Pibroch on the Harp

Bill Taylor recently sent me a recording to review of McLeod’s Lament played on wire-strung harp, closely following Colin Campbell’s 1797 notation. This is the earliest setting of Lament for the Harp Tree (PS 135). I am excited because we are nearing the end of a 12-month collaboration developing a CD and recital programme which, together with this website, form counterparts to my PhD thesis.

Building on experimental work initiated by Ann Heymann in the 1970s, Bill and I are exploring how to synthesise three strands of evidence:

  1. Oral transmission of pibroch, focusing on the performance craft of my mentor, Donald MacPherson (1922-2012).
  2. Early notations of pibroch, 1760-1841, focusing on the ‘cloud of witnesses’ for this tune.
  3. Early notations of Irish and Welsh harp music, focusing on the technical gestures presented in this PDF and this web resource.

We are grateful to the UK’s Arts and Humanities Research Council for supporting this creative collaboration through the research project Bass culture in Scottish musical traditions.

Bill and I began our experimental rehearsals on McLeod’s Lament twelve months ago and this is a preliminary result. We share an interest in what the craft of Gaelic harpers sounded like at the time when they were the pipers’ closest colleagues. Between around 1500 and 1746, when a decline in patronage for ‘classical’ Gaelic music was accelerated by civil war, pipers and harpers heard each other’s music from cradle to grave. They performed in the same halls for the same audiences. Working on the premise that something rubbed off in both directions, Bill and I have found this collaboration an illuminating one – artistically, technically and intellectually – pulling us both out of our comfort zones.

Bill made the following recording in Strathpeffer on 23 May 2015 – one take, no editing; not an end product but work in progress. He sent it to me for critical feedback and I am grateful to him for permission to publish it here, with my comments, as documentation of our journey from skeletal Campbell notation to fully-fleshed, living, breathing harp solo. A great deal of intuition and artistic judgement has been involved in this process! Almost everything could be done another way without contradicting the evidence. As Donald MacPherson often told me, “No-one has a monopoly on taste”.

That said, we have endeavoured to raise the bar by tuning in more sensitively to evidence that is culturally unfamiliar. We hope this post encourages others to investigate pibroch in a diversity of ways, treating it as evidence of a musical culture which has a depth and breadth we are barely beginning to understand.

Urlar (cycles I, IV & VII)

1:54 (bar 12). I love your unintended damp, creating a staccato note at the start of this bar. I think this could be put to good service, not here but in the counter phrase to this “rough” bar, the one we labelled “smooth”. A staccato first note would give it more individual character without reducing the contrast. So, please make the first note of  bars 7 and 11 staccato every time: not choked abruptly but fully sounded before the damp. What about trying a double choke? The point is to introduce the staccato effect taken up in cycle II. It would be better if this staccato idea didn’t arrive out of nowhere but emerged organically.

2:10. I’d value a more emotional pause before the 7th arrives, introducing the new pitch with a sense of ceremony and bated breath. It is not a mere upbeat – it is the point we’ve been waiting for. Unveil it with dramatic timing.

8:31. Please add a ‘Beat’ (the last Welsh ornament in your booklet) to the last F# in the Urlar. Why call it a beat, introducing potential ambiguity? Its fingering is midway between a short plait and a 4-finger plait, and the extent to which you use it, unbidden, suggests to me that it might be the basic ‘plait’ of medieval Welsh harp technique. It is a lovely sound, well worth adding another ‘plait’ here to increase the sense of culmination.

13:40 (bar 9) & 13:57 (bar 11). In your 3rd Urlar, I felt that these two bars could do with more space beforehand. Relish the dissonance of the C before moving on, especially in the last Urlar. I’d be happy if the three Urlars were more different.

7:10 (bar 5). I love your sruth mòr speed here. At 14:20, however, it could be slower: both broaden out the run and add a dramatic pause before it. The speed of the sruth mòr calls out for expressive nuance, the last one being the slowest, broadening out towards the end of the piece like a river delta.

8:14. Generally, your shaping is glorious, but could you phrase off here?

8:28. Lovely rubato – Yes!

8:35. The subtle difference of timing between these two successive little streams is heart-rendingly wonderful.

The last Urlar is outstanding. YES!!! Very beautiful indeed. Your first Urlar felt like it could flow a little faster – keep it moving. Not having them all at exactly the same tempo could be advantageous.

To help make them different, I suggest you develop the way you introduce bar 10, rising one scale-step higher in each Urlar (1:42, 7:48 and 13:50). This little touch would have a big impact on the overall shape. 1st time, start on E (a 3-note stream); 2nd time, start on F# (a 4-note stream, as you played in this recording); last time, start on A (a 4-note stream A-F#-E-D). They could all be emotionally charged, leaning on the first note like a cadence appuyée. I suspect that this loosening up brings us closer to the musical behaviour of a culture that persistently rejected musical notation.

Cycle II

Allow the staccato notes to blossom before you choke them, a full bell-like sound. What about a sprinkling of double chokes, possibly at the start of some or all phrases? There could be more than two staccato notes in the Urlar to help knit the cycles together. I’d like to hear the defining feature of cycle II growing out of cycle I naturally – something that was an attractive detail in the Urlar is now taken up as the main motif.

3:39. Expand the time, don’t cram everything in – this merits space. Story-teller timing, not dance musician – we’re in a completely different register and craft here, that of epic storytelling.

4:04. Again, I think a less measured feel would help at this point in the piece. Save such urgency for the Doubling – I feel you reach a ‘Doubling’ level of excitement and clarity of measure too soon. Keep that card up your sleeve. There is great craftsmanship in using timing to shape the large-scale architecture! Set the tension level in your rubato here midway between cycle I (very relaxed) and cycle III (quite taught) – we want equal steps, or a smooth climb. The goal is to captivate listeners with the gripping anticipation of an almighty crescendo.

4:25 (and 8:18). I would enjoy a slightly slower sruth mòr here – and a more open bee’s plait: less crushed, full-bodied, all internal notes palpable for sonic beauty. The sruth mòr at 11:58 is glorious: just right for that point in the piece.

Cycle III

Beautiful tempo, gorgeous timing, just right!!!

Ends of phrases could have even more time: in terms of the rubato tension level, keep a card up your sleeve for later. The crescendo (of rhythmic intensity) does not reach its culmination here, on the first journey, but on the final one, your Barrludh Fosgailte Doubling.

Before the Urlar returns, add more time. Take a deep breath, both here and (an even deeper one) after the Barrludh Fosgailte Doubling.

5:53 (bar 62). Could you delay the bass note to the same off-beat, keeping the pattern going? Instead of coinciding with the melodic C, strike the bass after it, coinciding with the run of the sruth mòr. See this revised score (page 3 only). Bar numbers have disappeared, sorry! Sibelius 7 has a bug and I have not found a work-around. Cycle III begins at bar 40.

In this revised score, I have added your bass notes and proposed three small changes to the ornamentation:

  1. You are currently doing a forked choke on both F# and E, which produces A and G grace notes respectively. As G is not in this melody, could you do a thumb choke on E, producing an F# grace instead of a G? This is in bars 45, 50, 53, 63.
  2. How about introducing the casluth (from Bunting) in bars 51, 58 and 65? This would develop the staccato characterisation of the previous cycle. I’d really like the two bars opening the 3rd and 4th Quarters to be recognisable as a contrasting pair, returning but in reverse order. To achieve that, they need to stand out both from each other and from the rest of the cycle. Lively sound bites, grabbing our attention! To help achieve that, I have suggested a 2-note sruth immediately after the casluth in bars 51 and 58. Would that work?
  3. I have added another bee’s plait at the start of bar 65, weighing down that bar with ornamentation to increase the sense of culmination and disrupt the pattern irretrievably.

Barrludh Fosgailte Singling & Doubling (cycles V-VI)

In the revised score, I have notated a rhythm in 5 that would, I think, help to lift the internal timing of each beat by avoiding the squareness of two equal subdivisions. In the recording, the ‘ghost’ note at the end of the barrludh fosgailte divides the beat equally in half and the triplet doesn’t have a solid place in the rhythmic framework – instead it is a floating prefix, the result of which is that it sometimes gets crushed.

To avoid these two issues – the squareness and the crushing – I’d recommend raising the status of the triplet, giving it independent metrical value as follows: subdivide each beat into five (1-2-34-5), start the triplet on 3 and place the ghost on 4. Perhaps use the 5 fingers of your bass hand as a metronome, tapping the soundbox, to practise fitting the barrludh fosgailte into this rhythmic framework, playing on 3 and damping on 4. The melody notes will come on 1 and 5.

Would you be able to slow down the little streams, opening them out and relaxing into the phrase endings? In the Singling, we want that ‘Urlar’ feeling to stay with us, intermittently. As well as playing the streams slower, you could lean on the first note if that took your fancy.

10:35. You could make more of this augmented 4th, adding time.

11:58. Glorious sruth mòr!

I love your timing from note to note in the Doubling. This is nuanced and expressive – playing of the highest order. Donald MacPherson would be delighted! He had no time for wooden, mechanical playing. Just listen to his Doublings on A Living Legend – even his Crunnludh a Mach bursts with musicality as well as being technically flawless. Every grace note is palpable, with artistic intelligence in the internal timing of his finger movements, opening out or closing up.

This is an awe-inspiring result, Bill – the most satisfying outcome of my research to date. Thank you!

A few questions

Bar 49. Should I resolve the final melody note with a short plait (C-D)?  It seems so bare without something!  It’s what we do in cycle II, bar 24. This is also a question for the same point in the Urlar, bar 5.

If you change it in the Urlar, then you should change it every time (bars 2, 5, 7, 11, and twice in bar 13). The same principle applies in cycles II and III – the ending of bar 19 is carried through cycle II, and the ending of bar 43 is carried through cycle III. We have signalled the end of the first half of each cycle with a heavier sruth on the penultimate note of the 2nd time bar. In the Urlar, we have a double sruth beag instead of a single sruth beag; in cycle III, a sruth beag instead of a forked choke. We could do something similar in cycle II – perhaps a rapid stream in bar 19 and a heavy stream (leaning on the first note) in bar 23 – but that’s the only change I’d make.

Fixing such small expressive details in notation is perhaps taking us away from what we’re after. Each rendition could be different and I’m sure would have been across the lifetime of a player like Ruaidhri Dall Mac Mhuirich (c. 1656 – 1714).

Bar 63, lower hand. Instead of playing A and then A, how about A then G, as we do in bar 14 of the Urlar?  I know G isn’t in the melody, but this is a very expressive moment.

Hmm. That would weaken it’s impact in the Urlar. We already hear the G four times as we also have it in cycle II (at 4:31). Repeating it in cycle 3 would mean hearing it five times (at 2:18, 4:31, 5:57, 8:23 and 14:26). I’d prefer to anticipate its removal in the Barrludh Fosgailte cycles and, in fact, I recommend we go in the opposite direction. Make it unique to the Urlar: remove the G from cycle 2 (bar 37) so we hear it three times only (at 2:18, 8:23 and 14:26). I believe that that would produce a more powerful experience, particularly on repeat listening. We are culturally habituated to a fast fix and many pitches; here, reward lies in restraint.

Bar 69, little stream to melody D. How about AFED rather than GFED? This would be better for damping, just stopping the E and allowing A, F and D to continue ringing.  Should we be consistent with this?  It happens quite a bit, as in the last bar of cycle II, bar 47, etc.

Yes, please change them all – the fewer Gs the better. The frisson of its appearance at the end of the Urlar is heightened by its absence  elsewhere.


6 thoughts on “Pibroch on the Harp

  1. Beautiful playing and interesting comments. Tks for sharing.

    Is there a link somewhere to the complete sheet music of this tune as we can hear it here (not necessarily with all the ornaments but with the basic melodic outline)? Or maybe that will be available when the cd comes out? Would be interesting to follow the comments and see where they apply in the specific bars mentionned. Tks.

    1. It’s in the Campbell Canntaireachd, which is online, on the Piobaireachd Society’s website, and elsewhere as well, under its name ‘MacLeod’s Lament’. Campbell recorded a different tune with the name ‘Cumha Teud nan Craobh’ (Kinlochmoidart’s Lament). A piece well worth learning.

  2. Only just had a chance to listen to this. Very interesting interpretation, nice selection of gestures for the different movements.

Leave a Reply