The Journal of Seventeenth-Century Music

Volume 11 (2005) No. 1

Editing the Keyboard Music of Bull and Scheidemann

David Schulenberg*

John Bull: Keyboard Music I. Edited by John Steele and Francis Cameron, with additional material by Thurston Dart. Third edition, revised by Alan Brown. Musica Britannica 14. London: Stainer & Bell, 2001. [xxxiv, 181 pp. ISMN M-2202-2024-1; ISBN 0-85249-867-5; ISSN 0580-2954. £67.]

Heinrich Scheidemann: Sämtliche Werke für Clavier (Cembalo) / Complete Harpsichord Music. Edited by Pieter Dirksen. Edition Breitkopf 8688. Wiesbaden: Breitkopf & Härtel, 2000. [84 pp. ISMN M-004-18084-6 €21.]

1. Introduction

2. Problems of Attribution

3. John Bull

4. The New Bull Edition

5. Attributions: The Fantasias

6. Further Questions of Attribution

7. Conclusions

8. Scheidemann Edition—Introduction

9. Instrument and Repertory

10. Attribution Again

11. More Conclusions

References

Appendix

1. Introduction

1.1 What is a composer? When confronted by a volume of compositions whose title declares them the complete music of a given composer for a given medium, we tend to assume that we have before us a body of work that represents or expresses some individual’s creative essence and personality, or at least a significant portion thereof. Depending on how much we know of that person’s biography and character, we may also be tempted to relate the works to events in that person’s life—both external events and internal ones, even moral or spiritual crises. The tradition of monumental collected works that began in the nineteenth century encouraged this approach, inspired as it was by such well-known figures as Beethoven and J. S. Bach. But the idea of assembling and publishing the collected works of admired composers goes back at least to the early seventeenth century, and its models in Western Europe can be traced back to the manuscript editions of poetry and drama of classical antiquity.

1.2 Yet what if the composer whose works are so assembled never existed? Homer is the archetype of such a figure—a poet who, despite the existence of ancient biographies and a large corpus of attributions, probably never lived. Less clear-cut is the case of the artist of whose historical existence we have no doubt, yet whose life and works are shrouded in legend or uncertainty. Among these are figures such as Boccherini and Pergolesi, who are best known today by works they never wrote. Another situation is represented by Machaut, and indeed, most pre- and early modern composers, whose biographies are too incompletely understood for us confidently to connect events in their lives with more than a handful of individual musical compositions.

2. Problems of Attribution

2.1 The repertories gathered together in the present editions are of the latter sort. In the case of Bull we have a composer whose biography is marked by one particularly momentous event—his flight from England and consequent shift from Protestant to Roman Catholic patronage—which has been used to explain some of the anomalies of style and transmission in the repertory under review. Yet there are serious questions as to how much of that repertory is actually his, let alone how it relates to his Continental exodus. With Scheidemann, attribution issues are also prominent, but because of the nature of this particular repertory, we also are faced with the question of why and for whom a distinguished professional organist-composer wrote a small body of music, allegedly for the harpsichord, that appears to be largely for amateurs and is sometimes quite amateurish in style and technique. Indeed, in both cases, the greatly varying style and quality of the music lead one to wonder whether we can draw any conclusions about these men or their compositional styles and concerns from these volumes. In what sense is each man responsible for what is here assigned to him? For each piece, did he compose the music or merely arrange it, or perhaps engage in what we might term creative copying? Quite apart from the issue of attribution is that of presentation: is the format of the nineteenth-century monumental edition the most apt for presenting texts that were far less stable than those of later traditions, using notation that departed significantly from nineteenth-century engraving conventions?

3. John Bull

3.1 Only specialists are likely to have given much thought to any relationships that might exist between the music of John Bull (ca. 1562–1628) and that of Heinrich (or Hendrik) Scheidemann (ca. 1595–1663).¹ At least in English-speaking countries, Bull is usually thought of as one of the Elizabethan virginalists. But he spent the last fifteen years of his life as an organist in the Spanish Netherlands, and, in addition to a body of pavans, galliards, and variations on popular English tunes, his name is attached to a diverse repertory of preludes and fantasias, plainsong settings and verses, and a few variations on Dutch carols. Pieces from both groups are preserved in Continental as well as English sources; music attributed to Bull circulated in Sweden, Germany, and Austria well into the latter half of the seventeenth century, alongside pieces by Frescobaldi, Sweelinck, and others. Among those others are Sweelinck’s students, of whom Scheidemann was perhaps the most original, if not the best known. He was Reincken’s teacher and predecessor as organist at St. Katherine’s church in Hamburg, and although neither he nor Sweelinck is likely to have known Bull personally, he and his contemporaries surely would have regarded his music as part of the same tradition, notwithstanding the present-day tendency to view Bull as a composer of the late Renaissance and Scheidemann as Baroque.²

3.2 As one of the best-known and most prolific of the virginalists, Bull was among the first composers honored by a collected edition of his keyboard music in Musica Britannica: A National Collection of Music (henceforth MB). The subtitle of the collection marks its connection with the nineteenth-century tradition of national monuments. A further link with that tradition was the division of the music between two volumes, one devoted to what were thought to be primarily organ works, the other to works for stringed keyboard instruments. This distinction was not made dogmatically, and it reflects a division in terms of genre and, to a lesser degree, of transmission and perhaps also of date and place of origin. But it is worth bearing in mind that although this organization makes sense within the tradition of the monumental Gesamtausgabe—and may also seem eminently practical to many present-day users of these volumes—it may not be the best way of conveying the historical reality of this repertory.

3.3 The present volume contains Bull’s fantasias, plainsong settings, and carol settings. It also includes several preludes that the original editors paired with larger pieces included in the volume. This pairing already hints at one of the problems inherent in the organization adopted here: no source actually has a title such as Prelude and Fantasia, attached to the opening work in the present volume (BuK 1).³ The textual commentary expresses varying degrees of doubt about each of the pairings in the volume, none of which can be traced to the composer.⁴ Other preludes were placed in volume 2 (MB 19) because they were unattached, or because they were thought more suitable to stringed keyboard instruments. In addition, the volume closes with five “Dorick” preludes (or short fantasias), the Greek term referring not to their mode, but to a category of serious but plain music—perhaps by analogy to the Doric order in architecture—to which they were thought, on sometimes doubtful grounds, to belong.⁵

3.4 The view of much of this repertory as organ music, although sometimes problematic, is helpful in understanding some of the pieces, which would probably make their points most clearly on a colorfully registered Dutch or Flemish-style organ. A few suggest division between solo and accompaniment manuals, and certain frequently occurring textures—for instance, extended bicinia comprised of one lively part against a sustained cantus firmus or ostinato—point to the organ as the optimum medium, since a harpsichord, clavichord, or the like could not sustain the long notes. The dissemination of these pieces in manuscripts containing primarily serious contrapuntal works, not simple dances, indicates their reception among professional organists and perhaps serious amateurs of the seventeenth century. Stylistically, the fantasias have more in common with Iberian and Netherlandish pieces of the period than those of other English virginalists, such as Byrd. The plainsong settings, although problematic from the point of view of modern taste—they have struck many students of this music as mechanistic—grow more clearly out of an English tradition. Yet these too found a resonance in northern Europe, where Samuel Scheidt, among others seems to have been influenced by them.

4. The New Bull Edition

4.1 Because this volume is the revision of an earlier one, a detailed consideration of its relationship to its predecessors is in order. Unlike the second edition of 1967, which involved only the addition of a small amount of supplementary material,⁶ the present third edition represents a comprehensive revision of the textual matter as well as a new checking of the musical text, which has undergone numerous small changes as well. The result, however, is an uncomfortable cross between a true new edition and a corrected reprint, as becomes all too obvious not only from the retention of some questionable features of the original but even from references within the revised text to the compromises that have had to be made.⁷ Thus, despite ample and conscientious additions and corrections by Alan Brown, this remains very much the edition that John Steele and Francis Cameron first published in 1960, “with introductory material by Thurston Dart,” as the title page declared, perhaps not sufficiently indicating his mark on the volume as a whole.

4.2 In fact, even the “Introduction to the First (1960) Edition,” with which the volume opens, has undergone “minor corrections and adjustments” to reflect recent scholarship.⁸ Yet “the opinions expressed by the original editors have not been changed,” and thus some dubious assumptions are allowed to stand.⁹ Dart’s “Calendar of the Life of John Bull” has been removed on the grounds that reliable biographical material is now available elsewhere. But Bull’s biography is summarized in a new “Note for the Third Edition,” which also provides a summary of the difficult problems of attribution that affect the works printed here. A separate discussion entitled “Editorial Method” has been rewritten, in part to apologize for those aspects of the original which have had to be retained. One feature, the conflationary approach to establishing the musical text, was not necessarily a bad thing, given the nature of some of the sources. But the halving of note values in a little more than half the works in the volume was most unfortunate. To be sure, where values have been halved this is clearly marked at the beginning of each score.¹⁰ And in a few pieces in an archaic contrapuntal style (such as the Fantasia, BuK 15) notation in reduced values is more legible and shows voice leading more clearly than the original. But in most cases the result is misleading at best and in some instances monstrous and almost unreadable, as in the fifth In Nomine, BuK 24, where the eye is faced by nineteen measures of unbroken thirty-seconds in counterpoint with steady sixteenths.

4.3 Barring throughout the volume is regularized, which forces some reduced-value pieces to open with a half-measure, although these probably cannot be considered upbeats in the modern sense. One might think that all pieces in reduced values would appear with a 4/4 time signature, the others in 2/2. But only the little Fantasia, BuK 16, (actually a prelude) is in 2/2. The In Nomine (VIII), BuK 27, which is reduced, appears in 2/4 although “the source is barred in 4/2 (original values).” In the Fantasia, BuK 10, opening also in 2/4, mm. 83–106 appear in 3/4 although “no source suggests the barring adopted here” (p. 168). No rationalization for these choices is offered. In the case of BuK 10, the shift to triple-time notation is implied by the music, whose bass shifts at m. 83 to one of those bizarre little ostinatos that appeal to the connoisseur of Tudor keyboard music (A–B-flat–A–G-sharp–A). The rhythm of this ostinato is effectively in 3/4, but we are not told what the original barring was. Moreover, the new barring obscures cross-rhythms created by the free (non-ostinato) voice, whose shifts between different types of rhythm—the prevailing note values vary between triplets, sixteenths, and thirty-seconds—now take place on off-beats.

4.4 The conviction that the editor must interpret tempo and meter for the unwary reader was only the most obvious manifestation of a prescriptive editorial attitude that today seems clearly inappropriate in a critical edition. Another was Dart’s advice concerning the interpretation of ornament signs, which has been unduly influential. Brown has somewhat softened Dart’s language (p. xxvi), but his discussion will still fail to convey the complexity of the issue to non-specialists, and he neglects to mention the considerable (if inconclusive) published scholarship on the subject.¹¹

4.5 Yet another section by Dart, entitled “The Instruments,” has also been revised, eliminating information about harpsichords and virginals as well as one organ specification. Yet the discussion retains the misleading account of the famous chromatic hexachord Fantasia, BuK 17, as requiring “equal temperament tuning [sic], or something very close to it” (p. xxix). One enharmonic equivalent within the piece would, if taken literally, rule out performance in meantone or just intonation.¹² But performance is possible in any of the circulating temperaments that began to be described late in the seventeenth century. Some such temperament must already have been in use by the end of Bull’s life, if not on the harpsichord—assumed, perhaps wrongly, to be the instrument used in BuK 17—then on chamber organs, required for the keyboard parts provided for certain chromatic fantasies for viol consort.¹³

4.6 Toward the back of the book, the “List of Sources and their Abbreviations” has been greatly expanded by the inclusion of up-to-date information and references (only one unimportant source has been added). Some doubtful copyist identifications have been deleted, notably that of Francis Tregian as compiler of the Fitzwilliam Virginal Book.¹⁴ Yet the familiar source sigla, based on supposed owners or copyists, remain in use (these are listed in Table 1 in the Appendix). The Fitzwilliam Virginal Book is still “Tr,” and another important source whose principal hand Margaret Glyn had regarded as Bull’s is still “Bu,” even though hers was a view that “those preparing this volume cannot yet sustain.”¹⁵ Moreover, two very uncertain scribal identifications are retained. The manuscript here referred to as Me (GB-Lbl Additional MS 23623) is described as “almost certainly by Guilelmus Messaus” (p. 164), which at least replaces the unconditioned certainty expressed in the first edition. Wr (GB-Lbl Additional MS 30485) is now said to have been “compiled very probably by Thomas Weelkes” (p. 165). The latter assertion is insignificant, since this source contains only the Canon, BuK 51, with a plausible attribution to Tallis. But Me is the most problematical of all the Bull sources, and in accepting Dart’s hypothesis as to its copyist Brown disregards the cautions sounded by Walker Cunningham and Alexander Silbiger.¹⁶ The most recent discussion of this manuscript, by Rudolf Rasch, affirms Messaus’ status as chief copyist.¹⁷ But this depends entirely on the assumption that the existing manuscript is in the same hand as a long-lost companion volume; an inventory of the latter indicated that it contained overlapping repertory and was signed by Messaus.¹⁸ Underlying the argument for Messaus’s involvement is a hypothesis familiar to students of the English virginalists: the manuscript was compiled by an associate of the composer, either with the latter’s cooperation or from the latter’s papers. As attractive as it is, the hypothesis has yet to be demonstrated for any of the manuscripts to which it has been applied.¹⁹

4.7 The Textual Commentary has also been considerably expanded, and it is now preceded by a page headed “Notes on the Textual Commentary.” In addition to revised lists of variant readings, the entry for each piece has sometimes been drastically expanded to include discussion of attribution and, where appropriate, the sources of ornaments, fingerings, and plainsong cantus firmi. As in other keyboard volumes of the series, however, there is still no comprehensive listing of readings for ornaments and fingerings, so it remains difficult to evaluate every individual instance of each. For instance, in the so-called “Great” In Nomine, BuK 28, we are told that the two sources known as Tr and Bu “agree on many ornaments, though Bu has slightly more. All ornaments in both sources are given” (p. 173). This is ambiguous; in fact it seems to mean that any ornament appearing in either of these two sources is given, thereby obscuring any pattern that might have existed in the transmission of these signs, any distinction in authority between the two sources, or any distinction in the meaning that these signs might have had for the two different copyists involved.²⁰

4.8 The same somewhat casual approach extends to textual relationships between sources. For the popular Miserere I, BuK 34, which survives in six manuscripts—a greater number than for any other work in this volume—we are told that “the nature of the variants suggests that Bu represents an earlier version of the piece” (p. 175). But one must comb through a daunting series of readings (run together in block paragraph format) to find the substantiation for this. The stemmatics of the five other sources are not explained at all. We are told that the chromatic hexachord Fantasia, BuK 17, is preserved in some sources with an attribution to Poglietti—but one must consult Cunningham to find what these sources are. In fact, they represent an Austrian manuscript tradition that includes Poglietti’s treatise Compendium oder kurtzer Begriff.²¹ According to a tradition transmitted by Mattheson, Froberger also composed a hexachord fantasia in which the ostinato appears on successive chromatic steps, but no such piece survives by Froberger. Might the reference have been actually to Bull’s piece, transmitted anonymously in Vienna and assumed to be by a more familiar composer?²²

4.9 The actual musical contents have not been changed save for the addition of one page containing a variant version of the Prelude, BuK 2/1, as well as four pages presenting the plainsong originals that are elaborated in BuK 20–31 and 33–49. The latter have been edited from sources that Bull might have known, although his settings of the In Nomine, like others, probably derive not directly from the plainsong but from its use in the Sanctus of Taverner’s Missa Gloria tibi trinitatis. Given the large number of such settings in the volume (twelve), it would have been useful to have a complete copy of Taverner’s setting as well, especially as this was the source not only of the cantus firmus but of texture and material in many subsequent instrumental settings. Indeed, at the opening of the first In Nomine in the volume, BuK 20, the lively soprano part clearly echoes Taverner’s treble line over a slower-moving cantus firmus (bass in Bull’s setting, alto in Taverner’s). Although the editors never explain their ordering principles, this particular setting bears the subheading “Prima pars” in one source, where it is the first of six given in close proximity.²³

4.10 Changes to the musical text are limited in the main to the occasional addition of accidentals, ornament signs, and the like. In all cases, the new readings are arguably better, and, although some might be judged arbitrary, others reflect currently preferred editorial practice. For instance, the eclectic reading in the Fantasia, BuK 10, for m. 105 is replaced by a reading from a single source, Bu, and a passage near the end of In Nomine III, BuK 22, now follows the second of the two copies in that source. The latter changes reflect a re-evaluation of the sources, Bu now being promoted ahead of Tr as the primary source of BuK 10, even though the copyist of neither can be identified. On the other hand, ornament signs have been added to the edition of the Fantasia, BuK 11, even though some are admitted to be later additions to the sole source and are “unlikely to derive from Bull” (p. 168).

4.11 Spot-checking of the revised text reveals occasional inaccuracies. A comparison of the edition of the carol setting “Den lustelijeken Meij,” BuK 52, with the facsimile of its source reveals one wrong note as well as some unreported editorial emendations. An unlikely editorial conjecture remains as well.²⁴ Cunningham proposed numerous corrections for the second edition, most of which have been made good by altering either the music or the textual commentary . In BuK 1/2, however, two of Cunningham’s readings would improve the text,²⁵ and in BuK 4 the eighth measure still contains an extra note.²⁶

5. Attributions: The Fantasias

5.1 Questions of attribution are widespread throughout the virginalist repertory. Not included in this volume are two plainsong settings that have been described as possibly Bull’s in previously published volumes of MB. John Caldwell proposed that an “Aurora lucis rutilat” preserved in Bu amongst other anonymous works known to be by Bull “may in fact be by Bull, but if so it is probably a very early work.”²⁷ This is plausible, considering that most of the works in Bu known to be Bull’s are also anonymous there. In addition, the composer Thomas Tomkins copied a Kyrie eleison setting alongside some In Nomines attributed to Bull. This was published as a doubtful work of Orlando Gibbons, to whom three other copies assign it, although the editor, Gerald Hendrie, regarded its style as “closer to that of Bull than of Gibbons.”²⁸

5.2 The problematical features of these two pieces—anonymous transmission and conflicting attribution—extend to a significant portion of the works in the present volume. Most are unica—40 of 66 separate pieces (counting preludes separately, but not individual verses of a plainsong setting). Eleven lack attributions, and two have conflicting attributions. The issue is most acute in the sixteen works that can be considered fantasias, since in musical terms these are the most varied and ambitious group of pieces; they are also the ones with the most problematic transmission.²⁹ Most of these are in what we would call Baroque, not Renaissance, Tudor, or Elizabethan style; they are akin in many ways to music of the Sweelinck school or even the tientos of the Iberian tradition, thus raising questions as to how Bull came to adopt such a style (really several quite distinct styles) at a relatively late stage of his career.

5.3 Unfortunately, what Cunningham wrote of the second edition remains entirely applicable to the new one: on questions of authorship or style in these pieces it “offers little help … for the sequence there follows neither a hypothetical chronology, preference for certain sources over others, nor the order of the keyboard modes.”³⁰ Assuming for a moment that all of these are actually by Bull, we gain the impression of a composer experimenting in radically different styles and rarely achieving satisfactory results; the most frequent problem is a lack of musical direction or shape beyond the immediate phrase. Paradoxically, the failure to achieve success or even to convey the sense of a single musical personality might actually be consistent with the arbitrariness and impulsiveness that seem common to Bull’s more solidly attributed pieces.

5.4 The problems of attribution in this repertory seem insoluble without the discovery of new sources, and one might question the wisdom of pursuing it, given that the pieces on the whole are a disappointment, especially by comparison to those in Volume 2. But so long as we care about Bull himself and his possible influence on Continental musicians, the issue is worth considering further, especially in the eighteen or nineteen fantasias, individually the largest of the pieces in the volume and as a group the most varied and musically interesting; moreover it is here that attribution issues are most acute.

5.5 We might characterize the styles of the fantasias in the following manner, at the same time giving them an ideal relative chronology (unica in Me are shown in bold face):³¹

Archaic Tudor-style fantasias: 10, 11, 15

Less archaic pieces still with identifiably Tudor style: 12

Idem, on ostinato subjects: 17, 18

Chromatic pieces with elements of both Tudor and Dutch style: 4, 5

Spanish or Dutch organ fantasias: 1/1?, 1/2, 2/2, 13

Idem, on ostinato subjects: 14, 32

Pieces perhaps modeled on Italian sonata style: 6

Adaptations of Italian pieces in more or less Tudor style: 3, 8, 9

Idem, in a perhaps Dutch or German style: 7

A few pieces might be assigned to different categories; for instance, “God Save the King,” BuK 32, is reminiscent at times of old-fashioned Tudor plainsong settings. The hexachord fantasias, BuK 17–18, both listed above as “identifiably Tudor,” are very different types of pieces, the first being chromatic and devoid of passagework, the second a tour de force of diatonic figuration. The single “less archaic” fantasia in more or less Tudor style, BuK 12, seems the sort of thing Bull might have produced in imitation of one of Byrd’s large multi-sectional keyboard fantasias. The best of the fantasias (or at least those most likely to be effective in performance today) are the shorter ones—the two fantasias after Palestrina’s madrigal “Vestiva i colli,” BuK 8–9, and the two chromatic ones BuK 4–5. Also effective in performance are the “Tudor” BuK 12—the closest thing in the book to Byrd’s great works—and the “Baroque” BuK 6, which although not particularly impressive in its craftsmanship has a great deal of energy and some of the varietas of early sonata writing.

5.6 Could one composer have been responsible for all of these? It is striking that, with three exceptions, the “problem” pieces—that is, all except those in the first three categories—are unica from Me.³² Because Me has been taken to be the work of a Flemish associate of Bull’s, preserving late works composed after his confrontation with Sweelinck and other regional composers, it is easy to understand why Brown and others have concluded that Bull’s style underwent a dramatic shift after his arrival on the Continent. Yet even Brown has flagged seven of the fantasias as problematic, tentatively accepting three of these (BuK 1, 12, and 15) but regarding two as probably by other composers (BuK 13 and 14); another two (BuK 6 and 7) are “best regarded as anonymous” (p. xxiii). It is impossible to discuss the issue comprehensively here; the Appendix sketches the principal issues and provides a tentative list of works in the present volume that bear “safe” attributions. Most pressing are the issues involving several of the longer fantasias in “Continental” style. All are serious pieces comprised of several long-winded sections, each on a single main subject, that tend to open imitatively and then gradually become more animated through the restrained application of melodic embellishment.

5.7 The Prelude and Fantasia, BuK 1, an unicum from Vi, consists, in fact, of a short fragment of a fantasia that was broken off by the copyist, followed by a second complete fantasia. Both are in what we would call D minor (without key signature). Brown suggests this may be an example of “Bull’s later style,” referring to “certain textural and motivic features” shared with the Salve Regina verse sets nos. 40 and 41 (p. xxiii). Brown’s “textual and motivic features” might be a reference to the occasional repeated-note figure (compare 1/2:45 and 40/2:11 or 40/5:11; something similar also occurs in 5:16–17), or to a type of phrase ending common in Jacobean keyboard music (as in 1/2:31 and 1/2:76; cf. 40/2:4, 41/2:14), or perhaps simply to the sometimes very sudden emergence of florid solo passagework out of a serene contrapuntal imitation in three parts. These things may not be sufficient to define a “later style,” but they are the types of fingerprints that suggest composition within a common “school” or group of associates, if not by one person.

5.8 BuK 13 is a “Fantasia in the Sixth Mode.” The title, which might more accurately be given as “Fantasy of the Sixth Tone,” is taken from Me, the sole source, whose macaronic heading is one of the signs that this is indeed a Dutch or Flemish manuscript (“Fantasia sixti toni A.4. Du Jan Bull: Doct^r”). But already in 1962, when Dart edited the second volume of Bull’s works, BuK 13 was listed in an appendix as one of eleven pieces “likely not Bull’s.” In subsequent editions of the first volume, this work has borne the heading “Cornet?,” one of six instances in which a possible alternative composer is named.³³ The reference is to Pieter Cornet (ca. 1570/80–1633), for a time a colleague of Bull’s as organist to the Brussels vice-regal court. The problem is that if Bull’s, the piece is unlike anything else known to be by him; yet if Cornet’s, it seems more successful than the six fantasias that have been preserved and published under his name. Stylistic affinities to the latter go beyond the use of a subject similar to that used in two other works—and in one by Peter Philips (also a Brussels court organist). Yet, as Cunningham observes, “transitions are smoother” than in Cornet’s firm attributions, “contrasts are handled with an ear for texture and register … phrases are thought out in shapes, rather than lengths.”³⁴ In addition, the piece occasionally builds to some real power, especially its one chromatic passage (mm. 197–8), sensibly placed as the climax to the third and last of the piece’s long-spanning sections. This alone is enough to discount Bull as the composer, given his tendency to ramble inconsequentially, but whether it makes the piece Cornet’s is one of those unanswerable questions of attribution.

5.9 The “Fantasia in the Eighth Mode” that follows in the volume (BuK 14) is another unicum from Me, where it appears twice in apparently independent copies. Dart’s suggestion that it might be Sweelinck’s was rejected by Gustav Leonhardt, editor of the latter’s fantasias, on the grounds that the piece bears “Sweelinck’s mark but not his personality.”³⁵ Thus the piece has been rejected by the editors of the two composers to whom it has been attributed, yet the writers of monographs on the two composers have each reclaimed it. Cunningham likened its treatment of the ostinato subject to that in “God Save the King,” BuK 32—another contested work—while Pieter Dirksen has found it “attributable to Sweelinck.”³⁶ Dirksen amply demonstrates that BuK 14 employs melodic formulas also found in Sweelinck’s music. He might have added that the overall organization, although not exactly comparable to that of Sweelinck’s mature fantasias, nevertheless follows a rational design. This design is no more characteristic of the few fantasies firmly attributed to Bull than is that of BuK 13:

mm.	Style
1–63a	bicinium: free figuration against cantus firmus
63b–112	tricinium: imitative figuration against cantus firmus
113–158	four-part imitative counterpoint, based on 1/2 diminution of subject37
159–180	coda (four parts), based on 1/4 diminution of subject

Several of Bull’s more extended plainsong settings are constructed on the same principle of gradually expanding the number of contrapuntal parts. Yet the actual figuration is frequently reminiscent of the Sweelinck style, and Dirksen is right to point out that the transitions between sections are handled with care more characteristic of Sweelinck: the climactic figuration in measures 60–3 and 110–12 shows “a mind at work which is much more conscious of the structural importance” of each juncture.³⁸ But none of Sweelinck’s works has precisely the organization shown above, and therefore his authorship must be considered at least as doubtful as that of Bull, who after all is the composer named in the source. This source shows no other connection with Sweelinck except for its attribution to the latter of the subject used in the fantasia, BuK 5.

5.10 The ostinato fantasia “God Save the King,” BuK 32, appears out of place in the volume among the plainsong settings, apparently reflecting Dart’s hypothesis that the royal acclamation was customarily sung to the four-note ostinato c''–g''–f''–e''. But this seems unlikely unless it was the second word rather than the first or last that was then emphasized.³⁹ That Brown has a sense of humor is clear from his expansion of the textual commentary to quote a nineteenth-century observation: Bull’s ostinato is no more like “the anthem now sung than a frog is to an ox.”⁴⁰ The book in which those words appeared is now the sole source for the present version. Another version, attributed to Sweelinck, occurs in the Lübbenau manuscript known as Lynar A1 (here Ly).⁴¹ In the latter, the final twenty-five measures (from the last beat of m. 110) are replaced by seventeen entirely different measures. Actually, the version given in the present volume is a conflation, for measures 24–39 are inserted from Ly; the fact is noted in the textual commentary, but no justification is given. If the ending preserved only in Ly is indeed by Sweelinck, it would be reasonable to conclude that the four variations inserted into the present text are his as well. And indeed, one of these—measures 26–29, with its repeated figure corte—is among the passages cited by Dirksen as superficially resembling Sweelinck’s style, although he asserts that the “schematic use” of the motive here is not actually characteristic of Sweelinck.⁴²

5.11 Hence, Dirksen agrees with the conjectures of Dart and Curtis that the ending in Ly is Sweelinck’s, and that the substitute ending corrects a problem in the original, which ends in the wrong key.⁴³ Far from a problem, this ending is actually a rather clever reinterpretation of the ostinato’s last three notes, and it is hinted at much earlier in the piece several times (e.g., in a very English-sounding cadence to A in mm. 80–1). Nevertheless, it may well be that Sweelinck or some member of his school did find the modal mixture disturbing. But it is not unusual in plainsong settings, whose tonal shape was, of course, determined by the vagaries of the cantus firmus. One might compare the setting of “Christe redemptor omnium,” BuK 33, which likewise might be seen as starting in C and ending in A. Perhaps this is even a point in favor of Bull’s having written BuK 32, or at least the portions not unique to Ly.⁴⁴

5.12 One other fantasia in this volume merits a word here, although it is certainly not an original work by Bull. The last of the three hexachord fantasias, BuK 19, is, as Cunningham recognized, a keyboard reduction of the “Trentehuictiesme Fantasie” from the Fantasies a iii. iiii. v. et vi. parties of Eustache du Caurroy. This was published by Ballard (in Paris) in 1610, hence shortly before Bull’s Continental exodus. Conceivably, then, this is music that was in vogue at the time of his arrival in Brussels, and perhaps he or some other keyboard player needed a score in order to study it or to accompany a consort playing it. The “Bull” version is a very literal transcription, but because the five parts cross frequently it is difficult to follow the original voice leading. The music, therefore, seems shapeless, at least in its keyboard form, making all the more puzzling the re-ordering of its measures, apparently first observed by Brown (p. 171). I cannot agree that the keyboard version reflects a deliberate recasting of the work, for the swapping of passages occurred not at the divisions between sections, but seemingly randomly, interrupting statements of the subject and at one point producing parallel octaves.⁴⁵ Why this was done, or what it has to do with Bull, must be left to speculation; the sole source is Me, where it bears a date several months after Bull’s death.⁴⁶

5.13 It seems hardly an accident, however, that the copy is immediately followed in the source by three anonymous ricercars in a similar style. The second of these is dated a week later (Oct. 19, 1628). The subsequent plainsong settings in Me are notationally very similar, and one of these bears the latest date entered into the surviving portion of the manuscript (Nov. 16, 1628, for verse 2 of “Telluris ingens conditor,” BuK 47). In other words, the viol fantasia is part of a group of austere exercises in counterpoint. Might Bull have copied out the Caurroy fantasia for study, then composed the works that follow in Me, in a deliberate effort to master a style appropriate for his new situation? Militating against this is the rudimentary nature of some of the plainsong verses. The last one that remains in the manuscript is the canonic “Alleluia,” BuK 49, which contains blatant parallel fifths (m. 12). Another is by Tallis, despite the attribution to Bull; this must be one reason why Dart had questioned Bull’s authorship of all but one of these pieces.⁴⁷

6. Further Questions of Attribution

6.1 Silbiger pointed to a total of fourteen pieces in Me that were omitted without comment from MB although they either bore attributions to Bull or were left anonymous.⁴⁸ Among them are the three ricercars just mentioned, which are anonymous and have no evident connection with Bull. Even less likely to be his are two decorated verses of “Alma redemptoris,” which are in a section of the manuscript evidently added later in the century. There seems no reason to attribute these anonymous pieces to Bull. On the other hand, a short monothematic “Fantazia du Jan Bull”—on a subject perhaps taken from a French chanson—seems at least as likely to be his as BuK 3. The latter is the piece closest to it in style and inspiration, despite being based on an Italian canzona francese. Cunningham’s argument against accepting the attribution to Bull is not compelling; some of what appears to be the piece’s “uneven quality” may be due to errors in the copy (not that unevenness would ever be a sign against Bull’s authorship). More telling is the presence of just two strains in the pavan and galliard, which would be rare although not unheard of in the work of an English composer of the period.⁴⁹ Still, one suspects that these pieces were eliminated from MB on the strength of Dart’s now discredited theory that these works were composed by Jacques Champion, father of Chambonnières.⁵⁰

6.2 Also omitted from MB 19 (not the present volume) are five courantes attributed to Bull later in Me, again all unica. These differ from nine preceding courantes, also attributed to Bull, in that the former have character titles—“Jewel, Battle,” etc.—whereas these have abstract Italian designations such as “Courante prima in A la mi re.”⁵¹ Moreover, these lack the varied reprises of the others, and they seem more Italianate in their occasional use of imitation at the opening and in their less regularly periodic tunes. They are also distinct from both English corantos and Italian correnti in their occasionally quite florid basses and inner parts. They are not like anything else Bull is known or thought to have written, except perhaps the doubtful fantasia on “A Leona,” BuK 7.

6.3 Were Bull to prove to be the composer of all or even some of these additional works, it would further expand the range of styles in which he seems to have composed during his Flemish years. We would need to entertain seriously the somewhat poignant situation of a brilliant, if inconsistent, keyboard virtuoso suddenly thrust into a foreign environment in which his old musical training was no longer relevant, forced by circumstances to learn quickly an entire range of new idioms and genres. Given the uncertainties, it would be wrong to publish these pieces without qualification as works of Bull. But it is equally wrong to continue to exclude them more than forty years after the first edition and more than a decade after Silbiger brought them to the attention of scholars. At the very least, they are more attractive than some of the clumsier, duller plainsong settings included in the volume, which are equally uncertain in authorship.

7. Conclusions

7.1 The remaining pieces in the volume, including the numerous plainsong settings, although of limited appeal today, were clearly an important part of the keyboard culture of several generations both before and after Bull. Given the less than compelling nature of much of this music, one may question the need for devoting further attention to it, especially in an age of diminishing resources for music-historical scholarship. Yet it is essential to rethink the treatment of repertories such as this one, for it will be a long time before the present music can possibly be reissued in a grand, handsomely bound and printed “national” edition. Despite the good intentions of the editor, this volume represents a lost opportunity to integrate recent scholarship on early keyboard music—not only writings immediately relevant to Bull’s music, but those dealing with the nature of keyboard sources, notation, and texts in general during this period.

7.2 Already several decades have passed since Silbiger and others began to point out the need for reconsidering how we approach repertories such as this one, in which texts and attributions are unstable, and the notation, although deceptively similar to that of today, conveys clues to interpretation that are vitiated when rewritten in modern form. Moreover, organizing the pieces in a modern edition according to genre and presumed composer can obscure information present in the seemingly disordered manuscript sources. That Bull’s music is part of a European and not merely an insular keyboard tradition is a point that might have been made by a more imaginatively conceived modern edition, yet the present volume, by maintaining the format and viewpoint of its predecessors, does little to foster such a view.

7.3 How exactly might a new edition of this music reflect such things? One can imagine new formats, made possible by electronic music editing and web-style hypertext, in which it is no longer necessary to define and order repertories by composer and genre, nor is it necessary to present a single Urtext—often as not, a creation of the editor—as the sole main text of the edition. Instead, from a database of readings a program might generate a particular text or group of texts corresponding to, say, a particular source, a particular title, or a particular attribution. Alternate versions of a piece might automatically be displayed simultaneously, for ready comparison, or elements of the text—just ornament signs, or just fingerings—might be shown as they appear in particular sources or from the hands of particular copyists. The result would be to substitute in place of the concrete monuments of a printed Gestamausgabe a virtual representation of the historical reality of pieces that were in some cases recomposed every time they were recopied.

7.4 Obviously the creation of such software would require sophisticated programming, and perhaps also types of electronic display that are as yet unavailable to musicology. Less radically, we might hope for paper editions that acknowledge the great changes that have occurred in how scholars and performers actually use printed music. Conditioned not only by the widespread availability of facsimiles but also by the diversity of musical notation in the twentieth century, musicians who are likely to use a scholarly edition are comfortable with notation that is closer to that of the sources. In particular, a new edition might not only eliminate most of the notation in reduced values, but might also improve the rhythmic notation in polymetric passages that occur particularly in the more archaic plainsong settings: twentieth-century composers such as Elliott Carter discovered useful ways of notating four-against-three and other cross-rhythms that prove more legible than the duplet notation employed here.

7.5 In addition, a new edition might observe more of the original beaming of small note values and division of notes between staves, as one already finds in recent editions of other repertories. For instance, in the carol setting “Den lustelijcken Meij,” BuK 52, a point of imitation is introduced in measure 72 whose smaller note values are consistently grouped in the source (Me) as a single unbeamed eighth followed by six sixteenths beamed together. In the next phrase (m. 79), however, a new motive (a figura corta) is introduced, beamed as either an eighth and two sixteenths together or a single eighth and two sixteenths.⁵² Whether one sees in this an indication for performance practice—perhaps a shift toward more a more heavily articulated style—or merely a way of understanding the motivic content of the music, the change is dramatically visible in the original source. But the present edition, following nineteenth-century convention, regularizes the beaming to produce groups consistently equal in value to a quarter note. Together with the 4/4 time signature, this obliterates important information and gives the novice reader a false impression of the metrical scheme of the music, which in the original notation moves within larger beats (at the half or whole note level) and larger groupings of smaller note values.⁵³

7.6 Although not all pieces or manuscripts are so consistently notated, one particularly persistent aspect of the original notation in this repertory is the isolation of a single note or chord that falls on a downbeat—typically an eighth note—from two or more shorter notes that follow.⁵⁴ Only the latter are beamed together. This strongly implies a break of some sort after the initial note, which accordingly can be understood as concluding the preceding gesture rather than commencing a new one, which begins with the smaller note values afterward. This particular detail of notation cannot be discerned within the conventions of MB. Consequently, the original phrasing of the music is often obscured, and many fingerings and ornament signs make less sense than they might otherwise.⁵⁵

7.7 One would not necessarily follow all original beaming. Sometimes the original notation breaks patterns for no intelligible reasons, and in some passages it is significantly less legible or informative than in others. Some manuscripts, notably including Bu and Wr, entirely dispense with beaming of small note values for some pieces or passages, without any evident reason. Other sources, including Me, sometimes group immense runs of thirty-two or more small values together. Both practices make it difficult to see melodic patterns or to line up the running notes with larger values in the other voices. There would be little point in reproducing this in a modern print edition, although one would want the option to display it in some future electronic format.⁵⁶

8. Scheidemann Edition—Introduction

8.1 Pieter Dirksen’s edition of keyboard music by Scheidemann raises many of the same considerations expressed above. It is what Dirksen has called a “composer (as opposed to source) oriented edition.”⁵⁷ This edition encompasses works deemed to have been intended for a particular type of keyboard instrument, chosen from a larger repertory; and many of the pieces within it raise questions of attribution, although not quite the same ones. In addition, the notation, which has been transcribed from both tablatures and scores, has been modernized in similar ways, although, fortunately, without alteration of note values.

8.2 Hardly a monumental edition, this is a small paperbound volume, evidently designed to provide non-organists with a sampling of music by a composer whose keyboard works are overwhelmingly sacred and frequently include a pedal part. Nevertheless, this is a bona fide critical edition, with a scholarly introduction (in German and in English) and a critical commentary (in German only) listing sources and variant readings. The musical contents are as follows:

1 each of praeambulum, toccata, and fantasia

1 set of variations on a song

1 embellished madrigal intabulation (Felice Anerio’s “Mio cor, se vera sei salamandra”)

29 dances:

    1 pavan (an arrangement of Dowland’s “Lachrimae pavan”)

    1 galliard (with variatio)

    3 mascarades (at least two based on existing melodies; one with variatio)

    2 ballets

    6 allemandes (2 paired with courantes; one with double)

    14 courantes (2 paired with allemandes; 7 with variationes; at least 5 are arrangements or based on existing melodies)

8.3 Stylistically, these pieces range from the post-Sweelinck manner of the toccata and fantasia through a combination of features reminiscent of both Dutch and English virginalist composers in the pavan and galliard, as well as French and Italian elements in the dances. The most substantial items are the first eight, that is, through the “Englische Mascarata in g,” no. 8 (the title seems equivalent to the masques found among the works of Orlando Gibbons and other contemporary English composers). The most impressive and distinctive work is the “Galliarda & Variatio in d,” no. 7, which integrates figuration of the Dutch or Sweelinck style into the large tripartite design, with varied reprises, characteristic of the English virginalists. Although all of the most important pieces have appeared in print previously—some of them several times—harpsichordists will be grateful to have this handy compilation, which provides a broad sampling of the types of keyboard music being played in northern Europe during a period unfamiliar to most.

9. Instrument and Repertory

9.1 Dirksen must have recognized that the title of the volume (“Complete Harpsichord Music”) would raise hackles, for the introduction includes a section headed “Delimitation of the Repertoire.” Dirksen believes that “Scheidemann must be reckoned to [sic] the earliest composers whose keyboard music can be distinctly divided into music for the harpsichord and for the organ” (p. 5). He mentions three considerations in selecting works for this “complete” edition of the harpsichord music: (1) abstract or “cantus-firmus-free” works, such as the toccata, must be preserved in sources “dominated by typical harpsichord repertoire”; (2) anonymous works in general are not included even where they occur alongside pieces with attributed concordances; (3) exceptions are made for two anonymous works whose style and transmission are thought to indicate Scheidemann’s authorship (pp. 6–7).

9.2 Perhaps if one substituted “stringed keyboard instrument,” the references above to the harpsichord would be less problematic. For much of what Dirksen considers typical harpsichord music—he mentions pieces by Froberger, Reincken, and Buxtehude—was surely played on the clavichord in the north-German and Scandinavian regions where the present repertory circulated. Virginals and spinets—only loosely identical with the larger harpsichord—were also used, and chamber organs might have also been available in the wealthiest households. None of the sources specifies an instrument, and Dirksen provides no rationale for assigning this music so specifically to the harpsichord. That is one of the ingrained habits that a modern edition such as this might seek to dispel. To use a category with a better-documented historical basis, what we have here is Scheidemann’s keyboard music for the chamber as opposed to the church.

9.3 Does this volume include all of Scheidemann’s keyboard chamber music? The collected edition of Scheidemann’s Orgelwerke includes fourteen praeambula as well as two toccatas, two canzonas, and two fugues.⁵⁸ Of these, the present volume reprints one praeambulum, one toccata, and one fantasia, since these happen to be preserved in sources that also contain dances or other such chamber pieces. Yet the one prelude is preserved in a Lüneburg manuscript alongside eleven other such pieces, nine of them unica. Some of these contain pedal parts or passages in sustained style that seem clearly idiomatic for organ, but others seem as idiomatic to the harpsichord as the one published here. The same can also be said for a short piece entitled “Fuga.” Another “fuga,” as well as a “Canzon” in F, both found in other Lüneburg manuscripts, are equally idiomatic to the harpsichord.⁵⁹ Indeed, the so-called Düben tablature, from which the present volume reprints Scheidemann’s galliard, prelude, and madrigal intabulation, also includes a canzona by Frescobaldi that was published as a work for cembalo or organ.⁶⁰

9.4 Dirksen considers the “Canzon in F” to contain writing idiomatic to the harpsichord, yet he excludes it from the present edition because it is preserved as an unicum “in the midst of a repertoire clearly intended for the organ” (p. 6). But it seems rash to assume that an accident of preservation tells us on what instrument a particular piece was played. The one prelude included in the present edition presumably was played by members of the Düben family on the same instrument used for the dances and the madrigal intabulation also edited here. But who selected this particular prelude for inclusion in the Dübens’ “chamber” manuscript? And what would have prevented those who owned copies of Scheidemann’s canzonas or fugues from playing them on stringed keyboard instruments, even if these happened to be preserved in manuscripts that lacked dances? Harpsichordists who purchase the present volume may well be disappointed to find only three examples of Scheidemann’s “free” compositions in the Sweelinck tradition. Moreover, it is misleading to assert that these three works constitute the “least numerous group in Scheidemann’s harpsichord œuvre” (p. 5). Not only is it anachronistic to speak of Scheidemann’s creating an “œuvre” for harpsichord or any other particular instrument, but the spotty transmission and preservation of Scheidemann’s music makes it pointless to count up the pieces belonging to categories that in any case are the inventions of modern, or rather nineteenth-century, musicologists. The volume might well have incorporated additional pieces belonging to the same “free” genres, especially as these works are more substantial and musically satisfying than many of the pieces included here.

9.5 Instrumental designation is not the only area in which conclusions are reached with scant documentary support. Two household manuscripts containing primarily little dances are dated to the 1650s and 1660s: “since they were both compiled during the last years of [Scheidemann’s] life, they can presumably be considered characteristic for a late phase of his œuvre” (p. 6). The logic is flawed, since we know that music as early as Bull’s was still being copied after these dates. To be sure, some of the little dances appear to show the influence of mid-seventeenth-century French music for lute or harpsichord. But the dating of such pieces is notoriously difficult. Equally problematic is the grouping of dances into what we would call suites. Dirksen avoids this latter term, but he does print each of two dance pairs as a single item under the title “Allemand & Courant.” The pieces in question do occur together in their sole source (the so-called “Celler Clavierbuch” of 1662),⁶¹ but each movement bears its own title and attribution, the pairing being made explicit only in one of the two cases, where the word “sequitur” is included in the title of one of the courantes. This seems a weak basis on which to conclude that “Scheidemann sought to give more weight to the short dances … by pairing an allemande with a courant [sic]” (p. 6).

9.6 Here, as with his designation of these works as harpsichord pieces, the editor is attaching authorial responsibility and authority to what might well be a chance or arbitrary circumstance of the pieces’ transmission. In fact, the dances in question (nos. 17 and 18) are among the more substantial in the book, showing what may be a relatively late style. They may well date from years when Froberger and other German-speaking composers were beginning to group their pieces into actual suites. But the unattached allemandes, nos. 13 and 16, show the same style in which the straightforward rhythms and simple tunes of the late-Renaissance alman and coranto—present in other dances in this volume, as well as in English examples—are replaced by a more elusive and elegantly embellished style. In addition, the texture involves more varied and irregular types of brisé arpeggiation.

10. Attribution Again

10.1 The most pressing problem is once again that of attribution, here complicated by the fact that a number of the selections are versions of music by other composers. Of thirty-two distinct movements, only one has more than a single independent attribution to Scheidemann; twenty-three are unica, and another eight bear an attribution in only one source. The greatest number of attributions are the fourteen in the Celler Clavierbuch, nine of which are unica. Three unica from another manuscript, known as the Drallius Tablature of 1650,⁶² are actually simple settings of pieces by other composers. Two of the unica are preserved anonymously, and the one piece that does have independent attributions to Scheidemann, the Courante no. 21, is an arrangement of a popular melody; two sources contain an additional variation which one of them labels “E: C D:”—perhaps a conflicting attribution. Needless to say, all these factors point to treating some of Dirksen’s attributions with skepticism.

10.2 The most significant attribution issue surely involves the Dowland arrangement, one of two pieces taken from an Austrian manuscript notorious for its poor texts and attributions.⁶³ Both are anonymous there, although the other piece (the “English Mascarade,” no. 8) has an attributed concordance. The two pieces were previously published by Werner Breig, who assigned the “Lachrimae” arrangement to Scheidemann on the basis of unnamed similarities in its melodic decoration to that in Scheidemann’s “monodic” chorale settings, as well as its proximity in the manuscript to two other works elsewhere attributed to Scheidemann.⁶⁴ The attribution is plausible, and some of the embellishment figures used here do match those found in Scheidemann’s chorale settings.

10.3 The placement of this piece immediately before the more firmly attributed Galliard , no. 7, naturally encourages pairing of the two in performance, even though they are in different tonalities. Yet the source of the Galliard presents another possibility: the Düben tablature opens with Paul Siefert’s arrangement of another Dowland pavan, known as “La mia Barbara,” followed by the Galliard. Both are in the same tonality, and as the manuscript contains three other such pairings, the first one may well have been deliberate, despite the involvement of two different composers. Of course, this provides evidence only for the copyist’s or owner’s preference, not Scheidemann’s, although as fellow Sweelinck students, Siefert and Scheidemann might well have known each other’s works.⁶⁵

10.4 The attribution issues in the many smaller dances might be considered trivial, given the musical insignificance of some of these pieces. Particularly rudimentary is the Allemande, no. 15, whose poor quality is noted in the textual commentary; it is attributed to “H.S.,” the only attribution of this form in its sole source.⁶⁶ But what exactly does it mean to assign any of these pieces to Scheidemann? The “Mascarata & Variatio” in G, no. 10, is a more finished piece, but it is anonymous in its sole source, and it has a melody concordance elsewhere. It is assigned here to Scheidemann on the basis of the style of the figuration in the variation, although the editor was surely influenced by the fact that the same source includes three other works by Scheidemann, one of them anonymously.⁶⁷ In this case it seems reasonable to speculate that Scheidemann made an arrangement of a popular melody, presumably for a student or amateur patron, and then wrote a variation on it.

10.5 Especially problematic are short dances described here as after (nach) Frescobaldi (no. 28), Gibbons (no. 20), Robert Johnson (no. 9), and Nicolas La Grotte (no. 24). At the very least, these dances illustrate the international sources of this repertory. Dirksen describes the Frescobaldi work as “a slightly revised version” (p. 6). But although it adds some rather obvious ornaments and a few minor rhythmic adjustments, the only musically significant alterations of Frescobaldi's published text involve the simplification of chords, sometimes by omitting an essential third (leaving an open fifth or bare octave); in the final 4–3 cadential progression the dissonant fourth is eliminated as well. The impression is not of a revision, but of a somewhat careless copy whose idiosyncracies diminish rather than enhance the music, eliminating intentional irregularities and refinements in the original. It is instructive to see that a northern musician—whether Scheidemann or a contemporary—engaged in the same inexact copying practices described by Silbiger for other Frescobaldi copies, and by David Fuller for Chambonnières.⁶⁸ But it seems wrong to consider what is in essence a faulty transcription—the source is a tablature of a work originally published in score—as a “work” of Scheidemann.

10.6 The Courant, no. 20, said to be “nach Orlando Gibbons,” is a simplified version of “Lady Hatton’s Galliard,” a keyboard piece with varied reprises otherwise known from three English manuscripts. The present version lacks the varied reprises, and, although clearly derived from Gibbons’ composition, betrays the same stripping away of refined details that might happen when a piece by a good composer is copied carelessly or written out on the basis of inexact memory. The piece is only sixteen bars long (without repeats), yet Gibbons’ texture broadens into four real voices before each cadence. Scheidemann’s version is in three voices throughout and at one point eliminates a lovely, characteristically English harmonic juxtaposition: an ascent to c''-natural in m. 5, following c'-sharp in the previous bar, is reduced to a banal ascent to a', merely prolonging the A-major harmony. Gibbons descends to A', a note found with some regularity in Jacobean keyboard music. Scheidemann avoids this while seemingly exploiting the availability of a short octave; the simultaneous notes D, d, and f in measure 8 are probably all to be taken by the left hand.

10.7 The “Courant and Variatio nach Nicolas La Grotte,” no. 24, presents a more complicated situation. Dirksen describes it as “based on a lute courante” by Nicholas La Grotte (ca. 1530–ca. 1600). La Grotte was organist and joueur d’épinette to Antoine de Bourbon, King of Navarre and father of Henri IV; he was later organist to Henri III. La Grotte was thus a keyboard player, not a lutenist as stated here (p. 81). Hence it seems possible that the version of the Courante published by Besard in 1617 was an arrangement of an earlier keyboard version. The present keyboard version is significantly more elaborate than Besard’s, so remote from it that one cannot be certain that it does not derive independently from another source. In particular, this version is followed by a double (variatio), whereas the lute version incorporates varied reprises within a single text. Conceivably, then, this version, attributed to Scheidemann in one of its two sources, provides a rare link back to the largely lost French keyboard repertory of the late sixteenth century.⁶⁹

11. More Conclusions

11.1 For works that have already appeared in print, Dirksen’s texts improve in small ways on previous editions. It is good to have fingerings from seventeenth-century sources that were omitted in earlier editions, even if they do not always seem to make sense.⁷⁰ I noticed one minor error in the “Lachrimae” pavan (m. 92), where the tenor’s dotted half d' should read as two notes, a half followed by a quarter. In the variations on “Betrübet ist zu dieser Frist,” no. 4, m. 2 contains parallel fifths between the two lowest voices, but the half-rest shown here as belonging to the alto should actually be in the tenor part, as is clear from a facsimile of the autograph tablature (p. 9). Other small errors in voice leading in the same piece, however, are present in the source, confirming the impression of Scheidemann as a somewhat careless copyist or transcriber.⁷¹

11.2 Perhaps more significant is the inconsistent listing of variant readings. Spot-checking revealed that the list of “ergäntze Pausen” for “Lachrimae” does not include some added rests (mm. 8 and 11), nor do the “Einzelanmerkungen” mention some more significant variants (mm. 11, 13, and 74).⁷² If these problems are representative of the edition as a whole, then the scholarly apparatus may not be as dependable as it appears to be. Certainly the editor might have provided more information about the sources, which have been little discussed outside publications that are difficult to find even in major research libraries (at least in North America). In the absence of even summary inventories, the reader is forced to rely on the editor’s assertions about the “harpsichord” character of these manuscripts. A promised study by the editor may explain the basis for his assertion that “the pieces in the Düben tablature were probably copied from autograph and perhaps even corrected by the master himself” (p. 7). This hypothesis, similar to that made for the Bull source Me and other virginalist manuscripts, would elevate the copy to the status of a primary source if true. Particularly curious is a second layer of corrections, relegated to the list of variant readings, that seems to have involved a systematic revision of the division-like passagework. These corrections substitute trill-like figuration for some of the circolo figures. But Dirksen provides no argument to substantiate the claim that some of the sources used here “were apparently written in the immediate entourage of the composer” (p. 5).

11.3 As with the Bull volume, the diversity of style and quality found here makes it difficult to draw conclusions about the composer himself or the significance of the music. We might imagine that the smaller pieces were used in teaching, the larger ones for professional study and performance, perhaps even in public organ recitals; and we might suppose that the pieces more closely resembling music by Sweelinck or the virginalists are relatively early—but there is no real evidence for any of these guesses. Particularly intriguing is the inclusion of “Mio cor, se vera sei salamandra,” an embellished transcription of a madrigal by Felice Anerio published as early as 1600. This piece is attributed to Scheidemann in the Düben tablature in a unique copy dated Oct. 12, 1643.⁷³ The original is an undistinguished setting in five voices, conceivably employed for the teaching of stile antico counterpoint; it contains not a hint of the more expressive types of madrigal being composed already in the 1570s and nothing of Monteverdi’s seconda pratica. Seventeenth-century listeners might have found the madrigal congenial because of the uncomplicated tunefulness of many of the simple points of imitation and the rhythmic regularity of its phrases, which in the modern edition tend to fall into four-measure units; at the end they form a sequence descending by thirds. Imitations at the unison between the two soprano parts create echo effects that might have been particularly attractive to Sweelinck’s students. Scheidemann’s keyboard arrangement is a decorated version of a literal transcription. It is quite straightforward—attractive, but neither inventive nor rigorous. The echo effects are not translated into echoes at the octave, and repeated or imitated passages are not necessarily embellished in the same manner, producing the same mixed impression of the composer as do the simpler dance arrangements.

11.4 One is glad to have both these volumes. Yet, especially in the Scheidemann edition, it is disappointing to find so many traces of an old-fashioned zeal to find clear, definite answers to insoluble questions of attribution, chronology, and instrumental medium. The titles and internal organization of both volumes make sense within the conventions of the composer-oriented edition, but those conventions prevent the editor or user from confronting the issues that scholars such as Silbiger and Fuller have been raising for the last thirty years. The result in each case is an apparently miscellaneous collection of pieces that may well reduce the reputation of the composer in the eyes of many players—surely not the end intended by either editor.

11.5 In our present postmodern deconstructive age we are no longer tied to the convention of the composer as the sole font of music—and we are quick to recognize approaches in which this is the unstated premise. A more current approach might be to accept explicitly, as a positive feature of these repertories, all uncertainties as to medium, date, or composer: each musical text may be available to various instruments and, as in some medieval repertories, may be the product of several hands working at different times and places. Nevertheless, interest in authors is unlikely to vanish; anonymous music for an unspecified medium is almost never performed and hard for publishers to sell. In the case of the Scheidemann volume, an appropriate compromise might have been reached by adopting the title Heinrich Scheidemann and His Circle: Works for Keyboard. For the editors of Musica Britannica, the task is more daunting, for no less than the entire “virginalist” repertory needs to be rethought as a coherent whole in light of recent research on the sources. Scheidemann and Bull did not write all of the music in these two volumes, but the music nevertheless represents an important slice of the musical culture of seventeenth-century northwestern Europe.

---