This avenue for research opened up for me in 2013 when a colleague at Queensland Conservatorium Professor Andrew Brown and I made successful application for a Griffith University Infrastructure Grant that enabled the institution to purchase two 7-foot grand Yamaha Disklaviers. They allowed Andrew to build on his extensive research on interactive music systems by incorporating acoustic keyboard instruments - a substantial aesthetic improvement over digital keyboards. He had been developing software applications that would interact with improvising musicians and, with the Disklaviers now available, he approached a range of pianists at our institution to explore their potential. I was one of them even though I had little experience or reputation as an improvising musician.
I had spent much of my teenage years playing around on the piano, making up various pieces in various styles. To refer to them as compositions sounds too grand to me now. Though some were notated and the occasional one was performed in a school concert, in retrospect I was just exploring at the keyboard to find sound combinations I liked and working out by ear how harmonies worked.
The activity was more akin to private doodling than improvisation. And when I came to realise how deeply derivative my “compositions” were – most sounded like second-rate Chopin or, later, second-rate Bartok - the allure faded. (I wish I had at the time come across the ideas in Austin Kleon’s, Steal like an Artist.) It now seems ironic that when I began to study music seriously at tertiary level that side of my musicianship soon disappeared from my musical identity. At the time it seemed indulgent to be doodling at the piano when it seemed that what I obviously should be doing, when I had time at the instrument, was to practise diligently in preparation for my next lesson or performance. In retrospect that creative play was laying the foundations of my musical awareness - not to mention a grasp of keyboard choreography – and so it is now a matter of regret that I stopped.
My professional identity as a performing musician has been primarily in playing in chamber music. In part, this has been due to opportunities to play with others – I’m very fortunate that my position at Queensland Conservatorium affords me lots of those – but it has become increasingly clear to me that I find chamber music a far more satisfying form of music-making than solo performance. Quite apart from the competitive stress of being a soloist, what I have come to relish is the often spontaneous interaction with other musicians. I have increasingly come to realise that that musical Dialogue - dealing with different individuals usually in contexts where rehearsal is limited and so much is left to spontaneity in the present - is what stimulates my musical life. But I was intrigued to play with the possibilities of playing with a virtual other. I will return to the idea of Dialogue (and to the ideas of Bruce Ellis Benson) below.
So I was rather pleased that this serendipitous encounter with Andrew’s work encouraged me to improvise to again. Of course, apart from being seriously time-poor, there is no reason why I couldn’t improvise by myself but I evidently needed some outside impetus to bounce off and stimulate my creativity. Moreover, I was acutely aware of my inexperience as an improviser and so in fact having an interactive instrument with which I could practise and build my confidence with was most welcome.
The Yamaha Disklavier might best be described as a piano that has been digitally-enabled. Essentially its construction and playing mechanism is identical to other Yamaha pianos – it can be an upright or grand - but it has additional technology under the instrument that enables the movements of the hammers to be precisely measured when the instrument is played and then played back or manipulated as midi data.
For the Doppelgänger Sweet, the instrument is operating through a laptop connected through midi cables at the back of the box underneath. By bypassing the much of technology built into the instrument, it operates essentially like a digital keyboard but with the great advantage of a full range of acoustic sounds together with the keyboard’s sensitivity of touch.
As mentioned above, my colleague Andrew Brown has been researching interactive digital music systems that have the intelligence to respond creatively. For several years he has been directing a concert series called Man and Machine at Queensland Conservatorium in which I have taken part.
To quote Andrew (from a recorded conversation we had some years ago)
“For me the interesting part of human-software musical improvisation is the stimulation of human creativity with fairly minimal automated input. In some of the research we have been doing with an Australian Research Council Discovery Grant, we have been examining the notion of a reflexive approach to creativity, that is, via a system that reflects back at the performer – in a slightly distorted way– their own inputs. …
Much interactive computer music software works with extensive computational machinery, that is, with complicated algorithmic systems. [By contrast] part of what we are interested in with our research is ‘how simple can it be?’. …
And so we have been experimenting with the notion of just employing simple reflections of what musicians play. … In a way what we are doing with simple echoing strategies is assuming that musicality is already there in the human performed data, and the computer software just has to grab bits of it and throw it back out during its performance.”
DUET (FOR ONE PIANIST)
Soon after we acquired the Disklaviers, Andrew introduced me to Jean-Claude Risset’s Duet for One Pianist (1989). At one stage, we even contemplated that I would play it. Widely recognised as a landmark work, Risset’s Duet
“explored the performative possibilities made available to pianists through the augmentation of emotive human musical gesture with the precise reactive and computational capabilities afforded by computer-based musical systems.” *link
Risset’s work was also composed for performance on an earlier model Disklavier but involved a hand-written score for the pianist to perform to which the Disklavier responded in predetermined ways. As such, the work did not involve any improvisation, and there no random element built into the Disklavier’s responses. (A video performance can be easily found via a Google search and viewed on YouTube.) My intention in Doppelganger Sweet was fundamentally different, as was the outcome. It was never intended to be ground-breaking in any technical or even musical way. More Postmodern than Modernist, my aim was primarily to explore how playing with the possibilities could stimulate my creativity. In some ways I just wanted to have some fun with it and see where that might lead but I hope that the somewhat quirky musical outcomes are novel and of interest to others nonetheless.
DATA - (PURE DATA - PD)
Such reflective tools have been used by musicians for some time across various musical styles. These have ranged from artists such as Lindsay Pollack who, since the 1980s, has used a delay pedal through to the more recent work of François Pachet who works for Sony in CSL (Computer Systems Laboratory) in Paris who developed a system called The Continuator that reflects back with some variation .
So, though the type of reflective tool is hardly new or in any sense ground-breaking in itself, I was keen to explore the music-making possibilities it could offer me. Previously I had explored with patches that Andrew had devised but I wanted was a simple way to develop my own patches. I wanted a tool that was simple and flexible enough to be manipulated a musician like myself – that is, one with little experience in music technology - that would allow me the freedom to explore and play with a range of the music-making possibilities. To this end, I worked with Andrew and a research assistant Lloyd Barrett with an open-access software called Pure Data (PD). I worked closely with Lloyd for several months in 2015 and, over that time, he developed some flexible templates would easily enable me to manipulate the parameters and ultimately devise my own patches. Doppelgänger Sweet was my first public performance that played with patches I devised from Lloyd’s templates.
DURATIONS, DELAYS (AND DYNAMICS)
Screenshots of some of the patches are below. (The patches themselves can be downloaded from the 10 Mirror Sketches section here.) Their visual complexity might seem pretty confronting at first and I will just point to a few things here.
The patch for the first piece (In C) provides a clear example. In this instance the setting is such that each single note I play generates a response of 3 notes. You can see Note#1, #2 and #3 on the right hand side where the arrows are. For each of these I can easily control several parameters: the interval from the note I play and the octave register of that responding note can be determined by selecting the preferred option. The duration of that note and the delay – that is, how long after my note the response comes – can both be controlled by a sliding scale.
Screenshot of PD patch for In C (based on template developed by Lloyd Barrett)
So from the above it should be evident that for each note I play on one Disklavier, the other responds with that same pitch class one and two octaves above (that is, Notes #1 and #3). Both are short duration and share the same delay (that is, they sound together). Note #2 is a semitone below Note #3 and, because of the shorter Delay setting, it sounds as a type of grace note before the top note. Its slightly longer Duration adds some “spice” to the response. As I hope will be evident, I tried to make the nature of the response as explicit and transparent at the start of each piece by starting with a simple idea on single notes.
Listen to No.1
The patch used for No. 9 (Underwhelmed) provides another example that involves a response of 3 notes for every one that I play. Unlike the patch of No. 1 (In C) in this case all three notes share the same Duration and Delay. In other words the response for each note is a chord that consist a semitone, major 7th and minor 3rd from the original note, all of them 2 octaves below.
Listen to Underwhelmed
Screenshot of PD patch for Underwhelmed (based on template developed by Lloyd Barrett)
As was done with other parameters, we could have devised patches that involved Dynamics. These could have been set to respond in different ways – for example louder or softer than notes I play or with various degrees of randomisation - but in these I liked this aspect remaining a direct reflection of whatever shapes I played. In fact, in all of the patches used for Doppelganger Sweet, the dynamics in the response reflect back exactly what I do. Artistically, I liked the double (Doppelgänger) mirroring effect this gives. (The term Reflective practice is rarely used as literally as this!)
To show but one more similar patch here, it should be evident that the responses of No.2, involve intervals of octave, Major 3rd (in fact 10th) and Perfect 5th thereby obviously reflecting its title of Major Triads.
Listen to Major Triads
Screenshot of PD patch for Major Triads (based on template developed by Lloyd Barrett)
DETERMINISTIC RESPONSES AND DENSITY OF FEEDBACK
The responses in the patches considered above have been deterministic in so far as the Disklavier responds to what I play in a totally predictable way - the response is determined precisely by whatever I do. While that might seem potentially uninteresting, in fact, when the delays are consistent, it is directly akin to canonic writing that does require a special contrapuntal form of listening from both player and audience if the musical process at work is to be grasped. (This canonic aspect is most easily perceived in No. 9.) However, in real-time when playing/ improvising, I found I was not as conscious of this canonic aspect as I would have expected, especially in the more complex patches involving responses of several notes for each one I played. In fact the density of feedback builds very quickly in so far as one has often played multiple notes between one note and the response it determined. Given that each of those will generate further responses the density builds very quickly. As such, even though one could get an interesting-enough effect immediately by doodling some simple pattern of notes, I found it required considerable practice to anticipate the response that was coming. (In fact this sense of listening both acutely in the present while also anticipating ahead as one plays is a fundamental skill that all musicians develop.)
Setting the Delays together or in simple proportions can easily establish a regularity of rhythm and, if desired, a sense of meter. No.6 (Waltz) is the most obvious example of this. Elsewhere however I particularly enjoyed working against this regularity and undercutting the rhythmic framework with some metric modulations. The following piece No.7 (“I do, I don’t”) is perhaps the clearest case in point where I wanted both to exploit and undercut its tendency towards metric regularity and predictability. This musical aspect was reinforced by the video editing.
Some involved some subtle degrees of randomness in relation to certain parameters. For example, no 3 (Alberti) was deliberately simple to draw attention to this. The patch is below.
Screenshot of PD patch for Ostinato (based on template developed by Lloyd Barrett)
Most interesting in this regard is no 8 (Overwhelmed) with its 4ths at randomised registers and uneven delays.
Listen to Overwhelmed
Screenshot of PD patch for Overwhelmed (based on template developed by Lloyd Barrett)
The patches can work on one Disklavier – as Risset did – where the response is made on the same piano that one is playing. In fact I was mostly practising on one instrument until the week of the performance. But having the two Disklaviers set up for the performance had some distinct advantages. Firstly it avoided the unsettling inconvenience of when “we” both wanted to play the same note simultaneously. In fact improvising around the notes that are responding to what one has just played can be a bit of a minefield. (It is not as issue if playing on digital keyboards as responses on those do not involve the keys being depressed.) I mostly avoided this by setting the responses an octave or more away from the note I was playing. Also, given that the responses often instigated multiple notes for each one of that I played, I mostly preferred the patches where the response were in registers above my notes giving a sense of glittering partials above whatever harmonic language I was playing with underneath. “Real Fast” no. 5 is a clear example of that.
(“Underwhelmed” no.9 is the obvious exception where the responses are set below. Apologies are offered to anyone disturbed by the poor pun in the title.)
However the use of two Disklaviers also provides a more striking visual dimension whereby an instrument appears to be playing and responding by itself. As was evident in the videos, a view of the second instrument’s keyboard was projected up on the screen during the performance to underline the effect. Many in the audience commented to me on the unsettling nature of this imagery of an instrument that seems to be playing itself. It is clearly responding to what the pianist is playing while nonetheless seeming to have a distinct way of its own. I wanted to exploit this visual aspect and make it an important part of the aesthetic experience.
DIFFERENT TYPES OF DOUBLES
Given the suggestive visual dimension, I was keen to underline some imagery that might help both for me and audience respond to what would be an unusual musical experience. That of the Doppelgänger appealed to me on many levels. The reflective nature of the responses was of course immediately suggestive and serendipitously the concept of doubles had intrigued me for some time. Moreover it came fascinate me increasingly as I looked into it further.
For many Classical musicians the word Doppelgänger of course is inevitably evokes Schubert specifically one of his last and most revered songs with that title. Listeners familiar with that song will hear easily enough that No.8 (Premonition - Overwhelmed) is improvised around the harmonic and melodic elements of that song. From the start, I had always intended to make reference to Schubert’s song central to the Suite. (It was played with in many patches before the final one was selected.) Unsurprisingly, it turned out to be the longest and most intense piece of the Suite. But, beyond Schubert in that piece, there were many more non-musical references to doubles that stimulated my imagination during the process of preparing for this performance.
The notion of a double is a very ancient idea that can be found in many cultural traditions across the world dating back at least a far as Ancient Egypt and Greece. The nature of the relationship between a person and his/her double varies across cultures. As in Schubert’s song, a setting of Heine poem, seeing a Doppelgänger can be like an omen of one’s impending doom, almost a personification of death. In some cases, Doubles may share similar memories and feelings. (Kieslowski’s intensely poetic film The Double Life of Veronique from 1991 is a memorable example.)
Fascination with the idea of a double can be observed across much 20th century literature, film and television and is showing no signs of abating in the 21st century. The most common form of double, one that has become almost a cliché of popular culture, presents the double as sharing appearance but being opposite in character. From Dr Jekyll and Mr Hyde through to Fight Club, Black Swan or Golum in the Lord of the Rings films, the Alter Ego or Evil Twin has become a commonplace across our contemporary culture in the contexts of both drama and comedy. Memorable episodes of both The Simpsons and South Park (Butters and Professor Chaos!) have, unsurprisingly, had a very good time playing with the idea. I should add that the project preceded the recent series of David Lynch’s Twin Peaks that abounds in doubles.
The ongoing fascination with doubles in film and literature is reflected in the fact that, not one but, two highly original Hollywood feature films about Doppelgängers were released in 2013; one titled Enemy, was based on José Saramago’s 2002 novel The Double, the other called The Double was based loosely on Dostoyevsky’s novella.
A rich European literary tradition involving Doppelgängers began with Jean Paul’s novel Siebenkäs (1796) where the term was first introduced (and explained in a footnote) and continued through the 19th century most notably through poems by Shelley in English and Heine in German and stories by Dostoyevsky and Edgar Allen Poe (William Wilson, 1839).
Of all these sources I was mostly stimulated by reading Dostoyevsky’s short novella called The Double (1846/1866) where, rather than an ominous sign of imminent death, the double is a treacherous figure that, at first meekly befriends and then taunts, upstages and ultimately humiliates the hapless protagonist, Mr Golyadkin. Ultimately, as in the Poe story, it drives him to a humiliating demise of shame and disgrace. But, without that ultimate consequence, for my Doppelgänger pieces I was attracted to the image of the double that one can’t escape from and who outdoes, upstages, outshines and achieves what one can’t achieve one’s self with apparent ease. This aspect of the Double, projected through the unsettling imagery of an instrument that appears to be playing itself - perhaps more brilliantly than the man! - seemed to me to be nicely appropriate.
With the imagery in mind, what sort of music could I make with these instruments?
Actually I experimented for a time with a language of clusters and dissonance so that it became essentially a language of atonal gesture. I find it relatively straightforward to do but actually I was intrigued by how it might be possible to work in more accessible languages involving tonal harmonies, melodies and references. I was not wanting this work to address exclusively either “new music” audiences or to those concerned with cutting technology of machine intelligence.
In fact, as alluded to above, the musical language I was playing with might be described as more postmodern than modernist. It was not trying to be innovative but I wanted it to be playful and incorporate typical postmodern aspects of irony, quotation and, above all, not taking itself too seriously. I would still hope there to be some potential to offer fresh perspectives – the nature of the interaction between the instruments would be, I suspected, unsettling enough.
I gave a basic, deliberately quirky title to each of the 10 pieces, titles that mostly reflect some obvious defining musical materials of each piece. This was in part so that the audience would be in no doubt about which piece is which.
DECISIONS (BOTH SPONTANEOUS AND PRE-MEDITATED)
Though the pieces were improvised, videos of the performances show that I have a page of notation as I play and that caused some to question whether these were in fact genuine improvisations. In essence the page was a cheat sheet. I could have done without but, as my first performance of this kind, I wanted to have it as a guide to keep me on track within the structure I had devised in advance. I wanted to have some recurring thematic/motivic material to relate the pieces. (For example, the prominence of both perfect and diminished fifths is clearly audible in the opening melodic material of most pieces, and especially in the first and last that also share the key of C.) I wanted the Suite to pass through a satisfying key scheme and I had decided how the melodic material and key in which each piece would start. Moreover I decided in advance that each piece would at some point make a more-or-less obvious quotation to a well-known source. (These range from Bach to Shostakovich, from Carmen to Star Wars …)
The extent to which improvisations are be planned in advance has been discussed insightfully in Bruce Ellis Benson’s book The Improvisation of Musical Dialogue (Cambridge, 2003)
“While the very idea of a ‘planned improvisation’ sounds almost like an oxymoron, there is no reason to think that being spontaneous is incompatible with thinking about what one will play or even practising ahead of time. Clearly the question of spontaneity is more one of degree – how much a particular improvisation is planned in advance and how much happens ‘in the moment’ - than a purely qualitative difference. … jazz musicians … have often recognised that improvisation - or at least what they consider to be good improvisation - is never merely spontaneous” (pp. 141-142)
He goes on to Duke Ellington’s claim that:
“There has never been anybody who has blown even two bars worth listening to who didn’t have some idea about what he was going to play, before he started.” (p.142)
Needless to say, as an inexperienced improvisor, I was heartened by such ideas.
“As odd as it may sound, the musician who is most prepared – not only in terms of having thought about what is to be played but even having played various possibilities – is most able to be spontaneous. It is when one is already prepared that one feels free to go beyond the confines of the prepared (with the assurance that one can always fall back on them if necessary).” (Benson, 2003, pp. 142-143)
The quotations from Benson resonate strongly with me in relation to these Doppelgänger pieces but are in fact absolutely congruent with my experience as a pianist whose performing career has been spent playing Western classical music. (Benson convincingly argues that all “music making is fundamentally improvisational” (p.xii) One does often a great deal of preparation for a performance of classical music but, increasingly as I get older, I greatly value the spontaneity of musical dialogue – the continual improvisation of nuanced micro-timing and micro-shaping - that responds to the acute listening that take place during performance.
I like the concept of Dialogue not only literally between one’s self and others but also metaphorically between the Past and the Present as one revitalises in the moment what has been prepared/anticipated – whether that be in relation to a notated work or a pre-conceived set of ideas. As in all music-making, one responds in the heightened moment to performative aspects such the instrument, the context, and the audience, not to mention to one’s self and to the other musicians/instruments with whom one is playing (whether they be real or virtual). The concept of Kairos is also pertinent here – that Greek idea of time (as opposed to Chronos, the sense of measured time) that involves recognising the precise moment when an event must happen. In such respects, all my musical performances share this profound similarity.
DIFFERENT RESPONSES AND DIFFERENCES OF DEGREE
In terms of reception, I was struck by some responses I received from musicians from outside the Western Classical tradition. Several of these were along the lines of “At last you are one of us creative musicians who doesn’t just reproduce someone else’s music. Isn’t it so liberating to be express yourself at last? Welcome to the club.” I can’t say that I appreciated such comments as, to me, they showed a limited appreciation of what performers in the traditions of western Classical music are usually doing. (I do not appreciate the insinuation that we are like skilled machine operators who merely execute a composer’s instructions.)
Though there are obviously differences of degree but, on reflection, in fact what struck me more about this sojourn into public improvisation – perhaps perversely - were the commonalities with my usual practice. Certainly the music-making was playing with a different sort of creativity but, for me, the experience was actually not as profoundly different as one might expect (or as those comments assumed). All music-making involves a myriad of spontaneous micro-decisions made in that remarkably heightened, at times intoxicating, state of presence that we experience when we perform music - a presence that involves the actualisation of a future that was anticipated and prepared in the past but remains open to welcome and embrace spontaneous improvisation in the present. Actually these are the values I most aspire to when I play any music, not just improvising with my double.
Behind the Performance