first two sections appears in its normal position at the top of the verso folio.
Close appears in its normal position at the top of the verso folio, where it fills the first six staves; the second section follows on directly after the first, beginning (economically) mid-stave, rather than starting a new one. The contratenor part for the same two sections
The contratenor part for the same two sections is placed directly opposite to the discantus.
Close is placed directly opposite to the discantus, and again occupies the first six staves. The first two sections of the tenor
The first two sections of the tenor, placed at the foot of the opening and divided between the recto and verso.
The first two sections of the tenor, placed at the foot of the opening and divided between the recto and verso.
Close , which occupy a total of only five staves, is then placed at the foot of the opening, and divided between the verso and the recto: the first section is placed on the last three staves of the former, the second on the last two staves of the recto. (A blank stave
a blank stave separates the contratenor and tenor sections
Close is left immediately beneath the contratenor part of the second section, separating it off from the second section of the tenor.) A diagonal line
Diagonal line drawn across the gutter of the opening.
Close was drawn across the gutter of the opening, connecting the last stave of the verso with the stave on which the voice-part is resumed on the recto. (Elsewhere in the manuscript, as noted under ‘Copying of text and music’ above, the link is often made by means of a pair of cues or signa congruencie.) The third section of the piece
The third section of the piece.
Close is roughly half the length of the first two combined, and could therefore be completely accommodated on the following verso, where the discantus part
Close appears at the top, followed by the tenor
Close , then the contratenor
Close , each more-or-less filling three staves. Note that this layout could have been used for the first two sections as well, with all three voice-parts of the first section placed on f. 245v, and all three voice-parts of the second on f. 246r. The scribe’s preference, however, was to combine the first two sections and display them together across the opening. This was the most commonly used method of presenting two consecutive sections of music, both generally during this period and specifically in this manuscript. For an exception to this rule, though, see no. 126 (Walter Frye’s Ave regina, ff. 204v–205r), where all four voice-parts of the prima pars appear on the verso, all four parts of the secunda pars (‘Funde preces’) on the recto.
All three voice-parts of this work are fully texted, and although much of the underlay is syllabic, the locations of brief melismas are consistently made clear. The decoration with initials, which are of the second type discussed above, is strictly hierarchical and helps to visualise the layout of the three voices arranged in three sections or partes. A large bold initial ‘P’
A large bold initial 'P'.
Close appears at the beginning of the discantus and marks the beginning of the work. The beginning of the tenor and contratenor
Initial at the beginning of the tenor.
Initial at the beginning of the contratenor.
Close are also indicated by large initials which appear, however, very thin and smoothly drawn. Small initials
Close , hardly bigger but more elaborate than the text underlay, appear eventually at the beginning of sections 2 and 3 in all three voice-parts. Voice-names for the two lower parts
Voice-names for the two lower parts appear in the margins, and have been partially trimmed off.
Voice-names for the two lower parts appear in the margins, and have been partially trimmed off.
Close appear in the margins, and have been partially trimmed off. Every stave (except the one left empty on f. 246r) has a clef (C2 for the discantus
Every stave has a clef (C2 for the discantus)
Close , C4 for the tenor
Every stave has a clef (C4 for the tenor)
Every stave has a clef (C4 for the tenor)
Close and F4 for the contratenor
Every stave has a clef (F4 for the contratenor)
Close ) and a B flat signature (doubled in the discantus and contratenor). The mensuration sign (cut C) appears in all three parts, but is placed, unusually, not on, but above, the stave in the discantus and contratenor, and in the left margin in the tenor part.
Example 2: No. 60. Anonymous, Patrem from Missa sine nomine, ff. 109v–112r
No. 60 (ff. 109v–112r), the Patrem from an anonymous Mass cycle (lacking the Agnus dei) preserved uniquely in this manuscript, of which it occupies ff. 105r–116r, and copied by scribe B, illustrates both the simplest form of the most commonly used four-part layout, and some of the ways in which this layout could be modified by scribes.
Despite the fact that the scribe reserved spaces to be filled with initials
spaces to be filled with initials.
spaces to be filled with initials.
Close 140, none have been carried out. Of the three openings on which the Mass movement is displayed, the last (ff. 111v–112r) contains the shortest section of music (‘Confiteor … Amen’), and therefore presented the fewest problems to the scribe, with the result that it has the simplest form of layout. The pages were pre-ruled with nine staves each, of which the first and sixth were indented on the verso, the first and fifth on the recto. The four parts each occupy a corner of the opening: the discantus
Close at the top of the verso, the bassus
Close beneath it, the altus
Close at the top of the recto, and the tenor
Close beneath that. Each voice-part occupies three staves, leaving three empty staves on each page: on both pages, two staves were left empty between the voice-parts, and one at the foot. The lower stave indentation on the verso coincides with the beginning of the bassus part, but on the recto, where the stave-ruler had placed the indentation one stave higher, the indented stave is one of those left blank, and the tenor begins on an unindented stave. (The beginning of the tenor part, nevertheless, is indented from the margin by about the same distance as the start of the altus part above it, leaving a stretch of empty stave to the left of the clef.) The voice-name of the altus part
the voice-name of the altus part written in the space left by the indentation.
Close has been written in the space left by the indentation at the top of the recto page; that of the bassus
the voice-name of the bassus written in the margin
Close , however, was written further left, in the margin. (This may actually have been the scribe’s preferred place for the voice-names, thus leaving the indentations free for decorated initials, but in fact there was insufficient space for him to do this on the recto.) The voice-name of the tenor
the voice-name of the tenor written beneath the fragment of empty stave that precedes the music.
Close is written beneath the fragment of empty stave that precedes the music, while the space left by the indentation at the beginning of the discantus
space left by the indentation at the beginning of the discantus is left blank
Close is left blank. The discantus, altus and tenor are all fully underlaid with the text, though apparently somewhat hurriedly, especially in the lower parts: the words ‘et exspecto’ were omitted in the altus part. The bassus has the incipit ‘Confiteor’, then no further text until ‘et vitam venturi seculi’, which coincides with a musically important point of imitation whose specific relationship to the rhythm of the music the scribe probably felt it important to indicate. All the voice-parts are fully cleffed (C2 for the discantus, C4 for the altus and tenor, F4 for the bassus)
Close , and fully provided with mensuration/proportion signs
Close and custodes
Working backwards through the composition, the second of the three openings (ff. 110v–111r) presents a somewhat longer section of continuous music, setting the text ‘Et resurrexit … et unam sanctam et apostolicam ecclesiam’. This time the stave-ruler indented the first and fifth stave on the verso, and the first and sixth on the recto (the reverse of the situation on the following opening). For each voice-part, the music occupies not less than four staves. On the verso, the upper five staves are devoted to the discantus, the lower four to the bassus. The discantus did not, however, require all of the fifth stave, and the last few notes are placed not at the beginning of that (indented) stave, but in the centre, or even towards the end of the stave
the last few notes are placed not at the beginning of that (indented) stave, but in the centre, or even towards the end of the stave
Close ; the bassus, by contrast, exactly fills the four staves allocated to it
the bassus fills the four staves allocated to it.
Close . The indentation of the fifth stave is effectively ignored, and the part-name of the bassus is written in the margin
part-name of the bassus is written in the margin.
Close , as on f. 111v. On the recto page, though, the altus exactly fills the five staves allocated to it
the altus exactly fills the five staves allocated to it.
Close , and the tenor more than fills the four remaining
the tenor more than fills the four remaining staves.
Close ; it was completed on a fragmentary stave drawn freehand at the bottom right corner of the page
fragmentary stave drawn freehand at the bottom right corner of the page,
Close . The altus and tenor voice-names appear in the spaces left by the indentations of staves 1 and 5 on that page. Only the discantus and tenor are fully texted on this opening; the bassus and altus each have only single incipits. All the voice-parts are fully provided with clefs and custodes, as before, but only the discantus and bassus (the voices notated on the verso) have the mensuration/proportion sign (cut C) that applies to this section of the music.
The first opening of the work, ff. 109v–110r, presents the longest section of music, setting the text ‘Patrem omnipotentem … passus et sepultus est’, and confronted the scribe with the greatest problems. The stave-ruler indented the first and fifth staves on both pages
first and fifth staves indented
Close . However, the music of each voice-part required a full five staves, so the bassus (at the foot of the verso) and the tenor (at the foot of the recto) begin not on the fifth, but on the sixth stave. The spaces created at the beginning of the fifth staves on both pages were filled in, freehand, with stave lines
spaces created at the beginning of the fifth line were filled in, freehand, with stave lines.
Close . Both the music and the text were more economically spaced on this opening than on the next two, notably towards the end of the discantus part. Even so, the discantus overflows at the end of the fifth stave: a few additional notes were entered onto a fragmentary stave (of which only the three middle lines were drawn) in the right margin of the page
a few additional notes were entered onto a fragmentary stave (of which only the three middle lines were drawn) in the right margin of the page.
Close , and the final notes were placed at the end of the sixth stave of the page
the final notes (of the discantus) were placed at the end of the sixth stave of the page.
Close , thus encroaching on the bassus’s space. (The last notes of the discantus are marked off from the bassus part by a double vertical line through the stave
double vertical line through the stave
Close ; the former’s attention is also presumably caught by the custos which follows the last note on that stave he is expected to perform, which directs him to the beginning of the next stave, while the latter is directed to the beginning of the following stave by a custos
Additional staves were ruled for the conclusion of the bassus and tenor voice-parts at the foot of the verso and recto pages respectively
additional stave ruled for the conclusion of the bassus.
additional stave ruled for the conclusion of the tenor.
Close ; these were drawn with a ruler, rather than the rastrum which had been used to execute the other staves on the opening. In the case of the bassus, the additional stave begins further left than the original staves on this page. The beginning of the altus, in the upper part of the recto folio, is indented from the left margin, but at the end of the part the three middle lines of the stave are extended into the right margin, as in the discantus on the facing page. The altus voice-name
Close was placed in the space left by the indentation of the top stave on f. 110r; the bassus voice-name
Close was placed in the margin (as on ff. 110v and 111v). The beginning of the tenor’s music, despite the absence of stave indentation at that point, was slightly indented from the stave edge, as on f. 112r, but in this case by a smaller distance; the tenor voice-name was placed beneath the stave
tenor voice-name placed beneath the stave.
Close , before the music begins. As on ff. 110v–111r, only the discantus and tenor are fully texted; the bassus and altus have only initial incipits. And as on ff. 110v–112r, all the voice-parts on this opening are fully provided with clefs and custodes, and all have the mensuration sign (cut C) that governs the whole of this section of the work.
The reason for the disparity between the three openings onto which this piece was copied is not difficult to surmise. The setting really falls into only two sections, ‘Patrem omnipotentem … passus et sepultus est’ and ‘Et resurrexit … Amen’, the end of the first being marked by a held cadence in all four parts (notated as a longa with a pause in all but the altus part, which has a simple breve). As we have seen, the scribe had difficulty fitting all of the music of the first section onto an opening; it may be that his exemplar was of a larger format, or at least was ruled with more than nine staves per page. (At any rate, the odd number of staves on the page in Speciálník was one of the factors that hampered the scribe, the music being for an equal number of voice-parts, each containing a similar number of music symbols.) The second section is only slightly longer than the first (129 breves as compared with 125), but includes a passage (23 breves) in sesquialtera proportion, which occupies (theoretically) 50% more space than a passage of the same length in duple time would. Presumably this second section of the work also occupied a single opening in the (larger-format?) exemplar. However, seeking to avoid the greater challenge of fitting all that music into the smaller space available to him, the Speciálník scribe decided to split the second section into two. However, there is no point in the section comparable to the end of the first section at which the voices cadence and come to rest. The scribe therefore split the section at the beginning of the triple proportion section. The music is continuous, however, so this expedient requires the performers to turn the page without interrupting the flow of the music.
Example 3: No. 30. Philipon Basiron, Sanctus from Missa de Franza, ff. 52v–55r
Philipon Basiron’s four-part Missa de Franza (nos. 27–30), copied by scribe C, occupies most of the fifth gathering (ff. 45–56) of Speciálník. The title ‘Officium Philippon’ appears on the first recto of the gathering, together with the first section of the Kyrie; and the rest of the Kyrie, and the Gloria, Credo and Sanctus, follow in order. (There is no Agnus dei.) The Sanctus (no. 30) occupies three consecutive openings (ff. 52v–55r). The staves on the last complete opening of the gathering (ff. 55v–56r) were left empty, and a different scribe began the next work, Compère’s Sile fragor (no. 31, ff. 56v–58r) on the last verso, continuing on the first folios of gathering VI
The music of the Sanctus falls into four main sections: (1) ‘Sanctus … sabaoth’, (2) ‘Pleni … Gloria tua’, (3) ‘Osanna’ (concluding with a passage in cantus coronatus for ‘in excelsis’), and (4) ‘Benedictus … in nomine domini’. The four voice-parts were copied in the four corners of the opening, with the discantus at the top of the verso, the bassus beneath it, the altus at the top of the recto, and the tenor beneath that, a disposition that is maintained on all three openings. (None of the voice-parts is labelled on the first opening, but the bassus and tenor are labelled on the second, as is (unusually) the discantus on the third.) Being shorter than the first and fourth sections of the work,the second and third are both presented on the second opening, of which the verso is the only page in the piece that has no empty staves. Elsewhere in the setting, one stave is always left empty between voice-parts that appear on the same page, and one or two staves are left empty at the foot of the page. (On the recto of the second opening (f. 54r)
the right-hand end of the last stave has been used for the conclusion of the tenor part of section 3.
Close , however, the right-hand end of the last stave has been used for the conclusion of the tenor part of section 3; even here, though, most of that stave has been left empty.)
Folio 53v also demonstrates one method by means of which a scribe could correct an error that he (or an earlier scribe whose work he was copying) had made. In the bass part for the ‘Osanna’ section, he had written one superfluous note
one superfluous note (within a ligature), marked by signa and the word 'vacat'.
Close . Ordinarily, this could simply have been lined out or erased. However, the note in question occurs in the middle of a ligature, so its removal could have caused an ambiguity. Instead, the scribe wrote a signum three notes before the superfluous note, and another (of a different form) several notes further on, thus defining a passage in the music he had copied. Beneath this passage he wrote the word ‘vacat’, signalling that the passage should be omitted. Then, after the bassus notes for the ‘in excelsis’ chords, he wrote out the correct version of the passage in full, marking the first and last notes with signa that match in shape those in the erroneous passage on the stave above, adding the following rest and a custos for the next note for good measure, ensuring an accurate pickup.
The music is mostly in perfect tempus, with contrasting diminished imperfect tempus for the ‘Osanna’ and ‘Benedictus’ sections. All the voice-parts are fully provided with mensuration signs, except the altus and tenor on the last page (f. 55r)
the altus and tenor are not provided with mensuration signs.
Close . The discantus, altus and tenor (but not the bassus) have single flat signatures throughout, and all voice-parts are fully supplied with clefs and custodes.
None of the staves in this gathering of the manuscript was indented. Nevertheless, the clef, flat signature, mensuration sign and first notes of each voice-part in the Sanctus were delayed a short distance from the left edge of the stave, creating space on which an initial could be superimposed over the first section of the stave. (At the beginning of the discantus part for the Sanctus, the beginning of the second stave was alsoleft empty, leaving space for a larger initial.) Most of these spaces in the Sanctus were subsequently provided with an initial, the exceptions being those at the beginning of the discantus and bassus on the last opening (f. 54v)
space left for initials, but none executed
Close . (The decoration was executed on double folios: f. 39 forms a bifolium with 54, and while the inner side (ff. 39v/54r) received decoration the outer side (ff. 39r/54v) did not). The first letter of the text or voice-name, as appropriate, had been omitted by the text scribe, and in general the initials inserted later supply the missing letters. Thus there are four different letter ‘S’s
four different letter 'S's on the first opening.
four different letter 'S's on the first opening.
Close on the first opening, ff. 52v–53r, for ‘Sanctus’. The second opening, ff. 53v–54r, has two ‘P’s for ‘Pleni’
two 'P's for 'Pleni'
two 'P's for 'Pleni'
Close (for the discantus – where the text scribe had erroneously begun by entering the text ‘[S]Anctus’, then continued with ‘Pleni’ – and the altus), four ‘O’s for ‘Osanna’
four 'O's for 'Osanna'
four 'O's for 'Osanna'
Close , a ‘B’ for ‘Bassus’
'B' for 'Bassus'
Close and a ‘T’ for ‘Tenor’
'T' for 'Tenor'
Close . The final opening, ff. 54v–55r, has only, on the recto page, a ‘Q’ for ‘Qui venit’
'Q' for 'Qui venit'
Close for the altus (the discantus and altus parts have rests for the word ‘Benedictus’, which is sung as a duo between the tenor and the bassus), and what appears to be an erroneous ‘P’ (rather than ‘B’) for ‘Benedictus’
an erroneous 'P' (rather than 'B') for 'Benedictus'
Close in the tenor. On f. 54v, the initials ‘D’ and ‘B’ required to complete the voice-name ‘[D]Iscantus’ and the first word of the text, ‘[B]Enedictus’ in the discantus and bassus parts respectively are missing.
The elaborate penned initials are three-dimensionally constructed through sharp chiaroscuro effects. Their vegetal shapes provide an endless repertory for the construction of a variety of letters in multiple shapes. As already mentioned, the opening of the Sanctus (ff. 52v-53r) displays the letter ‘S’ four times: the first (and largest) character appears highly entwined, the one below follows more closely a simple ‘S’ form, and the third letter exchanged its curvy shape for a more rectangular one, while the last one seems to sprout around a central branch. The following letters show a similar level of diversity, and one in particular deserves attention: the ‘B’ in the bassus part on f. 53v incorporates a scroll inscribed with the date ‘Anno domini millesimo cccco vijo’ (1407)
a scroll inscribed with the date ‘Anno domini millesimo cccco vijo’
Close . (Only two letters are similarly decorated with a scroll, another ‘B’ which introduces the bassus on the opening for the Mass on f. 45v, and a ‘T’ which introduces the tenor on f. 242r, although the banderols have been left empty in both cases.) On the basis of other evidence about this manuscript (see above, under ‘Place and Date of Origin, Production Context, and Provenance’), this date precedes the compilation of the manuscript, and its significance is uncertain. It is possible that the decorator copied the letter from a model-book which contained the date which he transcribed into the new context. It is also possible that his intention was indeed to include the year when he decorated the manuscript although he miswrote it for some reason. Its significance remains therefore uncertain.
Example 4: Nos. 88–93. Anonymous, Missa cum proprio, ff. 164r–168v
Folios 164r–168v contain an anonymous, three-part, incomplete but integrated, set of Mass movements, including Propers, for Advent (nos. 88–93): the introit Rorate celi desuper with its verse, Celi enarrant, a Kyrie, the gradual A summo celo, the Alleluia with verse Prophete sancti, the sequence Mittit ad virginem, and a Sanctus, all probably copied by a single scribe. (One or more later scribes may, however, have inserted clefs on some staves on which the original scribe had omitted them: see f. 164r, stave 5; 165r, stave 3; f. 165v, staves 2–5; f. 166r, stave 5; f. 166v, staves 2–5; f. 167r, staves 1–2; f. 167v, staves 2–5 and 9; f, 168r, stave 2. The following staves were also left without clefs: f. 164r, staves 8–9; f. 165r, stave 7; f. 167r, stave 9; f. 167v, stave 8; f. 168v, staves 1–5.) These works begin a new gathering of the manuscript, and share a number of features of musical style and of layout that are unusual and distinctive. Each of the works is based on a plainchant cantus firmus located in the tenor voice, sung in equal note values and presented in Bohemian chant notation. Above and below the tenor, discantus and bassus parts respectively, presented in void mensural notation, ornament the chant in more or less lively fashion. (The discantus is for some reason labelled ‘altus’ in the introit, but ‘discantus’ elsewhere.) The tenor part has adifferent appearance from the outer parts (even the text underlay of the tenor is in a heavier style than the text incipits of the outer parts), but may nevertheless have been copied by the same scribe, using different equipment (notably, a broader nib) and a different set of conventions (in terms, for example, of clef shape and custos) than when he was copying the outer parts. Even the text underlay of the tenor is in a more Gothic style than the text incipits of the outer parts, perhaps because this was seen as better complementing the chant notation, which is ‘heavier’ than the mensural notation used for the discantus and bassus.
The mise-en-page in this set of Mass music often reflects the style of the music, in so far as the discantus is placed at the top of the page, the tenor in the middle, and the bassus at the bottom, though the tenor is placed last, after the bassus, in some of the later items. In general in these pieces, only the tenor is fully texted, the underlay being fairly carefully indicated beneath the chant to which it must be sung. The two outer parts usually have only incipits; often, though, the syllables of these first few words of the text are spaced out beneath the notes to which they are to be sung, the remainder of the music being left untexted. Of particular interest with regard to the mise-en-page on these folios is an unusual lack of concern to present each item on a single opening. As a consequence of beginning the series of works on the front recto of this new gathering of ruled paper (when some scribes might have turned the page and begun on the first verso), the scribe repeatedly came up against the problem of having to split a composition into two sections, separated by a page turn. Furthermore, continuous sections of music sometimes run on from the foot of a verso to the top of the facing recto.
The introit (without its verse) occupies eight staves on f. 164r; the outer parts occupy three staves each, the tenor only two (the chant notation used for it being more compact than the mensural notation used for the other parts), and the stave beneath the tenor is left empty. The tenor is preceded by the six-notation intonation ‘Rorate’, presented in square notation and without text. (Presumably this would have been sung by a solo cantor, not by the same singers as the tenor part of the following polyphony.) The verse is only about half as long as the antiphon of the introit, and occupies the first five staves of f. 164v: the outer voices occupy one and a half staves each, while the tenor starts on stave 3 and concludes at the end of stave 2, after the end of the discantus part. The whole item (both antiphon and verse) could easily have been placed across an opening; as it is, however, a page turn is required between antiphon and verse.
The Kyrie, which starts with a large penned initial of type 3, then occupies most of the rest of this opening (ff. 164v–165r)
large penned initial of type 3
large penned initial of type 3
Close . It consists of four sections of music: ‘Kyrie primum’, ‘Criste primum’, ‘Criste secundum’ and ‘Kyrie ultimum’, probably indicating that alternatim performance beginning with a chant Kyrie was intended:
polyphony (‘Kyrie primum’)
polyphony (‘Criste primum’)
polyphony (‘Criste secundum’)
polyphony (‘Kyrie ultimum’)
The tenor for the two polyphonic Christe sections is identical, and is written out only once, and labelled ‘Criste primum secundum’. Although the tenor starts at the beginning of the stave without any decoration, the second and third sections are marked with large thin initials. Despite mensural elements that are introduced towards the end of the tenor voice of the Kyries, the whole of the tenor part for this work occupies only two staves (staves 2–3 on f. 165r). The beginning of the bassus, on the other hand, is indented. No initial ‘K’ is provided at the beginning, and the first word, ‘Kyrie’, has been written out in full by the text scribe; the space that precedes it has been left empty. The discantus and bassus parts for all sections except the second polyphonic Christe are each contained on a single stave, though the discantus of the ‘Criste primum’ and the bassus of the ‘Kyrie primum’ both conclude on freehand extensions of the stave into the right margin. The outer parts of the ‘Criste secundum’, however, are more elaborate than those of the ‘Criste primum’, and occupy two staves each. The four sections of the discantus part occupy the bottom four staves of f. 164v and run on to the top stave of f. 165r. The tenor, as already mentioned, occupies staves 2–3, and the bassus has staves 4–8, leaving stave 9 empty.
The gradual, A summo celo, and the Alleluia (without its verse, Prophete sancti, which appears on the following opening) occupy ff. 165v–166r. The gradual starts with a large initial ‘A’ penned in double lines. The verse of the gradual was omitted: it is, in fact, identical to that of the introit (‘Celi enarrant’), which may be reused here. A little economy was required in order to fit the discantus part of the gradual and Alleluia onto the first five staves of f. 165v: the Alleluia begins towards the end of the fourth stave with a smaller penned initial, rather than beginning a new stave, and the fifth stave is extended into the right margin in order to accommodate the last few notes of the Alleluia. Whereas in the previous movements of this cycle the tenor has followed the discantus, here the bassus is placed second. Its beginning is marked with an ordinary capital letter in the same position as the text underlay. The bassus of the gradual occupies the last four staves of the verso; the remainder of the last stave is left empty; and the Alleluia appears at the top of the facing recto, again with an inconspicuous capital. (As in the Kyrie, then, the bassus might be said to run on from the foot of the verso to the top of the recto, though here the material on the recto is actually the beginning of a separate liturgical item.) The tenor parts of the gradual and Alleluia, which require only three staves (two plus one), are then located in the lower part of the recto and start with large penned initials similar to that at the beginning of the discantus. An empty stave was left above the bassus of the gradual, another between that and the Alleluia, and two at the foot of the page. Again, only the tenor has the full text of the gradual underlaid; the discantus and bassus have the first three words as an incipit, followed by the remaining two words of the first metrical line (‘egressio eius’) underlaid to the music at the relevant point. In the Alleluia, the outer voices have the only word of the text as an incipit, while in the tenor the scribe has deliberately placed the four syllables beneath the first four ligatures of the music.
As with the introit, the verse of the Alleluia (Prophete sancti) appears not on the same opening as the main part of the work, but over the page, on the following verso
the verse of the Alleluia (Prophete sancti)
Close , where it becomes somewhat entangled with the beginning of the sequence (Mittit ad virginem / Fortem expediat). The layout of this opening is not dissimilar to that of the preceding one, though its decoration varies. The discantus part of the Alleluia verse, then of the first polyphonic verse of the sequence (only the ‘b’ verses of which are set polyphonically, leaving the ‘a’ ones to be sung as plainchant in alternatim fashion), occupies the first five staves and starts simply with a capital. The bassus of the Alleluia verse follows, on the next three staves. The bassus for the first polyphonic verse of the sequence then begins on the last stave of the verso and runs continuously on to the first of the recto. As with the discantus, the bassus is introduced only by simple capitals. The remainder of the first stave on the recto and the whole of the second stave on that page then contain the tenor part for the Alleluia verse, while the tenor of the first verse of the sequence follows on stave 3. All the following sections of the tenor part are introduced by large penned initials similar to those used previously for the tenor. The latter was underlaid first with verse 1b of the sequence (‘Fortem expediat’), for which the outer voices have the incipit. However, in the tenor the full text of verse 1a (‘Mittit ad virginem’) was later added – the first three words beneath the verse 1b text, the remainder above it. This latter text, of course – sung to the melody that appears here as the tenor to v. 1b – would have begun a performance of this sequence; v. 1b would then have been sung polyphonically as a response to it. Verse 2b of the sequence follows, in the lower part of this recto, but with the voices in a different order: the discantus first, followed by the tenor and (after an empty stave) the bassus – the same layout as the introit and Kyrie. Verses 3b, 4b and 5b of the sequence follow on the next opening (ff. 167v–168r), but – in contrast to the layout on ff. 166v–167r – with each voice-part kept together: the discantus for all three verses first (f. 167v, staves 1–5), followed by the bassus for all three (f. 167v, staves 6–9), then the tenor for all three at the top of the facing recto (f. 168r, staves 1–3). While the discantus and the bassus start with simple capitals, the tenor is introduced with large penned initials at the beginning of each line.
Like the introit, gradual and sequence, the brief Sanctus which follows also involves a page turn: the ‘Sanctus’ and ‘Pleni’ sections, both introduced by large penned initials, appear on the lower staves of f. 168r
'Sanctus' and 'Pleni' sections
Close , the ‘Benedictus’ (with the trope ‘Marie filius’) and the chant for the final ‘Osanna’ at the top of the following verso
'Benedictus' and the chant for the following 'Osanna'
Close . While the ‘Benedictus’ received a large penned initial, the ‘Osanna’ starts simply with a capital, although its first ‘Sanctus’ is introduced by a large penned more elaborate initial. Apart from this final ‘Osanna’, only alternate sections of this Sanctus are provided in the manuscript: the first and third ‘Sanctus’ appellations, the ‘Dominus deus sabaoth’ and the first and second ‘Osanna’s are to be sung to plainchant, of which only the last, the final ‘Osanna’, is actually supplied in the manuscript. (The other chant sections would need to be supplied from memory, aided where necessary by consultation of an appropriate written chant source.) The second ‘Sanctus’ and the ‘Pleni’ are presented together on f. 168r, the discantus part first, followed by the tenor and the bassus, on staves 4, 5 and 7 respectively, leaving staves 6, 8 and 9 empty. The ‘Benedictus’ is similarly presented on staves 1–5 of f. 168v, the final chant ‘Osanna’ appearing after the bassus.
Example 5: No. 195. Petrus Wilhelmi, Panis ewus / Pange exul / Panis ecce / Patribus / Tantum ergo, ff. 255v–257r
The acrostic ‘Petrus’ in the first two voice-parts of the five-part motet no. 195, Panis ewus / Pange exul / Panis ecce / Patribus / Tantum ergo, on ff. 255v–257r, copied by scribe A in full notation, reveals it to be another work of Petrus Wilhelmi. This piece demonstrates a number of features that may be associated with the repertory of Bohemian cantiones and motets copied in full notation in Speciálník. The five voice-parts, which in the first section have different texts, are identified in the outer margins of the first opening by the relative pitches of their first notes in relation to that of the part that begins on the lowest note: 8a, 5a, 8a, 5a, 1a
8a, 5a, 8a, 5a, 1a
Close [[ia119]. The first section (prima pars) of the motet appears on this opening, and concludes with a recurrence of the first few notes of each voice at the end, leading back to the beginning: it is thus a perpetual canon. This perpetual canon is followed (if that is not a contradiction in terms) by a shorter second section (secunda pars), copied on ff. 256v–257r, in which the five parts are identified in the margins not by their original relative pitches, but (presumably because their relative pitches are different at the beginning of this new section) by the incipits of the texts they sang in the prima pars. The motet begins with an empty space, and the first word, ‘[P]anis’, is left incomplete while voice-parts 2–5 have large bold initials of type 2 on the first opening. Although it would be logical to assume that the intention was to insert a more elaborate initial at the beginning of the first part, the reserved space is no bigger than the others, and this type of initial does not occur in combination with more elaborate decoration anywhere else in the manuscript. As the first space is quite often left empty when the remaining parts received large initials of this type it could be that the decoration was filled in economically, being executed only where needed for clarity. On the second opening, all five parts have smaller initials drawn in simple lines. On the first opening, each part begins immediately after the previous part ends, without necessarily beginning a new stave (though the beginning of part 4 does happen to coincide with a stave change), and part 3 runs on from the bottom of the verso to the top of the recto. Each voice-part of the secunda pars, however, occupies almost exactly two staves, so each begins a new stave. The first four parts are located on the verso, the fifth at the top of the recto.
All five parts are fully texted throughout. The texts of voice-parts 1 and 2, with the ‘Petrus’ acrostic, as well as that of part 4, are non-liturgical, and are likely to have been specially written for this composition. That of part 3 is an elaboration of the sequence verse (for Corpus Christi) Ecce panis angelorum, while that of part 5, Tantum ergo, is the penultimate verse of the hymn Pange lingua (also for Corpus Christi), with an additional pair of lines (‘Fides sitque firmamentum / Ergo intellectui’). The underlay of this section is almost literally one note per syllable, and when occasionally two short notes are sung to the same syllable, they are placed close together on the stave to make this clear (see, for example, the pairs of notes to be sung to the first two syllables of ‘lugubraminibus’, at the end of the first stave of f. 255v); longer notes to be sung to the same syllable are usually ligated (see, for example, the last syllable of ‘cibus’ on stave 2 of f. 255v). The text of the second section of the work, which (although there is no mensuration or proportion sign in the source to indicate this) is in triple prolation, and in which all voice-parts sing the same words, is based on the first verse of another Corpus Christi hymn by Aquinas, O salutaris hostia (actually the penultimate verse of Verbum supernum prodiens):
Hymn O salutaris hostia
O salutaris hostia
O salutaris hostia
Quae caeli pandis ostium:
Que celi pandis hostia
Bella premunt hostilia,
Da robur, fer auxilium.
Da robur fer auxilium
Ne bella nos hostilia
Melismas on the penultimate syllable of the third line of the text and the antepenultimate syllable of the last line are clearly and carefully indicated in the manuscript.
This motet is followed on the lower staves of f. 257r by a brief work (no. 513, Christus eternalia mundo)presented (also in full notation) in quite a different way, with a fully texted discantus and untexted tenor and contratenor. It starts with a large bold initial similar to those on the previous opening. The beginning of the tenor follows directly on from the end of the discantus, simply separated by a vertical line, rather than beginning a new stave, though the contratenor does begin a new stave. Seven additional stanzas of text are written over the last two staves and the bottom margin of the page.