There is a crucial twist here, which is that 3 Dreams is written for duo of baroque and modern violins, two instruments that never play together. It is crucial because we have to consider them as essentially different instruments; this consideration is central to the composition to the extent that the notes are themselves a product of the negotiations between the different physical and notational histories of the instruments.
The first and easiest way to understand what I mean by this is to consider the violins’ physical differences: one is strung with metal, the other with sheep gut; the necks are at different lengths and angles; the baroque bow is light, and speaks strongly only in one direction, whereas the modern bow is powerful and balanced, reinforced with silver. This means the violins sound differently but it also means the violinists hold their instruments differently and move differently (you can see this clearly in the picture mid-rehearsal below). Moreover, on the page the violinists read differently, too. In fact, the violins are tuned differently (A415 Hz and A440) so the same notation provokes a different attack, sustain, quality, and also frequency. Sadly, even at the very top levels of theory and practice there exists a perception of the baroque violin as a kind of primitive version of the modern violin, despite this argument being had and won over fifty years ago (indeed, we might more accurately consider the modern violin a nineteenth-century invention)!
One way I found to make this relationship clear was to focus on the open strings which, with these two violins, offered eight different frequencies and textures. The music plays on the open strings; it imagines more strings (G-D-A-E-B-F♯)… How better to demonstrate this relationship than canons at the unison which sound like nothing of the sort!
In the first instance I wrote a canon that could be manipulated in various ways—up, down, back-to-front—always familiar but often different, weaving it around three movements (the canons are performed something like twenty times in total). The first movement is technical: three doubly-serial canons—virtuosic number-crunching in musical terms; the second movement is the heart or, rather, lungs of the work; the third movement is a kind of analysis of the work as a whole; but the original canons are important above all—think of them like tendons holding the body of 3 Dreams in tension.
There are four canons, in two groups (1a-b, 2a-b). They have different characters and can be played in different ways. 1a and 2a are more open (2a can be tender), whereas 1b and 2b are more twisted (1b can be tortured). The canons can and should be combined. 1a-b-a is almost always an appropriate start, an attractive, dream-like arc, appropriate as a transition, for example; whereas, 1b-a turns away and looks upwards from something dramatic. Recordings of several combinations are included though there are more in performance.
In the first movement the performers share the same part, which examines the implications of tactus. The movement is itself in three parts, each halving the distance between canonic entries whilst maintaining the tactus and materials. The second movement grows from a kind of breathing as the performers alternate their open A-strings. The first section offers the baroque violinist space for ornamentation as the modern violin maintains the breathing figure A♭-A; the central passage conflicts two patterns of breathing, both heavy and wheezing, before reversing the relationship between the two violins, the breathing figure now A-B♭ in the baroque violin but sounding the same frequencies as its earlier versions; the modern violin is very high, the rhythms irregular as if floating above, and the accidentals reward further study: the B♭♭, for instance, if played correctly, sounds as an upper appoggiatura above the baroque violin B♭ though they sound simultaneously an interval that would usually be considered harmonious. I am particularly pleased with how this expresses and do not know another example of this expression. The third movement works towards a synthesis of the two violins. Read it as if the modern violin attempts to copy the baroque violin but is continually misled by the notational differences; there is a kind of negotiation in which the baroque violin takes charge, coaxing the modern violin towards a kind of synthesis whereby a high F♯, reached following upwards the open strings of the modern violin, sounds harmonious with a low G, the bottom string of the baroque violin, a conclusion of open string and imagined open string, gut and metal, A415 and 440.
I was once asked by a composer if I wrote everything at 440 and then transposed down for the historical instruments. I was caught slightly off-guard by the question and replied simply that there was no transposition, which perhaps came across as a sort of “hairy-chested” response but the truth of the matter was that to ask such a question is to fundamentally miss the point. To write a piece like 3 Dreams you must think in terms of the notation, move in terms of its physical realities. The A at 415 its just as much an A, physically, as the A at 440. It simply sounds lower, the same way that it sounds shorter, richer, less consistent.
About a year ago I prepared a score of 3 Dreams in which I wrote the baroque violin part at “concert pitch”, though such a thing does not exist, and notated the part with every nuance I would expect a good baroque performer to add—ornaments, articulation, dynamics, where time is taken etc. The result was a score that immediately resembled more conventional notational practices of New Music. I chose not to publish this because the likelihood of composers missing the irony of the task and actually preferring the score seemed to me a genuine possibility (I had experienced a similar response to passages of kreng (described in “tessitura a tratteggio and kreng”), whose irony was missed by some. In that state it made me think about Corelli’s op. 5 violin sonatas for whose publication beyond Italy, where the tradition of ornamentation was less familiar, included an ossia stave above the slow movements, containing an ornamented version of the solo line. Nicholas Cook has written excellently on this, concluding with the powerful phrase, ‘Schenker encouraged performers to engage with underlying structure; Corelli simply gave them no choice’ (‘At the Borders of Musical Identity: Schenker, Corelli and the Graces’, Music Analysis, Vol. 18, No. 2 (July, 1999), pp. 179-233).
I am not suggesting that my ironic edition of 3 Dreams represents a kind of foreground; this because the ironic edition deliberately undermined the fundamental structures of 3 Dreams. What I am suggesting, however, is that the level of prescription in New Music essentially undermines the fundamental structures of the instruments physically. There are all sorts of reasons of this, but a major culprit is surely that composers since the nineteenth century have became increasingly apart from performance—something that would have been unthinkable in the eighteenth century. That the music of today reflects the changing structure of the relationships between composer and performer, of notation and authority, is clear, and it is not surprising that the nature of these changes has led, by and large, to more prescriptive notational practice. While I am essentially in support of exploring different ways of making music, and different expressions for what is after all a different time, I do not believe this grasping of power by the composer is healthy for music; indeed, it seems to me a fundamental source of its elitism. One thing we could observe is that notational practice has been slightly neglected and is partly responsible for this prescription, whose music is forced to bend accustomed shapes in ways beyond their flexibility (I examined this in kreng, discussed in ‘tessitura a tratteggio and kreng’, and examine further the issue of changing musical tastes in the face of centuries of constancy in the relationship between performer and instrument in ‘The Sense of Disorder’).
3 Dreams was conceived alongside Lorenzo Bastida, a Dante scholar I met while based in Florence in 2019 who recites Dante's poetry exquisitely. The original idea was simply some combination of music and Dante, live performance and recitation. It is not as straightforward a marriage as one might think. One realises immediately the inherent music of Dante’s poetry as well as its immense strength (is there another poem that can claim to have founded a national language?). So influential is the poem that traditions have developed for its recitation. Simple adding music to the standard formula of lecture + recitation would likely detract from both so it was clear from the outset that we were looking for something different.
Searching for a structure within the great journey I became interested in the dreams Dante describes in the central of the three books, Purgatorio. There are three of them, symmetrically positioned throughout the poem at multiples of 9 (3x3)—typical medieval numerology. I was interested in the dreams as moments of transition, such as when Dante dreams he is picked up by an eagle and wakes to find himself elsewhere (canto 9). So dreaming became the theme of the project which we structured in three parts: Vita Nuova (Dante’s early writing), the three dreams of Purgatorio, and La divina commedia—as—dream. In other words, our own three dreams.
Small forces were necessary to balance the solo recitation, and violins emerged partly as a practicality (I would be able to rehearse and perform with my duo partner, Rachel Stroud). Then, the combination of historical and modern traditions set up an appropriate parallel with the mediaeval Dante and contemporary music, as well as building upon earlier experiments staging the negotiations of different histories (see description of Morphic Verse and ‘i, Byrd’ in ‘tessitura a tratteggio and kreng’).
3 Dreams is really a kind of theatrical manifestation of Dante and music. As such I will provide an outline of its performance.
The first performance was in the Round Church in Cambridge in October 2022. We performed in the round, setting up chairs around the central pillars with the slightly lower centre of the circle a clear stage. Lectures were from the circumference and recitations by heart from the centre. The canons were combined with movement usually entering the circle or departing it, always by heart, and chosen according to their dialogue with the Dante and its lecture. The first movement was performed twice, at different times, both featuring antiphonal violins (behind the audience), switching positions on repetition to provide the audience the effect of their canonic construction; the second movement was performed from the centre, invited to the stand by Lorenzo, who opened the music to the page; and the final movement was performed three times: twice antiphonally (switching) and lastly together. I do not remember the exact combination of canons but the performance lasted certainly over an hour. See below for a poster advertising the event.
We reprised 3 Dreams in the chapel at King’s College in April 2023 as part of the “Easter at King’s” festival. We managed to keep this under an hour, reducing canonic combinations which had previously always been in threes (1a-b-a, for example), sometimes to two (eg. 1a-b), focussing on different ways of performing the canons such that their gestures were clear (a forceful 1b followed by a sweet 1a is simple catharsis, whereas 2a is immediately sentimental, and offers complex combinations). We set up a kind of amphitheatre before the altar with a central stand and four more—two at the front and two at the back. We further experimented with positions amongst the audience in the front benches or occasionally standing facing Lorenzo (this was particularly powerful for the recitation of the passage from Inferno featuring Ugolino). The performance concluded, as in the round church, with the recitation of the very opening of the poem, joined by canon 1a as we all leave the stage, poet and violins branching off in different paths before fading away in improvisation, the only time words and notes sound together.
We reassembled in Florence in May 2023, touring 3 Dreams in theatres, churches, schools, palaces, and institutes between Florence, Arezzo, and Pienza. The performance became ever fine-tuned although the lectures were longer in Italian (the recitations were, of course, Italian even in their UK performances), and we adapted the staging often significantly in different spaces. In a theatre in Florence, which was perhaps the most conventional performance space, we found ourselves working against the architecture in a way that we hadn’t with some of the more eccentric churches such as the stunning pieve of San Pietro at Gropina. In the Florentine theatre we began the performance in the foyer (always 1a-b-a), like pied-pipers leading the audience firstly into the auditorium and latterly out a side door, under the theatre and up the other side to emerge onto the stage, the second of three areas we had set up for the performance. The remarkable passage about Ugolino was recited from the balcony, Lorenzo illuminated by infrared.
I hope it is clear from this that the canons are the fundamental connections between the movements and Dante. Included below is one of several “cheat sheets” we used in performance, this example in its final arrangement for the performance in the theatre in Florence. By this point the choreography was so complex we knew the project entirely by heart.
Click the icon below to view the score (this will open a PDF in a new tab).
In rehearsal in Florence (note differing violins, bows, angles, postures) (May 2023).
Pat Boyde and Lorenzo Bastida in King's Chapel on day of performance for "Easter at King's" (April 2023).
Lorenzo in rehearsal (Cambridge: April 2023).
Experimenting with lighting in Florence (May 2023).
"Cheat sheet" used in final performance in Florence (May 2023).
Poster for first performance in Cambridge (cut circular!).