Home page

John Dowland

Keyboard Music

Harpsichord, virginals or clavichord

edited by Christopher Hogwood


INTRODUCTION
Although the boundaries between the lute's repertoire and that of the domestic keyboard instruments of the 17th century were very permeable, it is notable that the greatest English keyboard writer, William Byrd, wrote nothing specifically for solo lute, and the most renowned English lutenist, John Dowland, left nothing in keyboard form. Yet such repertoire was enthusiastically transferred between these media by their contemporaries; an edition of lute arrangements of Byrd's music appeared in 1976,(1) and it seems not inappropriate to balance this now with a selection of Dowland's music set for the keyboard.
Many amateur players at all levels of society were proficient on both instruments, which increases the likelihood of transferred repertoire. When James IV of Scotland came to seek the hand of Princess Margaret in 1503 we are told that "Incountyent the Kynge begonne before hyr to play of the clarychordes, and after the lute, wich pleased hyr varey much, and she had great pleasure to here hym. . .". Mary Queen of Scots was also described in the "Report of England" delivered by Giovanni Michiel in 1557 as ". . . playing especially on the clavichord and on the lute so excellently (though she now plays rarely), she surprised the best performers, both by the rapidity of her hand and by her style of playing"(2) Abroad, the Swiss student Felix Platter learned to play clavichord, lute and spinet, and before he died in 1614 had amassed 42 instruments.(3)

The melancholic John Dowland (1563-1626), the greatest English lute composer of that Golden Age, was, even in his own lifetime, singled out by his contemporaries for the range of his musical achievements and praised for his 'learning'. His compositions were more widely transmitted in both British and continental manuscripts than those of any colleagues of his generation, and his restless travels throughout Europe (working in the service of Moritz, Landgrave of Hesse and King Christian IV of Denmark, and also visiting Italy) brought foreign musicians into direct contact with his playing and repertoire as well as broadening his experience of continental styles and dance music. The pieces selected here are best seen as "reflections" on Dowland's music; they range from literal and sometimes na‹ve transfers to elaborated sets of variations and highly "collorirt" transcriptions from both British and continental sources, dating from the first half of the seventeenth century. Since several of the numbers listed amongst his lute works are probably his arrangements of pre-existing dances acquired from elsewhere, the keyboard versions may have been derived from other sources. In addition, the quality of the sources available today can present a distorted picture; the Schildt setting of the Lachyrmae Pavan, for example, transmitted via a meticulous copyist, and the Sweelinck version from the untrustworthy Bártfa manuscript relate, in the words of Alan Curtis, "much like a beautifully preserved painting by a minor artist compared to a mediocre copy of a great master's lost painting".
Nevertheless, these settings not only give a context for Dowland's work, and provide a new repertoire for keyboard players, but they may also act as suitable models for modern keyboard players wishing to adapt other lute music of this period. Those English sources which are already available in accurate modern editions (especially the volumes of Musica Britannica devoted to Bull, Byrd, Farnaby and Gibbons) have been listed (appendix 1) but not transcribed here, and preference has been given to contrasted versions, both English and foreign, not previously available or not easily juxtaposed. At least one version of every Dowland piece known to exist in a keyboard transcription is included here; in the case of the most popular items, such as Lachrymae Pavan or Piper's Galliard, a selection has been made, favouring some of the less accessible sources. Where only a single version is known (Lady Hunsdon's Puffe, for example, or Lady Rich's Galliard) that setting is included, however basic. With a choice, such as the two settings of Come again, sweet love, the elaborated setting from the Turin tablature has been preferred to the very amateur attempt preserved in the London RCM source.
Inevitably, Dowland's most famous single creation, the Lachrymae Pavan, survives in the greatest number of adaptations, both for keyboard and consort. Its earliest form seems to have been as a lute solo: manuscript versions are found from the 1590s, and the piece was printed from wood blocks in William Barley's New Booke of Tabliture (1596). The song setting, "Flow my teares", is first found in Dowland's Second Booke of Songes or Ayres in 1600, with a metrically irregular text which seems to have been devised to fit the pre-existing melody. As "Lachrymae Antique" it heads the sequence of related pavans and other dances for consort which Dowland published in 1604 as Lachrimae, or Seven Teares Figured in Seven Passionate Pavans, with Divers other Pavans, Galiards, and Almands, Set Forth for the Lute, Viols, or Violins, in Five Parts.
The earliest (and most "English") lute version was in G minor (Tone 2), which gave rise to a number of continental rearrangements in that key. Other versions of Lachrymae in A minor (which fits the six-course lute better) and later D minor seem to stem from lost consort versions, or from the vocal setting (in some cases reduced to a two-part framework) which appears to have been the basis of many continental arrangements. Dowland's consort version is particularly rich in internally related thematic part-writing, and thus was specially attractive to the keyboard arrangers, for whom such complex textures were idiomatic. In cases where this counterpoint is preserved or glossed in the versions of the North German school, or in other instances where a quirk of the vocal setting is retained, the line of descent from Dowland is detectable. Gale and Crawford (see literature) have noted the variants in many continental lute and consort versions and pointed out, for instance, similarities between the virtuoso keyboard decorations added by Schildt and Scheidemann and settings by the lutenist Joachim van den Hove (in Florida, 1601, and later sources).
In all, with over ninety settings of Dowland's pavan and closely-related pieces now traced, it appears it was used, in fact, "as a starting point for compositional and performative elaboration rather than as a sacrosanct entity" (Gale and Crawford). The various settings of Lachrymae presented here, grouped by tonality, show a wide variety of techniques, from basic intabulation of the three strains, without elaborated reprises, to fiery and virtuoso decorated versions, with full repeats in the North German manner.
The settings also show the wide divergence in chromatic solutions and false relations that different composers could espouse. The spectacular clashes of Dowland's original settings (notably bars 4-6 and 23) were seized on enthusiastically by some writers, but modified or eliminated by others. Other versions reproduce such quirks as the late sharpening of a cadential figure (see no. 11) often found in tablature notation (where ficta is inapplicable). Individual niceties also include such decisions as whether to leave the rising sequence in the second strain (bars 23-5) with bare left-hand fifths (as chosen by Schildt and Paris BN 1186) or filled out with added thirds (as by Scheidemann and even more daringly by William Randall). The linking of two phrases with a motto suggested in the last bar of the first passage (such as the repeated-note links between bars 20 and 22 and 63/64 with 65/66 in no. 6) can also be seen in the Lneburg version of Piper's Galliard, no. 15, bars 30 and 32-5). Sometimes Dowland's familiar melodic line is altered (Sweelinck appears to raise the peak of the melody to A rather than Dowland's G in bar 6 of the pavan, though this is probably poor copying and has been relegated to the commentary here), and the opening four-note "tear" motive is entirely eliminated by Randall ( no. 5) - a rather puzzling deviation, maybe hinting that this tag was already showing fatigue. Giles Farnaby, on the other hand, gives a precocious pre-echo of the figure in the tenor part of his Lachrymae parody in advance of the melody itself (see Fitzwilliam Virginal Book [no. 290]).
In England and abroad, parodies and echoes abound in settings that allude to Lachrymae, yet are far from straightforward transcriptions. We find substantial references in Orlando Gibbons' Pavan: Lord Salisbury (Musica Britannica vol. xx, no.18), where the references are re-allocated to different strains (see Brown, p. 59); similar quotations are found in other pavans by Gibbons (Musica Britannica nos. 15-17). The 'Pavana' of Thomas Morley (Fitzwilliam Virginal Book, pp. 272-4, mentioned above), is followed by his galliard treatment of the Lachrymae motif, (pp. 274-5), both in A minor. The distinctive "tear drop" opening was borrowed for many consort pavans (Holbourne, Weelkes, George Kirbye and Tomkins in England; Schein, Brade and Johann Schop inter alia abroad), although the continuations are all independent of Dowland. Even Dowland's own Piper's Pavan appears to reflect on a phrase of Lachrymae in its second strain (see nos. 10 and 11). Lachrymae is distinctive in having no regular companion galliard, although several pairings were proposed in the 17th century: a galliard by James Harding seems to have most often been used in this position (see Drexel 5612, p. 188, Forster Virginal Book, p. 380, and Fitzwilliam Virginal Book nos. [121] and [122] by Byrd). The copyist of Tisdale's Virginal Book added to the Dowland galliard Can she excuse the suggestion ". . . and may serve to Lacrimae" (see no. 19), and an alternative candidate from the Ben Cosyn Virginal Book (loosely derived from the piece Dowland added - as an afterthought? - to his last book of songs, A Pilgrimes Solace, in 1612) is also printed in this collection as no. 22.
Like Dowland's original publication, this collection also contains "divers other Pavans, Galliards, and Almands". Of these remaining dance forms, the alman was, according to Morley, the nearest in speed to the pavan (though frequently performed more flippantly today); Lady Layton's Almain has a variable sequence of short and long strains, differently interpreted in the two versions given here (nos. 25 and 26). The final strain of no. 26 is clearly corrupt and has been corrected and completed editorially. Two exceptional galliards are the Frog Galliard (11-18), said to have been nicknamed after Queen Elizabeth's French suitor, the Duc d'Alen‡on, which has two rather than the normal three strains, and the exuberant King of Denmark's Galliard, which is a medley of established "battle" phrases familiar from such pieces as Byrd's Battell, set in warring tonalities, and flamboyantly varied by Scheidt. Pieces from the N”rmiger, Linz and Florence sources, by contrast, offer simple basic settings, again probably of pre-existing tunes, which would benefit from elaboration in performance.
The Paris source (Bibliothéque Nationale, R‚s. 1186) from which nos. 21 and 25 are taken also deserves mention for the long sequence of largely untexted Dowland songs included between V. 6r-13r and 58v-60v (see appendix 2). These basic settings (taken from the published song books and with their English titles) are patently short-score versions, possibly intended for voice(s) and continuo, rather than keyboard transcriptions. Although they could form the basis for improvised arrangements, the random allocation of parts to left- and right-hand staves, irrespective of which hand plays, and some impossible stretches disqualify them, as a whole, as idiomatic keyboard arrangements. (Interestingly, this source also contains more evidence of transferred repertoire with "A lute lesson" (f.66r; see English Pastime Music 1630-1660, ed. Martha Maas (Madison, 1974), no. 46) for keyboard which is anonymous and unknown in any lute source.) However, one such intabulation, a two-stave setting of Sleep, wayward thoughts, is included in this collection from an alternative source, BL Add. 15117, since, although the division between staves is that of a short score, it poses no problems of layout and is practical for keyboard performance (no. 30).
The choice of suitable keyboard instruments is wide; we know that amateur musicians in Europe, such as the aforementioned Felix Platter, often played both the lute and domestic keyboards such as the spinet and clavichord. The technical term "organ tablature" is not prescriptive, but simply the accepted technical description of the notation used by many German keyboard players but applicable to a variety of instruments. For modern performance the virginals, spinet, harpsichord, organ and clavichord are all equal contenders. Two pieces in the volume (La mia Barbara, no. 9, and Come again, sweet love, no. 29) call for short-octave tuning in the left-hand, the first in bar 42, the second in bars 10, 24 and 76, which can easily be adapted for instruments without this facility.
Dowland's titles (in their modern spelling) are used in the main text, and the original manuscript readings are given in the Editorial Notes, along with details of each source, variant readings and editorial corrections. The order of pieces is by dance genre and within that by title.
A full listing of all known keyboard arrangements of Dowland's music is included as appendix 1 to this edition. For ease of comparison with the lute originals, the Poulton (P.) numbers are included there, referring to The Collected Lute Music of John Dowland, ed. Diana Poulton and Basil Lam, where the music is given both in tablature form and in keyboard transcription. Where possible, dates are also given for the MS sources in this listing, though these should not be construed as dates of composition.


EDITORIAL METHOD
With such variety of sources, it would be counter- productive to impose a uniformity on these pieces; tablature and stave notation each offer different and discrete information (chromatic exactness, for example, from the first, variable beaming and hand-division from the second).
With sources in stave notation the original note values have been retained, but pieces from tablature have been transcribed in the values used for the same piece in stave notation to aid comparison (i.e. a minim pulse for a galliard rather than the conventional crotchet or, sometimes, quaver). Early key signatures often affected only the tonality of the register indicated and no other octaves (see the low Bs in the right-hand stave of no. 28, which would have automatically been considered natural unless specifically marked with a flat). The Italian source for no. 28 mixes the use of # as well as a dot to indicate sharpening. In the present text the indication of accidentals is modern, in accordance with Edition HH house style, and any ambiguities in the original - usually over whether a marking was assumed to have an 'outreach' for the remainder of the bar, or even beyond the bar-line - have been resolved with editorial accidentals (in square brackets), either reinforcing or cancelling the earlier indication. Editorial accidentals not preceded by a "real" accidental are suggestions only, and it should always be remembered that ficta is, to some extent, a facet of personal interpretation. Editorial interference aims to be minimally intrusive, and instances of 'odd' behaviour (the delayed chromatic alteration of cadences, for example, in no. 11) can easily be adapted by any player who would rather anticipate the accidental.
The time signatures and almost all the barring of the originals have been retained; shorter bars were frequently used for clarity in the florid reprises, although this gives some short bars, for example in pavans. In the few cases where editorial bar-lines are added, these are indicated with short lines (through the staves but not the system). The use of "Rep." to identify a decorated repeat in the FVB has been applied here editorially to other pieces, and the strains have been numbered. The copyist of the FVB also added an extra 'calligraphic' final chord to many of his settings, but this should probably be omitted in performance (see no. 20 bar 13), and possibly also the extra chord that ends The King of Denmark's Galliard (no. 24). Original beaming in stave notation has been retained wherever it can be shown to have a potential effect on performance (marginally useful divisions are mentioned in the Editorial Notes), though the direction of stems has been modernised. However tablature and other sources where the beaming follows a fixed pattern have been rendered in normal modern groupings. The allocation of music to right-hand and left-hand staves by 17th century convention indicates the division of hands, and has been retained, or indicated with \ where a stave change is unavoidable in modern transcription. Voice-leading indications have been added rarely, and only when the existing notation could be confusing. The fingering indicated in BL Add. MS 30485 (no. 13) has been retained; no other piece in this selection has original fingering.
Single- and double-stroke ornament signs remain a topic of on-going research; to date no definitive translation has been found, and possibly never existed. The symbols are variously written through the tail, note head, above or below the note or stave, although this does not seem to imply a different interpretation. Since a single-stroke ornament is sometimes found on the highest available note of the keyboard it must (at least there) involve a lower auxiliary: a lower mordent or appoggiatura is suggested. For the more frequent two-stroke sign, a wide choice could include short and long trills (probably beginning on the written note) or short appoggiaturas; two simultaneous double-strokes, as occur in The Frog Galliard (no. 16) and the TVB Lachrymae Pavan (no. 1), are exceptional, and may simply imply a rolled chord. Many extended trills are fully notated in the original sources, though with a varying number of notes in each group; where six notes replace the expected eight (see, nos. 3, 5 and 8), this has not been indicated as a sextuplet but simply as six notes, to allow the player to allocate or expand the figure at will; otherwise concordant sources often vary on this point, and the figure was obviously regarded as being variable ad libitum. Those settings which contain no notated ornaments are nonetheless open to any extempore additions the player may feel tempted to venture; 'naked notes' were not a normal feature of 17th-century style.

Christopher Hogwood
Cambridge, May 2005

LITERATURE
Bailey, Candace, Seventeenth-Century British Keyboard Sources (Warren, MI, 2003)
Breig, Werner, 'Die Virginalisten und die deutsche Claviermusik der Schütz-Generation,' in Wulf Konold (ed.), Deutsch-englisch Musikbeziehungen: Referate des wissenschaftlichen Symposions im Rahmen der Internationalen Orgelwoche 1980 (München, 1985), 51-74
Brown, Alan, 'England' in Alexander Silbiger (ed.), Keyboard Music before 1700 (London, second ed. 2004), 23-89
Curtis, Alan, Sweelinck's Keyboard Music (Leiden, second ed. 1972) Dirksen, Pieter, The Keyboard Music of Jan Pieterszoon Sweelinck (Utrecht, 1997)
Gale, Michael and Crawford, Tim, 'John Dowland's 'Lachrimae' At Home and Abroad', The Lute: Journal of the Lute Society, vol. XLIV (2004)
Hogwood, Christopher and Brauchli, Bernard, 'The Clavichord in Britain and France: A Selection of Documentary References before 1700', in De Clavicordio VI: Proceedings of the VI International Clavichord Symposium, Magnano, 10-13 September 2003 (Magnano, 2004), 177-84
Holman, Peter, Dowland: Lachrimae (1604) (Cambridge, 1999)
Hunter, Desmond, 'The Applications of (Ornamental) Strokes in English Virginal Music: A Brief Chronological Survey', Performance Practice Review 9 (1996), 66-77
Ladurie, Emmanuel Le Roy, The Beggar and the Professor: A Sixteenth-Century Saga (Chicago, 1997)
Merian, Wilhelm, Der Tanz in den deutschen Tabulaturbüchern (Leipzig, 1927)
Nagan, Doron, 'A German Organ Tablature Manuscript at the Hague Gemeentemuseum', Visitatio Organorum; Festbundel voor Maarten Albert Vente (Buren, 1980), vol. ii, 439-56)
Poulton, Diana, John Dowland (Berkeley, 1972; second ed. 1982)
Poulton, Diana and Lam, Basil (eds.), The Collected Lute Music of John Dowland (London, 1974; second ed. 1978)

ACKNOWLEDGEMENTS
Permission to consult and publish has been kindly granted by:
Biblioteca Nazionale Centrale, Florence (Dr Isabella Truci), Biblioteca Nazionale Universitaria, Turin, British Library, London (Chris Banks), Biblioteka Jagiellonska, Kraków (Agnieszka Mietelska-Ciepierska), Bibliothčque Nationale, Paris (Catherine Massip), Bibliothek der Oberösterreichischen Landesmuseen, Linz, Bodleian Library, Oxford (Peter Ward Jones), Christchurch Library, Oxford, Fitzwilliam Museum, Cambridge (by kind permission of the Syndics), S. R. Lancelyn Green Esq., (Poulton Lancelyn), Minoritenkonvent, Vienna (Fr. Karl, OFM Conv.), Museum of London, National Library of Scotland (by kind permission of the Trustees), Music Division, New York Public Library for the Performing Arts, Astor, Tilden and Lenox Foundations (Robert Kosovsky), Országos Széchényi Könyvtár, Budapest, Ratsbücherei, Lüneburg (Rolf Müller), Universitetsbibliotek, Uppsala, and Dulwich Picture Gallery, London (by kind permission of the Trustees). Assistance is also acknowledged from Deutsches Musikgeschichtliches Archiv, Kassel (Dr Rainer Birkendorf), Isham Library, Harvard University (Douglas Freundlich), Derek Adlam, Hannah Barber, Daniele Brancaleoni, Lucy Carolan, Tim Crawford, Pieter Dirksen, Michael Gale, Peter Holman, Curtis Lasell, Michael Millard, John Milsom, John Robb, Anthony Rooley, Stephen Rose, Michael Talbot and John Wellingham.

(1) Nigel North (ed.), William Byrd: Complete Works for Solo Lute, (Oxford University Press, 1976).
(2) See Hogwood and Brauchli, 'The Clavichord in Britain and France: A Selection of Documentary References before 1700' for a summary of the evidence for the presence of the clavichord in England.
(3) See Emmanuel Le Roy Ladurie, The Beggar and the Professor, A Sixteenth Century Saga (Chicago, 1997).