A TECHNIQUE FOR ANCIENT SOLO LYRE
Here I want to present an exploration of a set of ramifications of my research work on the Robert ap Huw manuscript. It resulted in what appears to be the recovery of the nature of the very backbone of music-making in the British Isles during the Middle Ages: the courtly harp music of the bardic tradition. Prior to the 1970s, the tunes and the harmonies of the pieces in the manuscript had both remained obscured by the difficulties of penetrating the meanings of the tablature symbols in which they were recorded, and to such an extent that it was commonly thought that the music was irretrievable. But the recovery of such a large quantity of music as the manuscript contains has filled in some of the absolutely vast void that the demise of the tradition, around the turn of the seventeenth century, left in our heritage. It gives us a bridgehead into what now appears as the extraordinarily distinctive and insular soundworld of bardic culture. It was every bit as highly developed and highly sophisticated as we should expect, from how highly prized it had been throughout the Middle Ages.
Piobaireachd and the Robert ap Huw Manuscript
I have already presented one of the most direct ramifications of the recovery: a reconstruction of the rhythms of medieval alliterative poetry in performance based on the rhythms of the harp music in the manuscript. But here I want to present some rather curious but hopefully very significant indirect ramifications. To introduce it, I need to explain some of my own personal history in the field. I initially came to the Robert ap Huw manuscript not from a baroque or renaissance perspective as others before me had, but from a piobaireachd one. I was, and fifty years on I remain, captivated by piobaireachd because of its intrinsic ‘divergent’ qualities. It is surely hard to believe that such music could be ‘diffusionist’ and have its origins in the Mediterranean lands or the Middle East. It bears the stamp of the Atlantic: the space of the long rhythm of its waves, the powerful height of its waves, the huge scope of the uncharted ocean. Surely this is music that is truly indigenous, that reflects a human response to the Insular environment. Whether or not that implies that piobaireachd’s qualities are ancient in origin does not affect the strength of the statement it makes. But of course it does set you wondering.
Throughout the modern era there has been a keen awareness that the bagpipe supplanted the harp as the main instrument of clan patronage in the Highlands and Western Isles of Scotland. If there had been medieval antecedents of piobaireachd’s ‘divergent’ and Insular qualities, then the harp music would surely have been where they existed; certainly not in ecclesiastical chant, for that was demonstrably diffused from Rome, and perhaps in the Dark Ages from Egypt and Syria. And I sense that chant carries the stamp of the Mediterranean, with, for example, that sea’s characteristically choppy wave action reflected strongly in the mobile melodies of chant, and particularly in its melismata. Francis Collinson had suggested in 1966 that some aspects of the former music of the harp – some melodic themes and its variation form – might have been taken over into piobaireachd. He pointed out: ‘This would of course offer an explanation of the always puzzling fact that piobaireachd seems to have come into existence fully fledged, without any of the rudimentary experiments which one could expect in the development of so perfected a form’. But as I sifted through the surviving records of harp music, or of what may have once have been played on the harp, in Scotland and Ireland, I was not struck by any immediate reappearance of any of piobaireachd’s most exciting ‘divergent’ qualities.
The apparent impasse began to break as I became aware of the Robert ap Huw manuscript. It was particularly Arnold Dolmetsch’s interpretations as recorded by Alan Stivell that immediately struck me as containing, unmistakably, one of the most interesting and revealing aspects common in piobaireachd – that rather stark, abstract, geometric quality (as opposed to lyrical, voice-oriented, ornate melody). That style of composition is rare in early music and I became convinced that it was not coincidence, but that in the manuscript we have material relevant both to piobaireachd and to the lost harp music common to Wales, Ireland and Scotland described by Gerald of Wales in the twelfth century. It was a revelation, and in the following year, 1972, I arrived as an undergraduate at Bangor University with the intention of studying the tablature. As I began to engage with the tablature I kept a watchful eye open for the possibility of the hallmarks of piobaireachd emerging out of it: familiar melodic motifs, the singling and doubling formula that links pairs of variations, the formulaic runs of gracenotes, the ‘cadence E’ appoggiaturas, the ùrlar-with-variations method of constructing pieces. By 1975 I had the tunes very much as I have them today, as fully-fledged, tuneful melodies, but I had found none of those piobaireachd hallmarks. Before I continue here with the consequences of that fact, I will set down briefly the salient features that I did find the harp music had in common with piobaireachd:-
- Extended form, although most harp pieces are much longer.
- In harp pieces, earlier variations tend to be melodic, particularly winding up the scale in a manner rather reminiscent of the thumb variation of piobaireachd. Later variations tend to be rhythmic in nature. Some of the ways of forming variations were formulaic and bore names.
- Heavy, if not exclusive, reliance on set pieces.
- Heavy reliance on melodic formulas, both within and between pieces.
- Melodic tonality. The harp pieces tend to be in these modes mainly (in order of frequency): Mixolydian, Dorian & Ionian, the flavours of which are relatable to piobaireachd.
- Double-tonic construction. A common one of the many patterns (or measures) used by the harp pieces, corffiniwr, is also detectable in piobaireachd.
- High degree of repetition.
- ‘Broken’ rhythms, somewhat reminiscent of the snap rhythms and a–mach of piobaireachd.
- Complexity of structure. Informed and disciplined analytical scanning is necessary in order to become familiar with a piece. Both types of music have a very strong contemplative element.
- Strong focus on minute details and differences.
- Heroic ethos. This really derives from many of the above points. I would classify both types of music as ‘divergent’ rather than ‘diffusionist’, in a way which might be characterized as heroic rather than romantic when placed into the wider European context. Formality, grandeur, majesty, are all qualities easily ascribed to both.
Nevertheless, the differences were such that it was immediately apparent that it could never be valid to extrapolate between the two idioms. I had to rule out the possibility of using the harp music as a simple means of shedding light on the nature of early piobaireachd, and vice versa. One particular area I struggled with coming to understand was the historically surprising lack of correspondence between the two in respect of particular measures. The Welsh records detail about sixty patterns or measures, which underpinned the harp music. Rather like DNA sequences, they can be used to gauge, at the deepest level, the degree of relatedness of any double-tonic material from elsewhere. They show a satisfyingly close relationship with double-tonic dance music in general, yet significantly not with piobaireachd. There are profound differences. One very revealing one I found is that where the harp music has a propensity to build melodic sequences in a pattern with two repeated segments: ABCDBD…, piobaireachd achieves the same but in the different order: BDABCD…; the ABCDBD… formula is rare in piobaireachd. Clearly, the historical relationship was far more distant than a father-son one.
It was becoming clear that the differences were such that Collinson’s suggestions were losing their appeal. It may be possible that some piobaireachd pieces contain motifs borrowed from harp tunes but, although the pieces in the manuscript are in variation form, they do not use the particular ‘ùrlar’-led variation form of piobaireachd. The recovery of the nature of the medieval harp music rendered Collinson’s suggestions as to how one might deduce that from piobaireachd redundant anyway, but of course it left unanswered the mystery of the origins of piobaireachd. In fact it deepened the mystery, by raising the question: how is it that piobaireachd developed on such different lines from the harp music? Could it be that there had been a temporal gap between them? Did the medieval harp music die out in Scotland earlier than we might suppose, and did piobaireachd develop there later than we might suppose? Was piobaireachd a conscious revival of half-remembered features of medieval music?
Alternatively, was it that the medieval harp music and piobaireachd had coexisted, but the differences can be explained by there having been an organological gap between them? That piobaireachd was developed in the Middle Ages on reed pipe precursors of the bagpipe? The discovery of the depth of the differences between the harp music and piobaireachd makes that more likely as a possibility.
Reed pipe precursors of the bagpipe
It has been the lateness of the evidence for the appearance of the bagpipe in Gaelic Scotland, along with the paucity of evidence of the development of high-status piobaireachd, that has prompted the speculation about piobaireachd’s origins on instruments other than the bagpipe. But given that medieval records concerning music in Gaelic Scotland are very scarce, and that some short piobaireachd pieces have been handed down which lack the more elaborate variations, it certainly remains possible that there was plenty of time available for piobaireachd to be developed and extended into its full high-status art form, either quite rapidly in some period between 1400 and 1570, or very slowly if the bagpipe was introduced much before 1400 or if its pre-existing music was also imported and adopted as the basis for piobaireachd. From Wales, however, where early records are relatively plentiful, comes evidence that mouth-blown pipes, precursors of the bagpipe and no doubt powered by circular breathing, had had high status.
The plain Welsh term for pipes, pibau, is generally taken, no doubt correctly, to have referred to bagpipes. I have come across only one reference to ‘bagbibau’. But an important early instance of ‘pibau’ in the Welsh triads (c. 1330) is of bag-less mouth-blown pipes:
Teir prifgerd megin ysyd, nyt amgen: organ, a phibeu, a cherd y got.
There are three principal types of wind music, namely: organ, and pipes, and bag(pipe) music.
Since ‘pibau’ here is plural, these will have been double pipes: chanter and drone, or triple pipes: chanter and two drones or two chanters and drone. And since cerdd fegin – wind music – is itself classified in another triad alongside cerdd dant and cerdd dafod – string music and tongue music – both of which had high status, there is an implication that the music of wind instruments was also, at least in part, of high status. So earlier instances of ‘pibau’ will presumably refer to mouth-blown pipes, and they bear out that pipes indeed had had high status. These are found in the Welsh Laws, which survive in c. 1250 copies but are much earlier in origin. They state that a pencerdd shall receive from the king either a telyn, a crwth or pibau, according to his usage. That the Laws should provide for a pencerdd – a master craftsman – of the pibau demonstrates that pipers and piping were organised on the same formal basis as cerdd dant and cerdd dafod were, with their penceirddiaid and accompanying administrative structures, that is, with high status. From this it is not surprising that the same three instruments are those listed in Gerald of Wales’ account c. 1185 of the instruments used in Wales, in his section of glowing praise on the distinctive instrumental music of Wales, Scotland and Ireland: cithara (harp), chorus (crwth) and tibiae (pipes). He does not notice the pipes in relation to Scotland or Ireland, but that may be merely a result of him being a native of Wales whilst at the time of writing he had only visited Ireland once and had not visited Scotland. The three instruments telyn, crwth and pibau also occur together in some of the accounts, c. 1330, of a great festival held at court at Aberteifi in 1176, at which a contest for a ‘chair’ was held between the three types of performers, with the implication that the music each type played was at the least comparable and therefore of similar status.
What was played on the pipes? I suggest it is more probable that it was closer in nature to piobaireachd than to the harp music. The piobaireachd piper Barnaby Brown has been performing early-style piobaireachd on triple pipes with two chanters and one drone, demonstrating that it is practically possible and also that it appears to be artistically credible. Essentially, if one were to develop high-status listening music rather than dance music on pipes, limited to around just nine notes, one would be hard put to do so without resorting to the sorts of gracenote runs and means of forming variations that piobaireachd uses. We can imagine, then, that early piobaireachd might well have existed alongside the medieval harp music throughout the existence of the latter, and not just for a brief period between the introduction of the bagpipe and the extinction of the medieval harp music. In Wales the pipes had lost their high status certainly by the fifteenth century if not before, so any proto-piobaireachd that was played on them had probably been abandoned there altogether. But for all we know in Gaelic Scotland the pipes and their music could have been gaining in status and popularity in the period after the time of Gerald. The politico-cultural conditions there were very different from those in Wales.
From the Welsh records it is clear that the music produced by the crwth and the timpan was essentially of the same nature as the harp music. In the High Middle Ages then, the probability is that there were two fundamental varieties of high-status instrumental music, one for the stringed instruments – known in Welsh by the collective term cerdd dant (string music) – and another for the pipes. If we project back what we know of cerdd dant and piobaireachd from later, both varieties of music in the High Middle Ages would have shared many characteristics but with significant differences. But what can we imagine of the Early Middle Ages? Certainly the crwth, and probably the frame harp, had not been developed or introduced; yet the timpan had, so that instrument may already have been the bearer of what we know later as cerdd dant. But what of that surely most ancient of Northwest European stringed instruments, the lyre?
If it were that proto-cerdd dant, the timpan music, was alone in our projected vision of the musical landscape of the Early Middle Ages, then we would have no reason to suppose that the same would not have been played on the lyre. But the identification of some sort of high-status proto-piobaireachd being also in existence, at least in the High Middle Ages, means there are two contenders for what was played on the lyre, not just one. The use of lyres also telescopes into two broad categories. The performance of poetry was accompanied by the lyre, and in Wales that continued until the fourteenth century, during which the lyre as the traditional instrument with which vocalists accompanied themselves was supplanted by the harp. The probability is that such lyres had around six strings, and the accompaniment they provided will have been similar in nature to that of the harps that replaced them, which is to say, referring to parts of the Robert ap Huw manuscript, partially if not wholly involving chordal accompaniment. But Irish records which relate to the Early Middle Ages refer not just to the three-stringed timpan but to other high-status stringed instruments designated ‘cruit’, with around nine strings. Technically, these could have been frame harps, but, with so few strings, they were almost certainly lyres. They presumably produced some form of high-status instrumental music, as did some types of lyres in the Ancient Mediterranean civilisations, and not, or not entirely, accompaniment music. Their music need not, therefore, necessarily have taken the same form as the proto-cerdd dant of the timpan.
In those first few years I spent working on the Robert ap Huw manuscript, it became apparent that there was no evidence there of counterparts to the gracenote runs of piobaireachd, not in the melodic lines nor even in broken chords in the bass. Yet the knowledge from Welsh sources that piping had been of high status in the High Middle Ages and therefore that piobaireachd, with its appearance of antiquity, might have ancient roots led me to consider if the lyre might have been used to make music of a similar nature. After all, the cruit and pipes are mentioned together in the early Irish literature, and certain types of plectrum action on a lyre can sound very like the gracenote runs and introductory cadences of piobaireachd. The need to work out the most viable stringing set-up and fingering for the lyre led me on the long, winding trail described in the following excerpt from my dissertation on various instruments in relation to the Robert ap Huw manuscript.
My own impression, from experimenting with various set-ups and fingerings on the lyre, is that my recommendations below offer the most viable solution to the problem and that they bear out my proposition that a form of proto-piobaireachd could have existed on the cruit in the Early Middle Ages. But the probability of that having actually been the case is more a matter for specialists in early piobaireachd to assess. To that end, Barnaby Brown has now also begun to explore the various fingering options that the lyre offers in relation to piobaireachd, and it is to be hoped that eventually the matter will have been pursued thoroughly enough to enable the specialists collectively to arrive at some definite conclusions as to whether or not the cruit was actually used to play a form of proto-piobaireachd. The issue is, of course, not just important because of the insights that it might offer on the nature of early bagpipe piobaireachd, but because of the insights it might offer on lyre music and lyre technique throughout the ancient world.
CHAPTER XIII: VERTICAL LYRES
[excerpt from ‘The Robert ap Huw Manuscript: An Exploration of its Possible Solutions’, Part 2: ‘Instruments’]
The cruit of nine strings
Some detail emerges from early records in Ireland of instruments with a greater number of strings than the six of the common lyre used for accompaniment. Instruments designated ‘cruit’ are referred to in a passage in the twelfth-century tale Agallamh na Seanórach in the Book of Lismore as nine-stringed. The number nine occurs again in the tale The Destruction of Dá Derga’s Hostel as the king’s individually-named nine cruit players, who are very convincingly interpreted by Frans Buisman as personifications of the nine strings of the instrument. Nine cuslennach – pipers – are also named, of which six relate to the cruit players’ names. An association of the cruit with the number nine is also found in the tale of the Dagda’s cruit: he calls it from its place on the wall and as it flies to him it kills nine men in passing. Nine strings is still well within the range used by lyres and stands far below the range recorded for harps in Europe, such as, most importantly here, the twenty-five strings the music in the Robert ap Huw manuscript requires and also the twenty or so strings of the angular harp or chang, such as is illustrated in the thirteenth-century Cantigas de Santa Maria of Spain. If, as some have thought, the nine-stringed cruit was what we know as a harp, the great opportunity presented by both triangular frame harp and angular harp designs, unlike that of the quadrangular lyre, to afford a great many strings covering a very wide compass had been passed over. Peter Crossley-Holland concluded that the cruit was indeed a lyre, in part following O’Curry and Galpin on a range of evidence, such as the name the Dagda uses to summon the cruit: Coir–cethair–cuir: ‘Quadrangular Sounding One’ in O’Curry’s translation, ‘Four Angled Music’ in Stokes’ translation, ‘Four Point Adjustment’ in others.
One difficulty here is that in Ireland the term ‘cruit’ appears to have been used sometimes as a generic term for stringed instruments, as a passage in an eighth-century Irish text demonstrates. In relation to the title of his book on the origin of the Book of Psalms, the author explains:
Nabla [is its name] in Hebrew; Psalterium in Greek ; Lauda-
torium, or Organum, in the Latin. It is asked, why it was
named by that name? Answer. From the Cruit through which
David chaunted the psalms; for, Nabla was its name in Hebrew,
Psalterium in Greek, Laudatorium, or Organum in Latin; in as
much as Organum is a generic name for all musical instruments,
because of its great nobleness. Nabla, however, is not a generic
name for every musical instrument, but Cithera is the generic
name for Cruits. Cithera, that is, Pectoralis; that is, the breast
instrument; for as much, as that it is at the breast it is played.
The Nabla is a ten-stringed Cruit; that is, which is furnished
with ten strings, which are played with ten fingers; in which
the ten commandments are concentrated. It is down upon it
[that is at top] that its belly [or sounding chamber] is placed;
and it is downwards it is played, or that music is performed on it.
From the description of the soundbox positioned above, the writer is evidently identifying David’s nabla with the angular harp, where the soundbox overhangs the strings. That such a feature was noteworthy and needed explaining suggests that his intended readership was more familiar with the soundbox positioned below, as it is in the case of lyres, or – just possibly, in the eighth century – the triangular frame harp may have already been developed.
As it happens at the limited range of nine strings, there is no playing method, and no resultant music, that either the one or the other instrument – lyre or harp – can accommodate which is excluded on the other. Although block-and-strum technique has no appeal on a harp with many strings, over just nine strings it is potentially very useful, and a plucking technique or any mixture of the two techniques are equally available no matter if it be a lyre or a harp. This is to say that in practical terms it makes no difference what the design of the frame was, as long as the strings were fully accessible to both hands. It is true that a quadrangular lyre affords much greater opportunities for re-entrant tunings than does a harp, but there is no evidence of re-entrant tunings for European lyres. Even the advantage the harp normally has over the lyre in terms of tone production, on account of having strings of different lengths for different pitches, can be broadly matched on a lyre with an oblique yoke, such as the asymmetrical lyres depicted in Ireland. Accordingly, I do not propose to pursue the issue of the morphology of the frame of the nine-stringed cruit here, but to concentrate on the musical possibilities it might offer. For convenience I will refer to it as a lyre.
To what extent is a terminus a quo for cerdd dant as we know it dependent upon the dating of the introduction or the development of the triangular frame harp? Alternatively, was there an antecedent of the many-stringed harp music of the cerdd dant in the manuscript on the early instruments: timpan and cruit? The timpan, as discussed before, was evidently capable of producing music of an adequate standard for classical professionalism, but, as a lyre-with-fingerboard, it will have developed from and, therefore, after the lyre. What is so interesting about the cruit is that it appears to have also been an instrument of classical professionalism, presumably in its most developed, nine-stringed form, so the question arises: could that instrument have produced an antecedent of the cerdd dant we are presented with in the Robert ap Huw manuscript?
Looking, firstly, into the possible capabilities of a nine-stringed cruit, the curious fact immediately emerges that a total of nine possible notes is, from the cerdd dant perspective, a very small number. Looking at the entirety of each piece in the manuscript, Caniad San Silin is the piece that uses the smallest number of strings: thirteen, followed by Caniad Cadwgan: fifteen, and Caniad y Gwyn Bibydd: sixteen. One of the main features of the music which distinguishes it so clearly from instrumental folk music is the way in which it achieves length in a piece, and this is commonly done by exploiting, in long drawn-out ways, the large compass offered by a harp of twenty-five strings.
It is not immediately apparent, then, that the nine-stringed cruit is an obvious and strong candidate as a platform for an antecedent to our many-stringed harp cerdd dant. But it is worth going into its possible capabilities in detail, for reasons that will become apparent later. The following outline of them is based on the single surviving playing tradition of the large types of lyres used by ancient civilisations: that of the bagana of Ethiopia, whilst bearing in mind the ancient lyres, particularly the kithara of Ancient Greece and Rome.
Early bagana technique
The large ten-stringed lyre, the bagana, is viewed in Ethiopia as having been brought there in remote antiquity from Israel where it had allegedly been the lyre of King David. It is, unequivocally, the Ethiopian equivalent of the various instruments in European depictions of David as musician, depicted in iconography beginning in the early fifteenth century. Although the instrument is now commonly played using the left hand only, it is apparent that the right hand used to wield a plektron, by which I mean a fairly rigid plectrum rather than a very flexible one, shaped with a rounded tip so that the strings are sounded against a curved edge, and the left hand used to be used for damping not just for plucking. It is unclear to what extent the full details of the original plektron technique have been handed down, but I have the impression that at some point, perhaps during the Italian occupation of 1936-1943, much was lost. Of most significance is the abandonment of the tuning up and full sounding of half of the strings. To the knowledge of Stéphanie Weisser, the author of the Doctorate thesis on the bagana, there was only one player in 2004-5, the master Alemu Aga, who still used the plektron, and it is from his plektron technique that its basic elements, so vital in unlocking the ancient use of large lyres, can be understood.
The technique is a combination of plucking and strumming, with subsequent damping of particular strings. The thumb and middle finger of the left hand are used to pluck a particular string assigned to each (string 1 and string 6 respectively, counting 1 as that nearest the player). The plektron is used by the right hand to either strike a single string (6, 8 or 10), a run of three adjacent strings (8 to 10 or 6 to 8) or five adjacent strings (8 to 4). In the case of a run from 8 to 4, all the strings but the last string of the run (4) are subsequently damped by the fingers of the left hand, by bringing them forward from their rest position on the strings beyond (6 to 9). The left hand is fixed in position such that each digit strikes or damps only one string (string 1 = thumb, string 5 = index finger, string 6 = middle finger: string 7 = ring finger, string 8 = little finger) whereas the plektron in the right hand is mobile, used over strings 4 to 10. Strings 2, 3, 5, 7 and 9 are the ones no longer tuned up to produce musical pitch, and of these only 5, 7 and 9 are activated, and then only in passing during a run, when they produce the friction noise of the plektron sliding over them. All the slack strings act as the physical rests for the plektron or the thumb or middle finger after their strokes on adjacent strings. The plektron strokes can be in either direction.
This is an entirely idiomatic technique. It is dependent on two important design characteristics of the bagana which distinguish it from all the other lyres still in use in the region and from the six-stringed North European lyres. Firstly, rather than being held horizontally or at an angle between horizontal and vertical so that the palm is flat and the fingers are parallel, or nearly so, to the strings, the bagana is positioned vertically or nearly so. This provides the hand with the opportunity to address the strings at an angle which is comfortable for plucking, by the left hand and by the plektron. Secondly, rather than being bunched quite closely together, the strings are more widely spaced and less splayed-out into a fan shape than on other lyres. This enables the plektron to be used selectively, on any selected single string or series of adjacent strings, and dragged across at a controlled pace, rather than swept briskly across all of them.
It is to be hoped that information on the full plektron technique when all ten strings were tuned up and sounded might yet be recovered. Michael Powne, writing in 1963, remarked that ‘a plectrum is almost always used’ and that the tuning system reported in 1798 and 1922 for all ten strings ‘appears to be the one used even today’. It differs radically from that in current use in the ordering of the notes the strings are tuned to, although both systems are re-entrant and pentatonic.
It is very significant that the technique appears to match what we understand of ancient kithara technique: the right hand uses the plektron, whilst the left performs the three functions mentioned by the first century classical writer Quintilian: it plucks (trahunt), damps (continent) and presents (praebent) i.e. leaves a string open to sound. There are two further similarities. Some of the names of the seven strings of the kithara are suggestive of a similar fixed position for the four fingers of the left hand around the middle of the series of strings. String 3 is named with the adjective lichanos – index finger; string 4 mese – middle, which may refer to its position in the middle of the strings or to the middle finger, or to both; string 5 trite – third, which may refer to its position third from the furthest string from the player or to the third finger, or to both. Secondly, string 1 is named hypate, highest or first, a term that was in general use to denote the thumb as the highest of the digits. As we have seen, string 1 on the strummed bagana is plucked by the thumb. Although it is clear here with the kithara that the index finger controlled string 3 and no other string, there is a pairing in the names of strings 1 and 2: hypate and parahypate – next to hypate, which might possibly indicate that the thumb had responsibility for string 2 as well as string 1. Similarly, the pairing of string 7 – nete – the last or lowest – and string 6 – paranete – might indicate that a finger, surely the little finger, had responsibility for those two strings. It is not secure, then, that the kithara had an entirely fixed left hand position. The thumb and the little finger may each have acted as toggles, now allowing one string to sound, then allowing its neighbour to sound. But if that was the case, then two strings would be left open to sound, and on current understanding of Ancient Greek music that seems unlikely, although many depictions of kithara playing show several strings left open, and a broad sweeping action with the plektron hand.
The number of strings on the nine-stringed cruit suggests that the combination technique of the bagana would have been used on it, the technique standing as it does half-way between the plucking technique of harp-type instruments with, usually, many more strings and the strumming technique of smaller lyres with six or so strings. There is within it certainly enough latitude and scope for complexity to accommodate sophisticated extended solo music, such as we know the kithara was used for in the Ancient World, rather than it being limited to accompaniment to the voice. In particular, the ability to drag the plectrum across several adjacent strings in a controlled way, including varying the speed during the strum, favours the development of a wide vocabulary of idiomatic, formulaic runs of notes, either to play simply a succession of plain notes or to create runs of grace notes, or mixtures of the two. A string can be approached from beyond or in front, with a single strike of an adjacent string or a whole series of them. In the case of a series, the notes do not have to be of equal duration and some or all of the strings involved can be pre-blocked, or post-damped at some chosen point in time, or left to sound. The ability to block or damp is of course contingent on a string being one which is covered by the left hand, and on the bagana that is restricted to just four strings, although there is no physical constraint that dictates the left hand should not be mobile. There can also be fairly rapid multiple strikes of a string by the plectrum moving back-and-forth on it.
If the left hand is fixed in position and there is a division but with some overlap as to which strings are plucked by the left hand and which by the plectrum, and, if the plectrum drags and left-hand damping are limited to certain strings, then the context, the phrasing, of each string would necessarily be entirely unique to it. The music produced, if it exploited these niceties, would be distinctively idiomatic to the instrument and highly idiosyncratic. The kithara and the nine-stringed cruit, limited to small numbers of notes, would have needed to exploit these same characteristics to their maximum potential in order to develop highly refined solo music, concentrating in particular on runs of grace notes rather than a wide gamut, to provide variety and sustain interest in extended performances. If indeed Buisman’s interpretation, introduced before, of the nine cruit players in the tale of Dá Derga’s Hostel as personifications of the strings and notes of the cruit is correct, their individual names supplied there can be taken as implying that the instrument produced nine different notes, but give little indication of whether the tuning was re-entrant or not. At present the etymology and musical significance of the names has yet to be investigated.
A particularly idiosyncratic feature of a lyre with fairly widely-spaced strings, when a fairly rigid plectrum is dragged across several strings, is the rippling effect of the sound produced. This is true when the strings are left open and equally so when they are damped as the last string is struck, as is the case today on the bagana. The sound is more percussive if some of the strings leading to the last string of the run are blocked before being struck. The rippling effect is very strong indeed, and attractively so, if alternate strings are blocked, since each sounding string is ‘framed’ by the ‘dud’ sound of the blocked strings that precede and follow it, so that each rippling note is surrounded by something of a rattle. In this case, unless those strings that sound are subsequently damped, on an instrument that has a tuning in a diatonic series rather than a re-entrant one, thirds will sound. So far as we can tell, that would not have been appropriate on the kithara (which was arranged in ascending serial order) in a Greek or Roman context, but of course it might well have been appropriate on the lyre in a Celtic context. The rippling effect still remains pronounced if it is not alternate strings that are blocked but some other pattern, where the rippling effect has an asymmetry to it. It is a very distinctive, elaborate sound, and one that is surprisingly different from the driving, percussive sound where blocking is used on lyres with narrower string spacing, less strings and more flexible plectrums, where all the strings are swept together in one strum. It is much more markedly rippling than a broken chord played on a harp, where there is no blocking. It can be replicated to some extent on lyres and psalteries with narrow string spacing, if the strumming is very slow and the plectrum is not too flexible, and with wider string spacing, the strumming needs to be faster. The rattling aspect is enhanced on the bagana by the presence of the buzzing mechanism – a bridge table against the top edge of which the strings are caused to vibrate intermittently, through the position of the strings above that edge being finely adjusted up or down with a U-shaped piece of leather that lies underneath each string and above the bridge table. This is the lyre equivalent of brays on a harp. The same bridge table can be made out on some depictions of the kithara.
Piobaireachd and the cruit
Now I have long believed that I can hear a very close parallel to that peculiar rippling in the idiomatic runs of piobaireachd, within which there tends to be, characteristically, an alternation of ‘muted’ notes that blend into the drone harmonics with notes that stand out sharply from them, so that the latter appear to sound as momentary pitched ‘clicks’ quite detached from one another. Further, the reed and the conical chanter of the great Highland bagpipe are such that there is a distinctive, percussive rattling or crackling aspect to the ripples, which is quite absent when the same figures are played on any other reed pipe (for example, the practice chanter), flute or whistle, where the crackling is replaced by a popping. To me, the resemblance of the crackling to some sounds produced by plektron action on a lyre is eerie. Can it be that the resemblance here is not coincidence? Might the origins of the formulaic figures of piobaireachd – its idiomatic gracenote runs and other runs – lie in the lyre?
The Scottish music scholar Francis Collinson suggested in 1966 that some aspects of the former music of the harp – some melodic themes and its variation form – might have been taken over into piobaireachd. He pointed out: ‘This would of course offer an explanation of the always puzzling fact that piobaireachd seems to have come into existence fully fledged, without any of the rudimentary experiments which one could expect in the development of so perfected a form’. His suggestion has caught the imagination of many, yet the harp music in the manuscript differs from piobaireachd in fundamental ways. Most significantly, to form variations it does not use changes of time signature or changes of grace-note figurations. Instead it is able to rely heavily on the incremental raising of notes, which piobaireachd, limited as it is to a compass of nine notes, cannot use beyond one or two variations in a piece. In contrast, one is hard put indeed to imagine how extended variation sets on a nine-stringed lyre could be achieved without resorting to the changes of time signature or of grace-note figurations used subsequently by piobaireachd. Given that the possible options for nine notes are so limited, it would not be in the least bit surprising if the same routes had been discovered twice over, independently. Nevertheless, we are confronted with such a remarkable similarity in sound that it may have been that piobaireachd evolved in the presence of fully-mature solo lyre music and is an imitation of it. Again, I draw attention to the evidence of the late survival of the lyre discussed in Chapters III and IV, and particularly that relating to the Highlands of Scotland in the sixteenth century. The concept that piobaireachd might have borrowed from the lyre rather than the harp has appeal of course, in that it is much easier to entertain the prospect of a variety of music with nine notes borrowing from that of a nine-stringed precursor rather than from that of a twenty-five- to thirty-stringed one.
The application of lyre technique to early piobaireachd material
Amongst the many types of gracenote clusters in piobaireachd, I have long been impressed by the rather extraordinary fact that there is one type in particular which conforms closely to a plectrum sweep of lyre strings where those strings are arranged in sequential scalar order, not re-entrant order. The bagana is re-entrant, but the Ancient Greek kithara was sequential and so was the lyre tuning provided by Hucbald, the only source for medieval lyre tuning in Europe. We could perhaps afford to discount the conformity as mere coincidence, were it not for the fact that this type of cluster stands out from the remainder as more mysterious and more idiosyncratic. It is also less comprehensible in terms of the bagpipe. The runs in question are the much discussed ‘introductory cadences’, by which is meant a run of notes which is entirely descending (in contrast to the ‘ripple’ effect of clusters with alternating high and low notes) and which prefaces important stressed melody notes, especially those at the beginnings and ends of phrases. They are considered to be in some sense extraneous to the melody and indeed have been considered by some to be, to some extent, at the option of the player as to whether they should be played or not, or as to where in a piece they should be played. They are mysterious in the sense that they are surely the most outstanding of the music’s many peculiarities, particularly since, unlike its other formulaic clusters, they are not particularly suggested by the nature of the bagpipe. That mysteriousness is surely reflected in the complexity of the historical record concerning their use in the tradition: from the outset of the records there have been disparities in the relative durations of the notes. The disparities seem to have multiplied through time, to also include the notes used, the circumstances that trigger the use of certain notes, and whether the notes take their time from the following plain note or the preceding plain note, or whether they draw time that is more-or-less extraneous to both. Most of the particular runs that used to be used have been dropped and have, in most instances, been replaced by the variety known today as ‘cadence E’.
The introductory cadences offer the clearest and most obvious practical point of entry into piobaireachd for the application of the combination lyre technique: that that combines plucking, blocking and damping by the left hand with the plectrum action of the right hand. This is because these cadences are so simply realisable in that way, and because they are (along with continuously ascending runs) the most obvious musical figures to produce on a sequentially-ordered lyre of nine strings using that technique. Indeed, for those reasons I suggest that the cruit was set up and handled in the following way irrespective of whether or not there was a historical link between the cruit and piobaireachd. I offer here an explanation of how I have found they can be played effectively on the lyre, by using the crucial principle provided by the old bagana technique of a fixed position for the four fingers of the left hand to cover a series of four adjacent strings. I have applied the technique to the whole range of introductory cadences, including those used by Joseph MacDonald in the late eighteenth century, examined in detail by Roderick Cannon.
Each of the cadences comprises a sequence of notes running down the scale and, because they appear not to have been played as fast throughout as typical gracenote clusters are, they can be ‘read’ as running at the speed of a relaxed sweep of the plectrum, with each note being voiced by the lyre for a carefully controlled duration. They all have in common the consecutive sounding of various notes in the limited range of high A down to D, followed by a plain note which is itself below the last short note of the run, in the range E to low G. If the left hand of the lyre player is fixed with its four fingers capable of blocking or damping the four strings: high G, F, E and D, the basic cadences can be played economically as follows: for high G-F-E-D: those four notes are swept in one stroke and damped by the left-hand fingers, and for high G-F-E and high G-F: the same actions but shorter sweeps. The other cadences would require blocking of some strings, shown here bracketed: high G-(F)-E-D, high G-(F-E)-D, high A-(high G-F)-E-D, high A-(high G-F)-E, high A-(high G)-F. Since, as Cannon penetratingly identifies, there used to be a meaningful tendency for those cadences that begin on high G to be used in G major contexts and those that begin on high A to be used in A major contexts, those beginning strings should perhaps be left open rather than damped, at least until the last note of the run is sounded. Indeed there is no means available of damping high A here anyway. But Cannon points out one instructive early setting of a piece in which each high A cadence has its high A held markedly longer than the very short high G of each high G cadence, implying perhaps that this is a relic of an early style and that only the latter should be damped on the lyre.
It remains to account for the means of sounding the plain note to which each cadence drops. In many cases this note is not adjacent to the last note of the run down, and in these cases it would be necessary to sound the plain note with the left thumb, since there is not enough time to skip the intervening strings to reach its string with the plectrum. This requires the thumb to cover strings for three notes: low G, low A and B. It might be helpful, to match the timbre of those, if the thumb was used as well to pluck C, D and E when those are the plain note, although covering a range of six strings is perhaps asking for too much mobility from the thumb. Obviously, the thumb being used for these low plain notes dictates that the lyre be strung with the lowest string nearest the player, and that gives us the direction of the plectrum stroke for these cadence runs: toward the player. There is slightly greater physical control over such an inward plectrum sweep than there is over a outward one, giving greater ability to vary the durations of each note as the plectrum passes over the strings, and that is a feature that appears to have been important in the expression of these cadences. It follows that if asymmetrical lyres were ever put to the purpose of such music, they would have needed to have the longest arm nearest to the player. That accords with the instrument depicted on the late eighth-century St. Martin’s Cross, Iona, although it is the reverse of the Irish depictions of asymmetrical instruments on the Cross of Muireadach at Monasterboice, the Durrow Cross and the Cross of Patrick and Columba at Kells, and also that on the Mal Lumkun Cross at Kirkmichael on the Isle of Man, eleventh century.
Moving out from the area of the introductory cadences, the various gracenote clusters can be addressed by ignoring those gracenotes that provide a single very quick click-like percussive articulation or cut, and by sounding the remainder. It has to be appreciated that separating out those two categories is fraught with historical difficulty, since the record has ambiguities and conflicts in the timing and in the significance of the component notes of many clusters. A plectrum sweep can be used where the notes are consecutive and where those that need to be damped can be covered by the left-hand fingers, as in the grip known in canntaireachd as embari: E-F-high G. Elsewhere, a combination of plectrum and the thumb or the middle finger, or of forward-and-back strikes by the plectrum or the thumb, makes most clusters possible. The thumb could damp where needed, for example where low G is followed by low A, or where C is followed by B. I offer suggestions for playing some of the more complex stereotypical clusters as follow.
Dithis and siubhal: plectrum plucks initial plain note, thumb plucks last note if it is low, plectrum if it is high.
Leumluath: plectrum sweeps across D (damped by forefinger) and E (damped by middle finger)
Taorluath: as above followed by thumb plucks low A (or low G where appropriate).
Crunluath: plectrum sweeps across D (damped by forefinger), E (damped by middle finger) and F (damped by ring finger), thumb plucks low A (or low G where appropriate), followed by middle finger plucks E.
Crunluath breabach: as above, followed by thumb plucks low A (or low G where appropriate), plectrum plucks final note.
Crunluath fosgailte: thumb plucks both initial plain notes, plectrum sweeps across E (damped by middle finger) and F (damped by ring finger), followed by middle finger plucks E.
As regards drone A, the left thumb is commonly unoccupied elsewhere, leaving it free for plucks of the low A string, which is conveniently placed within easy reach of the thumb. It could also add touches of low G to support the melody where appropriate, and also restore some low G to those piobaireachd clusters that include gracenotes on low G. Two strikes of low G by forward and back movement of the thumb could be used for many grips, such as in ibari. It could also damp the sounding of low G, albeit not very quickly.
Setting all these suggestions into a broad context, it is notable that my suggested fit relies on all the precise features taken from the bagana:
1) A tuning in a diatonic series with the lowest notes nearest the player, as was the case with Ancient Greek lyres but not with the re-entrant Ethiopian bagana.
2) The left hand being fixed in position, as may have been the case with the kithara and is the case with the bagana, and covering strings 5 to 8 as they currently are on the bagana.
3) Plucking by the left thumb, most commonly of the strings nearest the player, and by the middle finger of the string it covers, no. 6 sounding E here, as on the bagana. The middle finger is the strongest of the fingers.
4) Both inward and outward strikes and sweeps of the plectrum, as on the bagana, requiring free mobility of the right hand.
I think the extent to which these features enable the playing of so many meaningful and close approximations to piobaireachd’s gracenote clusters is remarkable. There are not a lot of clusters that do not lend themselves to the technique: I notice the grips dare and darodo in this regard. It is also perhaps surprising that the lyre does not simply accommodate ‘redundant A’ in the taorluath and crunluath, since that note was significant in some early styles, albeit not especially so in Joseph MacDonald’s. Perhaps it should be played, by the thumb, as part of the general working-in of drone A suggested above.
My hope is that these suggestions will prompt detailed explorations and enable some conclusions to be drawn, as to whether solo lyre playing would have had a significant impact on the shaping of piobaireachd in its early development, or whether it would have taken a form substantially unrelated to piobaireachd and perhaps more akin to our frame-harp cerdd dant. In turn this affects whether we should imagine our late medieval cerdd dant as having had a precursor played on the nine-stringed cruit, on the angular harp or on the timpan (as suggested in Chapter X), or even that it had been forged anew on the frame harp.
 Peter Greenhill, ‘The Robert ap Huw Manuscript: An Exploration of its Possible Solutions’, 1995- (dissertations deposited in the archive of the Centre for Advanced Welsh Music Studies, Bangor University), online at http://www.cl.cam.ac.uk/~rja14/musicfiles/manuscripts/aphuw/.
 Irretrievable, in the sense that the tablature was believed not to carry enough information on the precise details of the scales, rhythms, note durations, harmonies etc. of the pieces. That belief has led translators to abandon pursuing the historical solutions and instead to make ad hoc choices of their own when creating arrangements, and usually without explanation. Nevertheless, I was able to achieve substantiated answers to each and every open question, to produce, for the first time, coherent tuneful melodies which relate closely to folk tradition and which are supported by harmonies which operate on a comprehensible system. For a concise and clear exposition of the main outcomes of my research, see Paul Whittaker, ‘Interpretation’ (2006), online at http://www.pauldooley.com/aphuw_pages/intro.html.
 Peter Greenhill, ‘Bardic Rhythm: The Implications from Cerdd Dant Studies’, Studia Celtica, XLV (2011), 131-53.
 ‘Piobaireachd’ is the Gaelic term for piping or pipe music and refers to the classical music of the great Highland bagpipe traditional to Scotland. The word is commonly anglicized to ‘pibroch’.
 It has been suggested, on the basis of similarities in the modern era between Scottish Gaelic psalm-singing and Ethiopian Coptic chant, that both owe some of their origin to ancient Coptic chant in Egypt; see John Purser, Scotland’s Music (Edinburgh, 1992), 35-6.
 Francis Collinson, The Traditional and National Music of Scotland (London, 1966), 239-48.
 Collinson, 247
 These included the material alleged to have been formerly played on the harp in the Angus Fraser manuscript: Edinburgh University Library MS Gen. 614.
 The Robert ap Huw manuscript: London, British Library MS Add. 14905: written by Robert ap Huw of Anglesey, Wales, c. 1613 and containing a large body of otherwise unrecorded music in the form of a unique harp tablature.
 Arnold Dolmetsch: Translations from the Penllyn Manuscript of Ancient Harp Music (Llangefni, 1937); ‘Concerning my Recent Discoveries’, The Consort, 3 (Haslemere, 1934), 12-20.
 Alan Stivell, Renaissance of the Celtic Harp (Philips LP 6414 406, 1971), track 3.
 Giraldus Cambrensis, Topographia Hibernica (c. 1185).
 Bangor then was already something of a central point for its study, dating back to Dolmetsch’s involvement.
 Collinson, 242-8.
 A poem by Ieuan Deulwyn, fl. c. 1460: Ifor Williams (ed.), Casgliad o Waith Ieuan Deulwyn o wahanol ysgriflyfrau (Bangor, 1909), XLIII.
 Aberystwyth, National Library of Wales MS Peniarth 20, c. 1330.
 See W. S. Gwynn Williams, Welsh National Music and Dance (London, 1933), 28. Sources cited in Sally Harper, Music in Welsh Culture Before 1650: A Study of the Principal Sources (Aldershot, 2007), 40.
 The core financial basis of the piobaireachd tradition was a continuance of the cerdd dant one, in that the composer-teacher-musician held his land free of dues. For the office of pencerdd see Gwynn Williams, 18-29 and Dafydd Jenkins, ‘Pencerdd a Bardd Teulu’, Ysgrifau Beirniadol, XIV (Denbigh, 1987), 19.
 Gwynn Williams, 36-7.
 See Peter Greenhill, ‘The Forgotten Silver-Voiced Harp of Wales: The Accompaniment Lyre & the Accompaniment Harp’ (2005), 2, at http://www.pauldooley.com/aphuw_pages/silvervoice1.html
 Most notably, the two short pieces: Cainc Ruffudd ab Adda and Cainc Dafydd Broffwyd.
 Greenhill, ‘The Robert ap Huw Manuscript’, Part 2: ‘Instruments’, in progress but outlined in Greenhill, ‘The Robert ap Huw Manuscript: An Exploration of its Possible Solutions: Synopsis’ (1995), 6.
 My conclusion that the apparently common six-stringed lyre was used for accompaniment arises from the fact that accompaniment was itself common, that accompaniment requires few strings and that six was the number of strings on the lyre used for accompaniment by Bishop Gille-Pádraig of Dublin in the eleventh century: for which see Joan Rimmer, The Irish Harp (Dublin, 1969), 25-6. For its relationship to cerdd dant see Peter Greenhill, ‘The Robert ap Huw Manuscript: An Exploration of its Possible Solutions’, Part 4: ‘Technique’ (1996), 143-4.
 Ann Buckley, ‘What was the Tiompán? A Problem in Ethnohistorical Organology: Evidence in Irish Literature’, Jahrbuch für Musikalische Volks- und Völkerkunde, 9 (Cologne, 1978), 57.
 Frans Buisman, ‘The Pipers and Harpers in Dá Derga’s Hostel’: paper delivered at the Fifth conference of the Centre for Advanced Welsh Music Studies, Bangor, 1 August 1999.
 Peter Crossley-Holland, ‘Telyn Teirtu’: Myth and Magic in Medieval Wales (Bangor, 1997), 12.
 Peter Crossley-Holland, ‘Telyn Teirtu’: Myth and Magic in Medieval Wales (Bangor, 1997), 12.
 Eugene O’Curry, On the Manners and Customs of the Ancient Irish, W K Sullivan (ed.), (London, 1873), III, 214.
 Whitley Stokes, ‘Cath Maige Tured II: The Second Battle of Moytura’, Revue Celtique, 12 (Paris, 1891), 109.
 O’Curry’s translation, O’Curry, 238.
 See Joan Rimmer, The Irish Harp (Dublin, 1969), 18-19.
 Also spelt baganna, begena, beguena.
 Stéphanie Weisser, personal communication, 3.1.2008.
 Stéphanie Weisser, ‘Etude ethnomusicologique du bagana, lyre d’Ethiopie’ (Université Libre de Bruxelles, Ph.D. thesis, 2005).
 Stéphanie Weisser, personal communication, 3.1.2008. I am extremely grateful to Dr. Weisser for providing this information, so pivotal to my research.
 Michael Powne, Ethiopian Music: An Introduction (London, 1968), 57. Powne’s field work was between 1954 and 1960.
 Powne, 58. The 1798 reference is to the field work of M. Villoteau, Description de l’Égypte (Paris, 1808) and the 1922 reference is to C. Mondon-Vidailhet, ‘La Musique Éthiopienne’, in A. Lavignac & L. de la Laurencie (eds.), L’Encyclopédie de la Musique (Paris, 1922).
 Quoted in M. L. West, Ancient Greek Music (Oxford, 1992), 69.
 Most notably here the fixed left hand would tend to cause the resulting music to be laced with melodic formulas which were fixed in pitch, as they are in piobaireachd, not which change with the contours of melody as they normally do in the harp music, for which see Greenhill, ‘Melodic Formulas in the Robert ap Huw Manuscript’, 223-5.
 The harmony in the Robert ap Huw manuscript suggests that it would have been so. For in-depth analysis of that harmony built up from intervals of a third, see Paul Whittaker, ‘Harmonic Forms in the Robert ap Huw Manuscript’, Welsh Music History, 7 (Cardiff, 2007), 1-34.
 The absence of the crackling aspect on other pipes holds also for the Sardinian triple pipes. They produce a popping but not a crackling when used for piobaireachd gracenote runs. If the crackling on the great Highland pipes is to be associated historically with the lyre then it would probably be necessary for mouth-blown reed pipe precursors of the Highland bagpipes to have also ‘crackled’. I wonder, then, if Insular triple pipes were very different from Sardinian ones in respect of whatever design features are critical in this respect. I myself am not clear about what factors produce the crackling on the Highland pipes: the reed material, the conical chanter, their dimensions, the high level of air pressure on the reed, etc., but it is an important issue when it comes to reconstructing Insular mouth-blown pipes. The conical, flared Indian shenai comes close to crackling.
 Francis Collinson, The Traditional and National Music of Scotland (London, 1966), 247.
 The pieces in the manuscript do display the ‘divisions’ method of forming variations, where former single notes become divided into two or more notes, but those notes are not so brief as to qualify as ‘gracenotes’; see Greenhill, ‘Melodic Formulas in the Robert ap Huw Manuscript’, 220-221.
 These most commonly involve the introduction of the highest of the nine notes, High A, as a plain note, in ‘thumb’ variations; so-called because the note is produced by removing the thumb from the chanter.
 That is, once on the cruit and a second time on the pipes after the cruit and its music had been abandoned.
 Piobaireachd may well have evolved on the bagpipes, but it might have evolved on mouth-blown reed pipe precursors of the bagpipe long before its introduction (as explained above in the Introduction).
 The reference is to Chapters III and IV of this work, not included here.
 Hucbald, De Institutione Harmonica, c. 900, quoted and discussed in Christopher Page, ‘Anglo-Saxon Hearpen: Their Terminology, Technique, Tuning and Repertory of Verse’ (University of York, Ph.D. thesis, 1981), 188-91. It refers to a six-stringed lyre.
 That depictions of lyres in the British Isles contain quite a high proportion of examples with asymmetrical yokes supports the thesis of sequential scalar ordering in general.
 Unless an explanation is to be found in the continuous flow of sound that the bagpipe produces. I imagine that it is just possible that these cadences might have originated as substitutes, as fill-ins, for pauses for breath that occurred either whilst singing or whilst playing music on wind instruments that were not, or not entirely, powered by circular breathing.
 ‘Cadence E’ is so-called because in the modern tradition the note E in such runs is invariably played as an appoggiatura, and in stressed position.
 Roderick D. Cannon (ed.), Joseph MacDonald’s Compleat Theory of the Scots Highland Bagpipe (c.1760) (Glasgow, 1994), 14-16.
 Cannon, Joseph MacDonald’s Compleat Theory, 14-15.
 Angus MacKay’s setting of ‘MacRae’s March’; see Cannon, 15.
 Illustrated in Ann Buckley, ‘Musical Instruments in Ireland from the Ninth to the Fourteenth Centuries: A Review of the Organological Evidence’, in Gerard Gillen & Harry White (ed.), Irish Musical Studies, I (Dublin, 1990), 30, 32-33.
 It is not known whether the nine-stringed cruit would itself ever have been asymmetrical, or whether these depictions are intended to be exotic instruments with Biblical associations.
 I suspect a possibility that it was once strings 4 to 7 that were covered by the four fingers, since these had a special designation boz (‘fools’) on the bagana in the time of Mondon-Vidailhet, 1922. Mondon-Vidailhet suggested this was because they reproduced (by octave doubling) notes given by other strings, but strings 2 and 8 also duplicated a note and its octave doubling; see Powne, 58. Currently, the boz are the five strings that are not tuned up.