CHAPTER 6

BECAUSE NEARLY ALL THE SURVIVING manuscripts of lute music in England were written by amateur scribes, even knowing the name of the scribe rarely provides sufficient data to establish a history of that scribe's activity. Serious, that is professional, players rarely bothered to write their names all over their books: this was a practice of the amateur. Indeed, it seems likely that professional teachers did not own personal book collections, only loose leaves that could be easily copied one at a time by a pupil. Again, this is an area where circumstantial evidence is all that is available. It has been argued that teachers'books may not have survived because of the increased wear-and-tear that one might expect them to receive. However, the books of a professional player like Matthew Holmes have survived in very good condition, and many other surviving books show evidence of a long and active life. Considering the number and diversity of the surviving sources, surely at least one teachers' book,should such a thing have existed, would have survived among all the others.

It is very rare for a book to survive without additions by later owners to the original: what is interesting is the attitude of subsequent owners to those who came before. The inscription on the front flyleaf of Thistlethwaite has been so comprehensively scratched out that it is unreadable even under ultra-violet light; the inscription of Board has been damaged, obscuring the date, and Henry Sampson's name has been heavily crossed-out on f.7 of Sampson, though the title of the piece is left intact. The missing front end-paper in Sampson (1a) may have been a title page to the collection which was removed at the same time as the ascription was deleted, as part of the process of removing all traces of Sampson's previous ownership. It appears from these deletions that evidence of a previous owner of a book had to be removed, even though the music they copied was retained and used. There are a few books that show evidence of leaves being removed, though it is difficult either to find a reason for their removal, or to ascertain at what point in its history the leaves were removed unless they were removed before copying, which is usually obvious. The cost of paper, a scribe's pride in the appearance of his book, and the relative rarity of bound books would suggest either a serious and unrecoverable copying error, or--in the light of attitudes to ownership --a rather drastic attempt to remove evidence of a previous owner. This attitude to ownership is explored in relation to the dating of repertory, where problems in ascribing music are addressed.[1]

Pedagogical books are often well supplied with indications of the scribe's name, but since he or she is usually someone otherwise unknown it is rare to be able to establish any sort of biography for them that may give any indication of their period of activity. There are notable exceptions, such as Margaret Board, whose family records and the parish records of Lindfield in Sussex provide us not only with the date of her baptism, but also the information that she married between 1623 and 1631.[2] Jane Pickeringe, otherwise the closest scribe to Margaret Board in terms of her period of copying and parallels between the histories of their respective books, is an enigma. At present the only information we have about her is her name, though she did provide us with the date 1616 in her layer of the Pickeringe manuscript.

The biography of Edward Herbert, on the other hand, is known in extensive detail,[3] but in some respects this has served to confuse the dating of the manuscript rather than elucidate its history. When the manuscript first came to light, Thurston Dart published a lengthy discourse about its compilation based on this biographical material ,[4] and subsequent studies[5] did not question his dating to any greatextent. He came to the conclusion that the music was probably collected from 1608, but only copied into the manuscript between 1624 and his death in 1640, without adding more contemporary music to the repertory. Dart was not aware of the key scheme governing the arrangement of the book, and saw the manuscript as divided into sections related to different periods in Herbert's life, which he spent partly in England and partly in Europe, with some prolonged stays in France. Herbert was forced into a form of exile in 1624, which Dart concluded gave him the ideal opportunity to begin copying out his collection of music. This may be a correct assumption, but is in fact more likely to be false, since Herbert's ascriptions for his own compositions indicate that he could not have been active in the book before 1628, so the compilation is more likely to date from 1630 with the last pieces added in 1640.[6] The origins of the repertory are very varied, reflecting Herbert's cosmopolitan life, but even so the compilation is extraordinarily conservative for a manuscript to which additions were certainly still being made in 1640. In terms of the date of its compilation, it is out of place in this study, but going solely by its contents it is one of the most substantial extant Golden Age manuscripts, written entirely in vieil ton and preserving a repertory dating almost exclusively from the early years of the seventeenth century.

Occasionally the name and date of birth of a scribe are known, but there is no further evidence to suggest a date of the manuscript in which they are active. Where dates of birth and of copying are both known, the manuscripts compiled by young women seem to have been copied when they were no younger than the age of about 20 (Margaret Board), while the earliest age for a boy to be writing was 15, in the case of Richard Mynshall. Thus, if the name of the scribe and the date of his or her birth are known, it would be safe to suggest the ages of 15 for a boy and 20 for a young woman as being the approximate ages at which they would begin copying a lute book. Bearing this in mind, Jane Pickeringe was probably born c1595. Where the scribe is clearly a professional musician or writing later in life there are no such simplistic guidelines to follow: such scribes seem to have had a copying and playing life of 30-40 years, a not unreasonable span for any professional life. Some of the royal lutenists who held posts until their deaths were professionally active for considerably longer:[7] thus other criteria become more important when assessing their date of activity.

SSSCRIBAL CONCORDANCES

There are numerous instances in which scribal concordances for sources have been identified but that information has contributed virtually nothing to the dating or other information about either source. There are probably as many cases where the identification of a scribal concordance (or its absence) does have repercussions, either on the dating of a particular source or on our understanding of the repertory as a whole. The copying life of an amateur lutenist is likely to be considerably shorter than that of a professional musician who would continue to add to his repertory as new pieces became available. The work of the known amateur players like the scribes of sources such as Dallis, Pickeringe, Board, Sampson, ML, seem to be limited to a considerable burst of activity, probably at the time they were working with a teacher, and no other appearances. The more professional players, however, have a tendency to appear in more than one source, and possibly over a considerable time-span. John Dowland is one such known scribe, as is the scribe tentatively identified as Richard Allison. Women in particular seem to have learned to play the lute in the early years of adulthood, before marriage, particularly if they remained unmarried into their 20s, as mastery of the instrument seems to have been considered a highly marriageable trait.

Essentially, the identification of scribal concordance is usually acceptable unless it thereby overturns previous entrenched ideas about the sources. The insurmountable problem in identifying a scribe is that the answer can never be proved, and so even highly detailed analyses of hands by a number of experts may be discarded. The examination of scribes has been discussed in Chapter 4 and, since the subject is particularly controversial and therefore requires considerable attention, some specific cases that have raised particularly contentious problems have been examined in detail in Chapter 7. Table 19 summarizes the incidence of scribal concordances in the English sources.

Scribal practices in some ways are particularly predictable.The cost of paper and particularly ruled paper in bound books meant that scribes were almost always highly conservative in the use of paper. Blank folios are almost never left for no reason. In Herbert sections of folios have been left blank to accommodate music that was intended to be copied in the same key as that preceding, which is copied continuously without any empty pages. In sections where the quantity of music exceeded the space available, some of the spaces reserved for other keys were used. The compilation of Hirsch implies that the scribe was keeping dance music and fantasias separate, and like Herbert apart from the large gaps between sections, blank folios are not to be found within groups of pieces. Frequently, a scribe may return to early portions of a book and fill single or double line spaces left by longer pieces that did not fill a page with short song settings of arrangements of popular tunes. In the case of Board, the scribe was clearly filling the smaller gaps with these types of piece as she went along.

For this reason, if the work of one scribe is interrupted by the work of another, it is almost certain that they were working in the book contemporaneously. Very often, the secondary scribe changes or corrects the work of the primary, as in Sampson and Board, and in some cases a piece of music may be started by one scribe and completed by another, as in Hirsch and Swarland. In pedagogical books the only fully satisfactory explanation for the 'invasion' of one scribe into the copying of another is that the second scribe was teaching the first, and the style of the invasion usually justifies this supposition. Personal anthologies seem to have encouraged the intrusion of secondary scribes, as the purpose of these books was more social. In these sources, sections of different scribes interspersed with the principal scribe are not unusual, but even so, the practice of maintaining an uninterrupted flow of music without blank folios seems to have been important.

End-papers were not usually used as copying pages, mainly because the quality of the paper was not of a 'writing' standard. Normally, only pen trials and possibly an ex libris inscription are to be found here. That Dowland used the flyleaf and not the first folio in Board for his table of mensural equivalents indicates that he probably wrote it after the first folios were copied, though he may have preferred this leaf as it was not ruled. If he had been active at the inception of the book though, one might expect to find some evidence of his activity in the early folios up to 11v.

In a group of about 50 sources that might be optimistically 100f the original generation, one would not expect to find many scribal concordances. Six of these sources, the Holmes books, were written more or less end to end by a single scribe, however, bringing the real total down to 44. Even so, the number of such concordances again exceeds expectations, in some cases quite dramatically. The secondary scribe of Sampson is also the secondary scribe of Swarland, Dd.9.33and Dd.4.22,⁸ linking the largest and most comprehensive professional collection to survive with three otherwise unconnected sources, not only in apparent provenance but also in the type and purpose of the manuscripts linked. A further three sources are linked to the group through concordant primary scribes or stemmatic relationships.[9] If fragments preserving unrelated repertories and earlier sources that are limited to the previous generation of music are excluded, then the percentage of the remaining sources that are clearly linked is over 50 This statistic is 'alarming' because it shows that our ideas about the survival of the sources of this repertory may be quite substantially incorrect. These are not the only manuscripts to be linked by scribal concordance either. Those that remain are frequently linked, though rarely in more than pairs. Have we been overestimating the extent of use that a domestic instrument received and assumed inaccurately a widespread popularity for the lute, and have we also failed to recognise the unusual status of manuscripts devoted to music for this instrument?

TABLE 19
SCRIBAL CONCORDANCES IN ENGLISH LUTE SOURCES
Source	Scribe	date	Concordances
408/2 A		c1605	6402
408/2 B		c1605
6402		c1605	408/2 A
2764(2)		c1585-90	Willoughby G
31392 A		c1605
31392 B		c1605	Dd.9.33 B
31392 C		c1605
41498		c1590
60577		c1540
Andrea		c1570	Lodge C; cf. Willoughby G?
Ballet A		c1590
Ballet B		c1610	Folger D
Ballet C		c1610	31392 B; ML C?
Ballet D		c1610	Swarland C
Board A	Margaret Board	c1620	Hirsch B
Board B	John Dowland	c1620	Folger
Board C		c1625	ML C
Board D		c1625	cf. Willoughby E
Board E		c1625
Brogyntyn A		c1600
Brogyntyn B		c1600
Cosens	C.K.	c1610	cf. Pickeringe A
Dallis	'Dallis's Pupil'	1583-5
Dd.2.11	Matthew Holmes	c1585-95	Holmes books
Dd.3.18	Matthew Holmes	c1585-1600	Holmes books
Dd.4.22 A		c1615
Dd.4.22 B	?Richard Allison	c1615	Sampson B, Swarland B, Dd.9.33 C
Dd.4.22 C		c1615
Dd.4.23	Matthew Holmes	c1600	Holmes books
Dd.5.78.3	Matthew Holmes	c1595-1600	Holmes books
Dd.9.33 A	Matthew Holmes	c1600-1605	Holmes books
Dd.9.33 B		c1600-1605	31392 B
Dd.9.33 C	?Richard Allison	c1600-1605	Sampson B, Swarland B, Dd.4.22B
Dd.9.33 D		c1600-1605
Dd.9.33 E		c1600-1605
Edmund		c1635
Euing A		c1610
Euing B		c1650
Folger A		c1590
Folger B		c1590
Folger C		c1590
Folger D		c1590
FolgerE		c1590	Wickhambrook A; cf. Welde A
Folger F	John Dowland	c1590	Board B
Folger G		c1590
Folger H		c1590
Genoa		c1600
Handford	George Handford	1609
Herbert A	Herbert's secretary	c1630
Herbert B	Edward Herbert	c1630
Herbert C	Cuthbert Hely	c1640
Hirsch A	H.O.	c1620
?Magdalen A cf.
Thistlethwaite H
HirschB	Margaret Board	c1620
Board A
Hirsch C		c1620
Hirsch D		c1620
Hirsch E		c1620
Krakow A	?John Sturt	c1615
ML B; cf.
Folger E
Krakow B		c1615
Krakow C		c1615
Lodge A		1559
Lodge B		1559
Lodge C		c1575
Andrea A
ML A	Margaret L.	c1620
ML B	?John Sturt	c1620
Krakow A
ML C		c1620
ML D		c1620
ML E		c1630-40
Och532 B
Magdalen A		c1605
Hirsch A
Magdalen B		c1605
Mansell		c1600?
Mynshall B
Marsh A		c1595
Marsh B		c1595
Thistlethwaite A (title)
Mynshall A	Richard Mynshall	1597	Swarland A
Mynshall B		c1600	Mansell
Mynshall C		c1605
Nn.6.36 A	Matthew Holmes	c1610-15	Holmes books
Nn.6.36 B		c1610-15
Nn.6.36 C		c1610-15
Northants		c1625
Occ254		c1610
Och439 A		c1620
Och439 B		c1620
Och439 C		c1620
Och439 D		c1620
Och439 E		c1620
Och439 F		c1620	Dd.9.33 E
Och1280		c1580
Osborn		c1560
Pickeringe A	Jane Pickeringe	1616
Pickeringe B		c1630
Pickeringe C		c1650	cf. Folger D; Ballet B
Pickeringe D		c1630
RA58		c1530	cf. Ballet ?
Richard		1600-1603
Rowallan A	Anna Hay	c1605-8
Rowallan B	?Mary Hay	c1605-8
Rowallan C	Sir William Mure	c1615-20
Sampson A	Henry Sampson	c1610
Sampson B	?Richard Allison	c1610	Dd.4.22 B, Swarland B, Dd.9.33 C
Sampson C	[title only]	c1610
Stobaeus	Johann Stobaeus	c1635
Stowe389	Raphe Bowle	1558
Swarland A	Richard Mynshall	c1615	Mynshall A
Swarland B	?Richard Allison	c1615	Sampson B, Dd.9.33 C, Dd.4.22 B
Thistlethwaite A		c1575
Thistlethwaite B		c1575
ThistlethwaiteC		c1575
Thistlethwaite D		c1575
Thistlethwaite E		c1575
Thistlethwaite F		c1575
Thistlethwaite G		c1575
Thistlethwaite H		c1575	cf. Hirsch
ThistlethwaiteJ		c1575
ThistlethwaiteK	Thistlethwaite	c1575
Trinity		c1630
Trumbull A	William Trumbull	c1595
Trumbull B		c1605
Welde	professional copyist	c1600	cf. FolgerE; Wickhambrook A
Wemyss	Lady Margaret Wemyss	1643-4
Wickhambrook A		c1595	Folger E; cf. Welde
Wickhambrook B		c1595
Willoughby A	?Richard Greene	c1560-85
Willoughby B	?Francis Willoughby	c1560-85
Willoughby C		c1560-85
Willoughby D		c1560-85
Willoughby E	?John Edlin	c1560-85
Willoughby F		c1560-85
Willoughby G	?Richard Greene	c1560-85	2764(2)
Willoughby H		c1560-85

The common belief, predicated partly on fact and partly on a sort of common-sense extrapolation, is that the further back in time you proceed from your own point, the less of each succeeding period's belongings survive for your perusal. This is reasonable as far as it goes, since the equation: age = fragility = degeneration, is undoubtedly correct. If, as seems to be the common assumption, a calculation based on rates of survival of books of any kind in more recent times leads to a figure of around 50.000000or 1600, this needs to be offset by such considerations as the rarity and monetary value of manuscript music-books. Furthermore, the frequency of scribal concordances can only strengthen the suspicion that survivals of lute-books, in particular, represent a far higher proportion of what once existed.

Lutes appear repeatedly in sixteenth- and seventeenth-century paintings in the hands of men and, equally often, women, and they find their way into lyric poetry of most decades up to c1650.[10] That the lute and its symbolism thus pervaded Elizabethan and Stuart life is therefore not an issue. However, the problem of how many of these people--often clearly middle- or working-class--compiled lute books is unresolved. The courtesans of England, Italy, Fance and the Netherlands, whose lute was a badge of trade,[11] almost certainly did not use written music, nor, as already explained, did the professional players of the Royal Music. This would also give some explanation for the paucity of printed collections for the solo lute in England compared to printed lute-songs. The only players who compiled books were the aristocratic ones, and even these would play from memory, not music, in public. To them, though, the lute book was as valuable as the instrument--perhaps more, as the instrument could be easily replaced--and was preserved and passed-on with great care.

The number of scribal concordances between the lute sources indicates a high level of communication between them. If the number of books to survive was really only about 50f the original distribution, then there should be dramatically fewer scribal concordances. Perhaps having a pupil write a lute book was the particular practice of a specific school of teachers in a single region? This would certainly provide one explanation for the distribution of concordances. Unfortunately, however, the evidence of common exemplars that would confirm this situation, is also absent. Taking into account natural losses from fire and other accidents, deterioration through use, and the likelihood of unbound sources having a considerably shorter life, the conclusion is that a more accurate figure for the rate of survival might be 50

SSTYPE OF LUTE AND TUNINGS

The consensus among modern players and from the prefatory matter to sixteenth- and seventeenth-century publications, particularly Dowland 1610B, is that the standard sixteenth-century English lute had six courses and eight frets, and that the seven-course lute in England dates from almost precisely 1595.[12] However, the six-course instrument persisted, and certainly a lutenist would not have changed his instrument simply to stay in fashion. The overall result is a variety of six-course pieces being adapted for larger lutes by the addition of notes requiring bass courses, and the reverse change in which later music is adapted to suit a more conservative instrument. Thus the presence of these bass notes themselves could indicate the relative age of a source, but their absence might not. Euing for instance, was written by a scribe who played a six-course lute, but contains much music that was intended for at least seven courses and has been adapted, not always successfully.On the other hand it is easier to adapt a lute to accommodate additional frets than it is to dispense with the need for them in the music. Dowland 1610B refers to Mathias Mason as having invented the ninth, tenth and eleventh frets on the old English eight-fret instrument by glueing strips of wood to his soundboard, and adds that the French subsequently lengthened the neck of their instruments to accommodate ten tied frets. He implies that these improvements had been around for some time, and that the most popular instrument was the long-necked French variety, which probably had three extra frets on the belly. How long the instrument with these high frets had been in use is nowhere specifically stated, though it is probably unlikely to have been before 1600. Certainly, music that uses tablature letters above 'k' is likely to date from after 1605 or even 1610, depending on whether it also requires more than one added bass course. This is more an aid to dating repertory than sources, although some adaptations, more re-writings than simple alterations, are so successful that is impossible to see which version came first. Composers also complicated matters by improving on old pieces by re-composing them. Dowland's 'Battle Galliard' appears in Sampson (c1610) in an easier form than is found in books of more accomplished players, and may have been an early version of the piece rather than a simpler adaptation.

Lutes with large numbers of added courses were generally used late in the first half of the seventeenth century, but those with six to seven courses, although probably made early in the century, were nevertheless still in use later. Thus, although we have approximate dates for technical changes such as the addition of extra frets and courses, the lack of precise information, and the undoubtedly variable take-up of any new innovation, makes this an extremely volatile weapon in the dating armoury. Even an attempt to pin down a decade during which the seventh course came into use is inadvisable, since it may have been in use in some places long before, and was certainly not taken up by every player as soon as a few had tried it.[13]

An English manuscript that is heavily influenced by the French repertory and by trends in design and construction originating abroad may appear to be later than it really is because it is progressive in its geographical context. On the other hand, a manuscript such as Herbert looks as if at least parts of it must date from the early years of the seventeenth century because of its consistent vieil ton tuning and the lack of numerous bass courses in many pieces. Considering the unavoidable links that its owner had with the continent before and during the copying period, and the actual date of copying, it is extremely surprising that there are no new French tunings employed, and that there is so little that is truly representative of the up-to-date repertory. On the other hand, Lord Herbert used a lute with up to ten courses, which would preclude an early dating despite those features that make the manuscript a conservative product for the 1630s and 1640s.

Clearly there are numerous caveats to be taken into account when examining the type of lute a player used. Again there seems to be a significant gap between what we might expect the amateur or the professional to be using. The amateurs probably bought up-to-date instruments, but may have found themselves copying music for a more conservative instrument. In this case, we may expect to see some bass courses added to the music for extra resonance on chords that would otherwise not have had much bass range. The professional, on the other hand, probably changed instruments or altered old ones (as Mathias Mason is reputed to have done) in order to keep up with modern trends, thus assuring the popularity of his repertory and improvising style as well as making it easy to play newly composed music. A single scribe, depending on his (or her) wealth, enthusiasm or professional needs, may easily have changed his lute to suit changing fashions, and a scribe who copied over a considerable length of time, such as a thirty-year span,[14] would be likely to change aspects of his notation to accommodate the changing construction of the instrument and increasing demands on its range.

SSREPERTORY, DATEABLE ELEMENTS IN ASCRIPTIONS

The single most important feature of a source that qualifies every aspect of its dating, and particularly the date of its repertory, is the original intended purpose of the book. It has been shown in Chapter 3 that the currency of the repertory can depend entirely on the category of book. In particular, pedagogical books may be expected to preserve a particularly 'old-fashioned' repertory, while professional books are probably the only ones in which dating principally by the repertory they contain is plausible. However, if the book was not compiled in London (or, possibly, Oxford) or the immediate environment of the composers it contains,the time new music took to percolate outwards into more provincial areas places constraints on the effectiveness of this method of dating. Dating by repertory alone can, in some cases, be useful, but more often than not leads to a distortion of the historical position of the book.[15] Overall, the conclusion that must be reached is that although the repertory must not be ignored, dating by repertory alone is unwise, and even proposing a date by using the repertory as a prime factor has proved dangerous.[16] Relative difficulty or simplicity of the repertory clearly has no relevance to its date. The repertory may support other evidence and may even contradict it, but essentially its consideration must be qualified by first ascertaining the original intendedpurpose of the book.

Dallis is undoubtedly a pedagogical book, and this factor would lead us to expect it to contain a repertory from a wide chronological background with some emphasis on early music.[17] The evidently short copying span also brings its dating by Poulton and Ward[18] of 1565-80 into question, as the consistency in script implies strongly that it was copied over a significantly shorter time, even though the music in it was copied from prints dating from the early 1500s right up to Adriansen 1584. In some ways, Dallis is a rather extreme example of the divergence between the date of compilation and the actual date of copying, but since we know that it was prepared in Cambridge under the direction of a Cambridge teacher it is far more likely that the lutenist and his master would have had access to a fair library of old books of music, and particularly of the continental prints that seem to have exercised the authors of Ward 1967 to the extent that they felt justified in disregarding the scribe's quite clear statement of the date of his work. Sampson is similarly impossible to date by the repertory, and for the same reasons. This book is also one of the pedagogical books, and the repertory copied by Henry Sampson is considerably earlier than the 1609 of the watermark.

Appendix 5 lists titles or names that appear in ascriptions that provide the piece of music concerned with a date, sometimes only approximate, before or after which the piece must have been composed, thus adding to the body of information that has to be reconciled in order to reach a satisfactory conclusion regarding the can contain a number of hands, apparently all copied within a few years of each other, if not virtually simultaneously, with such widely differing characteristics that they appear to originate from chronologically opposing spheres. This is a feature of a source that cannot be used as a dating aid unless it is in the identification of scribal concordances. However, in the case of fragmentsThere is no little danger in using the appearance of a scribe's hand to determine his abilities or his profession. Comparison of the tablatures written by young amateurs and mature professionals show that usually the more neat and accurate a hand is, the more likely the writer was to have been an amateur writing under the supervision of a professional, while the more cryptic and inaccurate texts probably belonged to the professional musicians who had no particular desire for an immaculate display copy, and relied on memory to make up for the deficiencies arising from the lack of leisure-time required to copy meticulously. In addition, the professional musician, as distinct from the highly accomplished amateur, was not usually a member of the classes that employed a writing master--certainly not in England--and may have learned to write late in life. The amateur was, and one would therefore expect the amateur hand to look neater, more regular, and more aesthetically pleasing than that of the professional.

The hands of two known teachers and professional lutenists, John Dowland and the anonymous teacher of the scribes of Sampson, Dd.2.11 and Swarland (possibly Richard Allison), both wrote positively unattractive hands when compared with the elegance and visual attraction of their pupils. John Ward frequently refers to the various styles of handwriting to be seen in his discussion of Thistlethwaite:

The repertoire, style of music, difficulty and length of most of the pieces, certain aspects of the notation, even the handwriting support a dating of the MS in the late 1560s and early 1570s and a conjecture that the chief contributor was an accomplished Italian lutenist who may have composed much if not all of the music he wrote out.[19]

Subtle variations in the handwriting of scribe C point either to two scribes with almost identical writing (unlikely) or two periods of copying (more likely, but difficult to prove as there is evidence of disturbance to the original collation). Ward's comment to the effect that the handwriting suggests a particular date should be viewed with suspicion, since the 'vigorous, hastily written tablatures' of [his] scribe B look as far from the date he concludes as any in the repertory. He also goes on to point out that this very early example of the use of rhythm-change flagging in a manuscript of this date 'appears to be without parallel', but does not find it anomalous enough to re-assess his original dating. Had Ward been able to ascertain the original purpose of the book, its repertory alone might not have proved such a satisfactory means of dating the source as it eventually was.

Ward also uses flagging styles, particularly continuous or rhythm-change flagging, to suggest periods of copying in Thistlethwaite:

All of [the scribes] wrote Anglo-French tablature ... but some ... indicated note values the redundant way, providing a sign for each stroke ... others ... indicated them the economical way, providing a sign only when the value changed. ... The former system is found in all English manuscript tablatures from the 1550s up to the end of the century, and was not completely abandoned until well into the reign of James I. The latter system prevailed on the continent during the last half of the century, may have been introduced to the English with Rowbotham's reprintings of French lute and gittern books in the 1560s, and was employed almost exclusively in tablatures printed in England, no doubt because it required less type than the redundant or 'grid' system. Its use in an English manuscript of the late 1560s/early 1670s appears to be without parallel.[20]

Thistlethwaite, though probably copied in a short period of time, shows rhythm-change and continuous flagging, both of which seem to have been in use more or less currently. The only reasonably dateable alteration in flagging styles is one that is not relevant to Thistlethwaite. Sources in which mensura gallica flagging is used usually date from after 1615, and the same is true for time signatures, with the notable exception of the practices of the Dowland family, whose familiarity with continental sources led them to use time-signatures far earlier than other English scribes and sources. Mensura gallica in a source that is observably entirely English in style, content and layout would be likely to date from the second decade of the seventeenth century, though Sampson (c1610), which appears to be highly 'English', shows both types of flagging. A copyist who had contact with European printed and manuscript sources in any respect may easily have contaminated his style with a practice that seemed easier or quicker than the one he originally used. Scribes familiar with lute song are far more likely to use mensural notes for flagging, as this would match the rhythm of the vocal part. Flagging styles should clearly never be seen as independent dating evidence, but this part of the notation is not as equivocal as the script itself.

It has been seen that continental trends had very little bearing on the central English repertory, either in the genres of the music composed, or in the way it was laid out in the sources.[21] The same is true of trends in notation. In particular, any continental innovations would have been particularly slow to show effects in English music.

In the most simplistic terms, it would be reasonable to say that the greater the number, variety and complexity of the graces in a source, the later is likely to be its date of copying. This 'rule', however, cannot be dogmatically applied, since graces are those signs added by a scribe to indicate his own performing practices, or occasionally the influence of his exemplar. There is no reason why a scribe who graces his own book most freely should not copy into another book with a complete absence of signs. In fact, if a scribe is copying from an un-graced exemplar, then it would be expected that he would simply copy the graces in the original, only adding his own if he were intending to play (or perhaps teach) from the copy. The use or absence of graces may be related to the intended purpose of the book. The more personal and amateur the source, the more likely the scribe is to grace it generously in keeping with contemporary practices. Pedagogical books would therefore be likely to present the most likely candidates for gracing, a probability borne out by Poulton. Poulton observes that Folger, ML, the middle section of 31392, Sampson, Welde and Board are the manuscripts that contain the greatest number of signs in each piece, and their scribes have evolved the most sophisticated systems for indicating each carefully differentiated grace. She also indicates that in England,

Evidence suggests that ornamentation reached a higher degree of complexity at the end of the 16th and the first two decades of the 17th century than in any other country in the pre-baroque era. Nevertheless, in spite of the very large quantity of ornamented source material that has come down to us, interpretation is difficult since there was no standardization of the signs used either by scribes or printers, and we have exceptionally little information on the subject.

Early MS sources bear no ornament signs, but some of these are fairly unsophisticated in the way they are written, and it is perhaps not to be expected that methods would be found to convey the subtler aspects of playing. All the Matthew Holmes lute books ... are ornamented, although Holmes is somewhat capricious in his use of the signs. At least seventeen other MSS are also ornamented.[22]

Even with the paucity of printed lute tutors in England, we might expect those that do exist at least to touch on graces and grace signs. Although Barley 1596 includes both the + and # signs throughout the tablature, nowhere does he describe what these signs are intended to represent. Dowland 1610B includes the instructions from Besard 1603 and the essence of his advice is that a student should go and listen to a virtuoso or more experienced player to learn how to grace a piece of music. Neither describes specific graces or signs, but the Dowlands agree with Besard that graces should be learned and applied to all music, whether notated with signs or not:

You should haue some rules for the sweet relishes and shakes if they could be expressed here, as they are on the LVTE: but seeing they cannot by speech or writing be expressed, thou wert best to imitate some cunning player, or get them by thine owne practise, onely take heed, lest in making too many shakes thou hinder the perfection of the Notes. In somme, if you affect biting sounds, as some men call them, which may very well be vsed, yet vse them not in your running, and vse them not at all byt when you iudge them decent.[23]

This last reference to 'biting sounds' is somewhat enigmatic, and hardly serves to illuminate the reader. Robinson 1603, though it does not give examples of the signs to which terms refer, does list 'A Relish', 'A Fall' and 'A Fall with a Relish', with a verbal description of their respective interpretations.

TABLE 22 GRACES IN BOARD.
Sign	Scribe's description	Explanation
		Margaret Board
		probably a relish or shake starting on the main note, used before tablature letters a, b, c, d, e, f, g, h and i.
		a half-fall of a tone or semitone, used before tablature letters b, c, d, e, f, g, h, i and l. It may also mean a backfall. She occasionally combined both the fall and shake signs.
		possibly a single relish, used only on tablature letter h, and only from 10v on.
		possibly a whip (interpretation unknown) or vibrato on tablature letters f, g and h, on ff.11v and 15.
		possibly a whole-fall, only used on tablature letter d on f.20v.
		a backfall on tablature letter c on f.29 only.
		John Dowland
		shake, on tablature letters a,b, c, d, e, f and h.
		fall on tablature letters a, b, c, d and g, it may also mean a backfall depending on context.
		possibly a whole-fall (on tablature d).
		possibly a double backfall, on tablature letter g.
		perhaps a whip or vibrato on tablature letter i.
		Scribe E
	a pul back	a
backfall, used on tablature letters a, b, c, d, e, f and h
	a fal forward	a
half-fall, used on tablature letters b, c, d, e and f
	to beat down the finger with a shake	although it is ambiguous, I think this means a half-fall repeated, used on tablature letters b, c, d, e, f and g. It appears to apply to tablature letter a twice, but in each case I think the grace belongs to the neighbouring note
	3 prickes to be struck upward with one finger	one right-hand finger playing a chord from higher to lower sounding strings
	for a long shake	used on tablature letters a, b, c, d, e, f, g, h and l, presumably beginning and ending on the same note, rather than beginning on a note lower or higher than the ending note
	for a slide	slurring or hammering two or three notes with left hand fingers having plucked only the first with a right hand finger.
		These signs appear in the tablature on folios 33v and 38v, but are not explained in the table--perhaps indicating a beat.

The most useful of all the manuscript sources is Board in which one of the secondary scribes has supplied a table of grace signs and a verbal description of their interpretation. Dowland and Margaret use a generous variety of personal signs, several of which do not appear in the work of any other contemporary copyists, though Margaret's more complex signs clearly date from her contact with Dowland. The usefulness of the interpretation of these symbols is therefore limited to the single source in which they appear, but they serve to demonstrate the variety and subtlety of the graces that were commonly used by the second and third decades of the seventeenth century. The table above gives the signs used by the Margaret, Dowland and Scribe E in the left hand column and the probable interpretation suggested by Robert Spencer[24] in the right hand column, including the description of the interpretation of the signs given in the table of Scribe E on f.32v.

SSCOMPILATION AND LAYOUT

Despite the apparently lengthy periods of compilation proposed by Lumsden for some of the English sources, and the dates of sources such as Pickeringe and Board, very few sources show evidence of copying over more than a few years. Common sense suggests that the Holmes books were probably compiled end to end rather than with the overlapping periods of copying suggested by dates proposed by Lumsden, Harwood and Poulton. Because of the extended activity of their scribe, these books each show a longer compilation span than is common among many other books. As might be expected, the purpose of the books virtually defines the length of activity of the scribe. Nearly all the pedagogical books were copied only within the duration of the pupil's instruction, and perhaps for a short time after, though their actual use for playing may have been considerably longer. This may be partly due to the fact that the addition of new repertory to the pupil's collection depended on what the teacher was able to offer. Any period of learning was likely to be finite, since once the approximately desired level of competence was attained further instruction should become unnecessary, particularly if the pupil then married and found less time for the pursuit of this particular past-time. Once the teacher ceased to visit, the immediate source of new music was cut off. Professional collections such as the Holmes books would reflect the fact that the owner/scribe was continually coming into contact with new music, and would to all intents and purposes continue adding to his collection for the duration of his working life, though possibly in patches rather than purely continuously. The household or personal anthology would again show perhaps a lengthy compilation, but only as long as the owner's enthusiasm for the lute persisted; probably not as long as the single span seen in a professional book. Thus with the possible exception of the professional books,the compilation of any book is unlikely to exceed ten years,and in the case of pedagogical books less than five years. This does not take account of later additions to the book unrelated to the work of the initial scribe, though each of these self contained layers are similarly likely to be very limited in duration. Single amateur players would therefore not be expected to show more than a single layer of activity in a book. Thus returning to a book frequently over a period of thirty years, for example, would be highly unlikely, and is not seen in any of the surviving books, despite previously held ideas about both ML and Herbert, both discussed in Chapter 7. If a source is encountered where an amateur player appears to have copied in the source over a considerable time, then three more likely probabilities should be considered: first, the later appearances of the scribe were not made later; second, they were not made by the same scribe; third, the original scribe is actually not an amateur.

The layout of a source is governed by many factors, and rarely by one principle alone. From the point of view of dating, the most important factor to be understood is that if the scribe is able to determine his layout in advance of copying, he must therefore already have assembled the music he intends to copy--or at least the majority of it--and therefore the finished source is likely to contain almost entirely music that pre-dates the start of the copying period, perhaps by up to several decades. This is the sort of book in which, in rare cases, the repertory may be entirely from a significantly earlier period than the book, or it may come from a highly chronologically diverse background, and include reasonably recently composed music (assuming it was compiled within easy reach of the compositional activity).

There are two types of pre-determined layout: firstly, and most commonly found, the arrangement of the contents in order of difficulty, beginning with technically simple pieces, usually represented by a group of duets, and progressing through stages to more difficult music of a standard that might be found in any source. Books compiled along these principles are invariably pedagogical sources, preserving a repertory that pre-dates the copying period sometimes by as much as 20 years. It is the easier and therefore first-copied music that is likely to be chronologically out of step, while the later-copied and more difficult music is likely to be considerably more up-to-date. In the larger pedagogical sources, which cover a significant period of learning and progress to a high standard, the apparently anachronistic early music tends to be balanced by the later works. In the case of sources where the student does not continue copying up to the high standards seen in ML, Pickeringe and Board, the overall repertory may appear to be anachronistic, as in Sampson and Dallis. Sources with a deliberately pedagogical layout are usually easily distinguishable both from the immediately obvious arrangement by technical standard and also by the inclusion of other features.

However, pedagogical books are not the only ones to exhibit signs of pre-arrangement, and sources that use different criteria, as well as being much rarer, are also far more difficult to recognise. The arrangement of Herbert into sections by key was not recognised for many years, despite some quite detailed investigation of its contents. Despite the clearly delineated division of Hirsch into dance music and contrapuntal music (fantasias) this was not considered significant in dating terms. In both cases, the importance of the fact that the scribe must have known what he was going to copy in advance has been overlooked. In the case of Herbert, it appears from the repertory copied that the music must have been collected over a substantial period and from music current in both England and France at different times and been virtually complete before copying into the book began. All the music pre-dates the copying time by ten years or more. The same must be true of Hirsch, not in this case copied by a single scribe, but by a number of scribes clearly directed by the primary scribe. Neither book has the appearance of having been completed, and there is no reason to suppose that the scribes did not have every intention of including new music as it came their way. So far, these are the only sources known to exhibit evidence of a pre-determined order governing the whole compilation. One is a personal anthology (Herbert), while the other occupies a grey area in professional books. There are deliberate gaps throughout the compilation of Marsh that suggest strongly a pre-determined order, rather than the scribe simple copying where the book happened to fall open, but in this case the order is one that is indistinguishable to the modern eye. The pages copied by the original (and main) scribe were carefully prepared before he copied on them,[25] and those left blank lack preparation, suggesting that the scribe certainly intended to leave these pages blank and knew precisely how much space each section of copying would require. The implication of the blank pages, however, is that the collection was not completed. Table 23 lists the contents of the manuscript with the key of the first and last chords of each piece in the first column (lower case letters indicate minor key, and upper case major).[26] Additions by the second, later, scribe are in italic type. There is no demonstrable correlation between the copying divisions in any category (single unused pages are shown by a single horizontal line, and gaps of more than one unused page are shown by a double horizontal line), though the key column, in spite of its miscellaneous appearance, may be the most likely of all the possibilities.

TABLE 23 CONTENTS AND DIVISIONS OF MARSH
Key	Page	Title/Genre	Composer
c-C	10-12/1	P.A. Pavan
c-C	12/2-13	P.A. Galliard
c-C	14-17	P.A. Variations
G-G	18	Mounsieur's Almain	[may have been added later]
C-C	25/1	Almain
C-C	25/2	Lord Hereford's Galliard
Bb-Bb	26-27	Good Night,duet treble	John Johnson
d-D	28-29	Fantasia	Alberto da Rippe
g-G	30/1	Chi Passa
G-C	30/2	Chi Passa	Cotton
C-C	35	Nusquam Galliard [incomplete]
F-F	/1	Galliard
Bb-Bb	37/1	Round
C-C	37/2-36/2	Galliard
C-C	38	Ruggiero
f-C	39	Ruggiero,duet treble
C-C	40-41	Downright Squire
C-C	42-43	Sellenger's Round/Est-ce Mars
C-C	44-45	Pepper is Black
C-C	46-48	French Galliard	John Johnson/Francesco da Milano
Bb-Bb	49/1	Part song arrangement?
Bb-Bb	49/2	Fancy	Newman
c-C	50-54/1	P.A. Pavan/Weston's Pavan	Weston
Eb?-Bb	54/2-55	Cantus Firmus setting?
c-C	56	Galliard
F-F	57	Galliard
Bb-Bb	58/1	Galliard
a-A	58/2-59	Galliard	Henry Lichfield
C-C	60	Lesson
c-c	61/1	Lusty Gallant
F-F	61/2	Queen of Scots Galliard
c-C	62-63	Galliard
D-g	64	Quel Bien Parler	Pierre Sandrin arr. Alberto da Rippe
c-C	71-72	Almain
D-G		Chi Passa
d-G	74/1	Chanson, Je Suis Desheritée	Cadéac/Lupus arr. Alberto da Rippe
d-g	74/2-75	Si Comme Espoir	Jean Maillard arr. Alberto da Rippe
g-C	76/1	Galliard
d-G	76/2	Galliard
F-D	79	E Lume Alta Galliard
Bb-Bb	80-81	Galliard
c-C	82-83	Pavan
f-F	84	Galliard
C-C	89	Galliard
C(G?)-g	90	Galliard
C(G?)-G	91	Galliard	John Johnson
Bb-F	92	Galliard
G-G	94	Fantasia	Francesco da Milano
F-F	99	Scottish Galliard
c-c	102	Galliard
C-C		Labandalashot Galliard
g-G	107	Dont Vient Cela	arr.?
d-D	115	Galliard
F-F	116-117/2	Galliard
C-C	117/1	Scottish Galliard
G-G	118/1	The New Year's Gift Galliard	Anthony Holborne
f-F	118/2	Galliard
C-C	120-121	Quadran Pavan	John Johnson
c-c	123	Almain	Richard Greene
c-C	-125	Galliard
F-F	126	Sinkapace Galliard/Church's Galliard
F-F	129	Fantasia	Francesco da Milano
f-F	130-131	P.A. Variations
d-G	132	Fancy
D-G	133	Fancy	Fernyers
D-G	134-135	Fantasia	Alberto da Rippe
A-A	136-137	In Nomine	Robert Parsons arr. H R
F-F	138	Fantasia	Francesco da Milano
C-C	139/1	Trenchmore, duet	John Johnson
C-C	139/2-141	Trenchmore, duet	John Johnson
g-d	142-144/1	P.A. Pavan, duet treble	John Johnson
G-C	144/2	First Dump, duet ground
G-C	144/3-145	First Dump, duet treble	John Johnson
Bb-C	146-148/1	Wakefield on a Green, duet treble	John Johnson
Bb-C	148/2	Wakefield on a Green, duet ground
c-c	148/3-149	French Galliard, duet treble
g-g	150-151/1	Dump, duet treble
g-g	151/2	Dump, duet ground
C-G	151/3-153	Chi Passa, duet treble	John Johnson
C-C	154/1	P.M. Variations, duet ground
C-C	154/2-156/1	P.M. Variations, duet treble
c-C	156/2-157	duet treble
g-g	158-160	Good Night, duet treble	John Johnson
D-D	162-165/2	Folia ground variations, duet treble
c-C	-165/1	Delight Pavan	John Johnson
c-c	166	Delight Galliard	John Johnson
G-G	168-169/1	Quadran Pavan
G-G	169/2-171	Quadran Pavan
F-F	173-175/1	Fantasia	Francesco da Milano
C-	175/2-176	Arthur's Dump [incomplete]	Philip van Wilder
F-F	182	Sellenger's Round, duet treble
C-G	183-186/1	The New Hunt's Up, duet treble	John Johnson
C-G	186/2	The New Hunt's Up, duet ground	John Johnson
c-C	187-186/3	Pavan
G-C	188	Weston's Pavan	Weston
G-	190	Lady Rich's Galliard [first six bars]	John Dowland [No.43]
C-C	225	Quadran Pavan
C-C	-228/1	Quadran Pavan	Clement Cotton
C-C	228/2-229	Fantasia	Francesco da Milano
Bb-Bb	230-231	Fancy	Newman
F-G	232-233	Conde Claro/Hornpipe	Guillaume Morlaye arr.?
C-Bb	234	Fancy
G-C	235	Fancy
f-F	236/1	Galliard
G-G	236/2-238	Fantasia	Francesco da Milano
G-Bb	239	Part-song arrangement?
G-C	243	Dump	E. E.
C-C	-246/1	Galliard Rondo
D-D	247-246/2	Pavan	Ambrose Lupo/John Ambrose
G-C	248-251/1	Variations
c-C	251/2	Chi Passa
c-C	257	Galliard
G-G	263	Quadran Pavan [unfinished]
G-G	264	Lavecchia Galliard
c-C	268	Packington's Galliard, first part of duet
c-c	270-271	The Old Medley	John Johnson
D-D	272	The Old Medley	John Johnson
D-G	273	Part-song arrangement?
G-G	274	In Nomine	Robert Parsons arr. H R
C-C	279	Galliard
G-C	280-282	Dump ?
c-C	287	Omnino Galliard	John Johnson
C-C	289	Quadran Galliard
c-D	295	Almain
C-C	305	Ruggiero
C-C	319	Knole's Galliard	Knowles
G-G	328-329	Lesson?
g-G		Fancy
f-F	357/1	Lord Strange's Galliard
G-G	/2	Galliard
F-F	358	Quadran Galliard
g-g	359	Militis Dump [Bergamasca Ground]
G-C	360-361	Chi Passa
Bb-Bb	362-363	Goodnight, duet treble
F-F	364	Galliard to Westminster/To Me I Must
G-G	365	Galliard	John Johnson
D-g	366-367	Chanson L'Oeil Gracieux	arr. Alberto da Rippe
F-F	368	Labandalashot Galliard
c-C	369-375/1	P.A. Pavan, duet treble	Marc Antoine
g-G	375/2	P.A. Pavan, duet ground, bass lute
c-C	376-378	Pavan
g-F	379/1	Quadran Pavan
g-g	379/2	Quadran Galliard
F-g	380/1	Chi Passa
c-d	380/2	Chi Passa
c-C	380/3	Change Thy Mind	Richard Martin
G-G	381	Lady Rich's Galliard/Dowland's Bells	John Dowland [No.43]
G-G	382/1	Mistress Norrish's Delight	?John Dowland [No.77]
d-D	382/2-383/1	The Emperor's Almain/Alliance Almain
F-F	383/2	The Duke of Parma's Almain
g-G	383/3	O Dear Life when shall it be
F-D	384	Sir John Smith's Almain	John Dowland [No.47]
G-G	385	Pavan	Mathias Mason
g-G	386/1	Galliard	Alfonso Ferrabosco/John Dowland/Francis Cutting/Robert Hales
c-C	386/2-387	Galliard	Anthony Holborne
F-F	397/1	fragment[bandora]
d-G	397/2	Goodnight [bandora]
Ab-	397/3	fragment [bandora]
F-F	397/4	The Hunt's Up [bandora]
g-g	398/1	P.A. [bandora]
c-C	398/2	Expectare Pavan [bandora]
c-C	399	Expectare Pavan [bandora]
C-C	400-401	Madrigal: Dormendo un Giorno	Philippe Verdelot arr.
g-G	419/1	P.A. Galliard
g-G	419/2	P.A.
D-C	419/3	Chi Passa
G-G	419/4	duet ground
G-G	419/5	duet ground
c-C	420-422	Galliard Variations
C-G	423-424/1	Bergamasca Variations
g-G	424/2	P.A. Variations
D-[G]	425	fragment
G-C	426/1	In Nomine	John Taverner arr. N. Strogers [?]
C-C	426/2-428	Arthur's Dump	Philip van Wilder
f-F	429/1	Psalm: Where Righteousness
F-F	429/2	Psalm: Where Righteousness

Most professional books exhibit evidence of hurry in their copying and have a tendency to a complete lack of overall organisation that is quite surprising. There is, however, a grey area occupied by a few books that have a single section of one genre or one composer within an apparently un-organised collect ion. Most often this involves separating duet music out from solo music. The original scribe of Ballet copied his duets at the back of his book, keeping only solo music at the front. Matthew Holmes was a prolific copyist, and it might be reasonable to expect him to foresee a lengthy copying life, and organise his books into sections that he would fill as appropriate music came his way. The opposite seems to be true though, with the exception of the non-solo lute music that is nearly all put into Dd.3.18. Apart from this, and with the possible exception of Nn.6.36, in which a large group of preludes is copied in a group, Holmes apparently simply copied pieces one after the other (leaving no blank folios) and at such a rate that he frequently copied pieces more than once. This may have been because he was copying faster than he could include the music in his playing repertory--collecting avidly and indiscriminately the best music he could find--but it is more likely that he copied different settings of pieces deliberately, to take advantage of a different set of divisions or transposition. The group of pieces by Holborne in Dd.2.11 are all for bandora, and even the unusually large number of pieces by Bacheler in Nn.6.36 may simply have been expedient because Holmes was working with Bacheler at the time. As noted above, scribes did not leave blank folios between their work and that of other scribes. Standard layout in any book does not allow for blank folios, and their presence should suggest a pre-determined order that may be apparent, or may only have been obvious to the compiler because he was copying from a variety of exemplars that he was conflating (or even just a single one that he was adjusting or copying in the order that he learned the music, not the order it was written in).

Concurrent scribes, therefore, will always copy within each other's work without gaps. In fact subsequent layers of scribes also follow this practice if they are working within a few years after the previous scribe's work. So far, none of the vieil ton sources have shown the original scribe progressing to transitional tunings, though if that were the case, it might be expected that the scribe would lay the manuscript out so that music in certain tunings was grouped together to facilitate its performance. Vieil ton sources that do contain music in transitional tunings have always had it added by a noticeably later layer of scribes. Transitional tuning scribes seem to prefer to invert the book and begin copying from the back, giving themselves a new 'front' to work from, rather than continuing on from where the old scribe concluded. With notable continental exceptions, the music of the older scribes is not mixed with the new-style music, even when the new scribe also makes use of vieil ton. Perhaps surprisingly, even these scribes do not deliberately layer their music into sections of each tuning.

It appears that even when a scribe might have been able to anticipate a lengthy copying span, and therefore be able to predict certain copying divisions, he did not organise his music into deliberate layers, and to have done so was fairly unusual. Many sources seem to have been deliberately miscellaneous in layout, so that a continuous playing from start to finish would give quite a varied and diverse programme. The layout of Hirsch (c1620) and its other features suggest a special purpose or intention behind its compilation, discussed further in Chapter 7. Even the most obvious of layouts, putting music in order of composer--possibly the type of layout that might be most expected--is only seen in Nn.6.36 (c1610-15), where a large number of pieces by Daniel Bacheler are grouped together at the beginning of the manuscript (with some music by other composers),and music by continental composers such as Bocquet or Perrichon only appear in the last folios, though not grouped together.The section of music by Bacheler is the largest group of works of any single composer in any of the sources. Pavans and galliards with titles and thematic material that indicate pairing are to be found separated by other music as frequently as they are found in tandem. Very occasionally up to four pieces by a single composer appear together, but this seems to be more by accident than by design, and there is no evidence that blank folios had been left to accommodate the addition of later pieces by the same composer. It seems that the music simply came into the copyist's hands in this order. Put simply, and based only on the statistically limited evidence provided by Nn.6.36, Hirsch and Herbert, pre-arranged layout of any manuscript seems to have been undesirable, and is unlikely before 1615.

SSSTEMMATICS

TABLE 24 BANDORA MUSIC IN 31392 AND Dd.2.11
31392		Dd.2.11
Scribe C		Holmes
39v-40	-	27v-28
40v-41	-	28v
41v-42	-	35
42v-43	-
43v-44	-	85v

As has been noted above, in spite of the presence of clear scribal links between many of the vieil ton lute sources, there is presently no observable stemmatic lineage between them. One exception is the possible connection between Dd..2.11 and 31392 seen most clearly in the bandora section of the music. The concordances between the two sources are shown in tables 24-25 and the duplicated order of copying for some of these pieces implies some relationship between the two sources.

The relationship of the bandora pieces is limited to a matched running order, with one piece omitted, and the bandora pieces in Dd.2.11 spread among those for the lute. Copying of lute music by the same scribe in 31392 from f.27v shows a similarly related order with Dd.2.11, as if the compiler was flicking through Dd.2.11 choosing a selection of Holmes's pieces to copy. The size of Dd.2.11 and the very popular repertory it contains would make concordances between it and 31392, copied chronologically very close to each other, to be expected, but even this rudimentary type of relationship is rare in the extant sources, and the corollary is that Scribe C in 31392 either copied from Dd.2.11 or from an exemplar common to both sources.

TABLE 25 LUTE CONCORDANCES BETWEEN 31392 AND Dd.2.11
31392		Dd.2.11
Scribe A		Holmes
14v/2-15	-	58v/1
17v-18	-	41v-42/1
18v-19v/1	-	88v-89/1
19v/2-20	-	43
Scribe B
25	-	53/2 and 66/3
25v	-	66v/2
Scribe C
26	-	100/1
26v-27	-	3
27v-28	-	46v/2-47/1
28v-29/1	-	53/1
30v-31	-	71
34/2	-	71v/2
35v-36	-	75v-77/1

In this case, the stemmatic relationship of these two sources does not have any bearing on their dating, neither of which is particularly dubious, but should a relationship of exemplar and copy be established for a source in which the dating factors are considerably more controversial, such as Hirsch, this could have significant repercussions on its chronological position. Since no recognisable exemplars survive however, unless Hirsch is a copy of an earlier one as is suggested in Chapter 7, the probability of quantities of loose sheet exemplars that did not survive is strengthened.

Diana Poulton has noted that the contents of Euing are closely related to the Holmes books,

even, occasionally, to the reproduction of identical mistakes made by Holmes in his copies. Of its [Euing] 71 pieces only three have titles or composers' names. The first part appears to be contemporaneous with all but the latest of the Cambridge books, but another hand, probably of the mid-17th century, has added an extremely interesting set of instructions for the realization of figured bass on the theorbo.

Despite the apparent corollary assumed by Poulton that the Euing scribe must have been copying from Holmes (or vice versa) the order of the music (shown in table 26) does not support the idea, and it may be that a third source (now extinct) is indicated. The similar lack of concordance with the linked 'Allison' group of sources suggests that those concordances that do exist may simply be fortuitous, as are concordances between the 'Allison' sources themselves. It is possible that the unrelated order of the music may have come about because the Euing scribe was copying from Holmes's music before it was bound, but since the Euing concordances cover the manuscripts compiled from c1585 up to 1605 (excluding the rare pieces concordant with Nn.6.36 which may simply be coincidental), it may be more likely that Holmes was copying from Euing, although the fact