Improvisation vs. Notation: DJ Spooky Meets the Orchestra

By: Felicia Miyakawa (Round Rock, TX) //

On March 19, 2004, I had the good fortune to attend the West Coast premiere of Devolution, a new orchestral work by composer Anthony Paul DeRitis. I was compelled to buy a ticket after reading a feature essay by Andrew Gilbert called “New Work Brings DJ into Orchestral Mix” that appeared in the San Jose Mercury News.1 The title was a play on words meant to draw attention to the renowned artist who would be featured in the concert. DJ Spooky (Paul D. Miller) was to join the Oakland East Bay Symphony in a new 32-minute concerto for DJ and orchestra, conducted by Michael Morgan. Spooky is known for his innovative “mixing,” live blending of records and digital files to produce new sounds. Gilbert’s title hints not only at the unusual (for 2004) combination of DJ and orchestra, but also to the technique Spooky was expected to use during the performance.

Flanking Devolution on the program that night were two orchestral works familiar to even the most casual of symphony aficionados: Ludwig von Beethoven’s Symphony no. 7 (1812) and Maurice Ravel’s Boléro (1928). These two pieces were selected for the program because they influenced the composition of Devolution. In fact, Devolution’s musical landscape is a collage of snippets and quotations from Boléro and the Beethoven symphony. DeRitis pulled out small passages of the two scores and recombined them with new musical connective tissue. The concert programming was intended to help the audience (re)connect with the original works, to remind listeners of (or introduce them to) the “source” material before experiencing the sounds in a new context. The performance forces for Devolution are standard for a large orchestral work: the score calls for a large complement of strings, brass, woodwinds, and percussion. But the details of how the instruments are used also stretch our expectations: the percussion section demands five players who play a huge variety of instruments, including three separate drumsets; among the string players is an electric guitarist who doubles on electric bass; and, of course, the soloist in this concerto is a virtuosic DJ, who performs on two turntables and a laptop loaded with digital mixing software (specifically, Spooky used MAX/MSP).

Spooky took the stage armed with samples from Boléro and Beethoven 7, original ideas supplied by DeRitis, and additional resources of Spooky’s choice. As is typical in orchestral works, the orchestra played from a notated score. But Spooky does not read music (nor do most DJs, who learn their complicated craft from listening and hands-on practice), so DeRitis had to find a way for the soloist and ensemble to work together. In the spirit of both hip-hop DJs and classical-era concerto performers, Spooky improvised. DeRitis gave Spooky precise timings for when to start and stop (specifying, for example, that “DJ Chunk 2” should enter at 6:12 and will last for 56 seconds), provided expressive prompts, and indicated relative volume levels. The DJ’s first entrance is marked with the following direction: “rhythmic, wide range for volume, lots of activity, beat driven.” (Readers interested in orchestral scores might be interested to know that the DJ’s part is notated in between the Electric guitar—just below the harp—and the upper strings.) Everything else is left to Spooky’s discretion, including—within certain parameters—what to play. There are no notes in Spooky’s part, no indication of pitch direction or sound quality. The DJ plays with no further direction until measure 152, when he is advised to fade out. When he enters again a few minutes later, the mood of the piece has changed. Here he is instructed to use “soft, atmospheric, individual gestures.” In one instance (measure 291), the DJ is instructed to provide a quotation from Boléro to extend over 4 measures, and in other places DeRitis indicates volume changes. Late in the score, the DJ is instructed to provide a cadenza—a solo intended to showcase his skills—of 2–4 minutes. The orchestra rests, presumably until the DJ indicates that he is ready to continue.  

Spooky is not the only improvising performer in this mix. A bit over halfway into the score, the percussionists are also asked to improvise. Their staves are clear of all rhythmic and melodic notation, making way for them to provide “free improvisations with very sparse, metallic sounds; rarely above mp, and never too busy.” The guitar also has a chance to improvise. The rest of the orchestral musicians, however, must stick to the score. (To be fair: for most of these musicians, reading from a score is likely to be more comfortable.)

There is certainly precedent for the combination of notated, pre-planned sections of music and sections of unnotated music in a single composition. Take, for example, Milton Babbitt’s Philomel (1964), an electro-acoustic piece often encountered in undergraduate music history surveys. Entire sections of this piece are for electronic sounds produced through a synthesizer. There’s no need for notation here: Babbitt input computer code, which generated sounds through the synthesizer. The only reason for notation at all is for the security of the live performer; after all, she must have a way to learn the music. (Scholars and students of music history also appreciate having something to follow.) But there is a striking difference here: Babbitt ultimately controlled all the music. The electronic sounds may not be notated, but they are closely controlled and shaped by the composer.

Soprano Tony Arnold performs Milton Babbitt’s complex piece Philomel in 2010

By contrast, the collaboration in Devolution between DJ, composer, and ensemble challenges the traditional boundaries between “composer” and “performer.” DeRitis, however, was not bothered by this ambiguity: “I wanted to give [the DJ] freedom, which raises questions of who is the final owner of the piece. The choices Spooky makes will profoundly affect the final outcome, so that Devolution is a dialectic of musical activities with fixed composition and free improvisation as its two endpoints.”2 Such freedom, however, comes at a price: there is no way to guarantee how a performance will play out. As conductor Michael Morgan quipped in advance of the concert: “I’ve been telling people for over a year now I have no idea what’s going to happen.”3

DeRitis has taken an “open source” approach to this work, especially after he oversaw a recording in 2011 by DJ Spooky and the Boston Modern Orchestral Project, conducted by Gil Rose. The composer loaded two versions of the full score to his website (one is the Oakland performance version, and the other is the version performed in New Haven, CT, earlier in 2004), as well as audio files, reviews, and program notes. Given that the DJ’s part is largely improvised and unnotated, the likelihood of a copycat performance is slim. Even if the DJ’s part were written out, anyone who attempted to “realize” the score would need all of Spooky’s original files and samples, as well as exact cues and timings, EQ and crossfader indicators, and other tools of the DJ’s trade. Every live performance will be unique, brought to life not only by a DJ’s skills and discerning ears, but also by the specifics of his record collection.

In the 18th and 19th centuries, concerto soloists would also have been expected to improvise, especially during the cadenza. Cadenzas were opportunities for the soloist to show off within the safe boundaries of convention and style. Today, however, most classical musicians look to transcriptions of what the original soloist (or another performer closely associated with the history of the particular concerto) played and try to recreate what the music might have sounded like when it was first composed. The skill of a soloist certainly influences how a concerto will sound in any given performance; a conductor shapes the sound as well. But note the difference in Devolution: the DJ provides unnotated music throughout the concerto, not just during the cadenza. As the concerto unfolds, the composer has little control over the sound of the music.

On the night of the concert, I remember being surrounded by a “typical” orchestra audience: women of a certain age with elaborate hair and perfectly chosen evening ensembles; men in suits (or at least wearing ties) and shined shoes. But there was also a phalanx of younger audience members with prominent tattoos and piercings, dressed for the club rather than the symphony. The elder patrons clapped politely between movements, clearly familiar with the conventions of a concert hall performance. The twenty-somethings pumped their fists in the air whenever Spooky hit on an especially fresh breakbeat or executed a particularly tricky transition from one sample to the next. I had a permanent grin plastered on my face, struck by the wonder of both the performance happening on stage and the intense musicking going on in the audience all around me.

For Discussion

Listen to the audio files of Devolution available here. Then watch this 2001 video of DJ Radar’s Concerto for Turntable. How would you compare how the DJs and orchestral forces interact in each?

How do the conventions of soloist vs orchestra in Devolution compare to those of the Classical era? (For comparison purposes, try Mozart’s Violin Concerto no. 3). To what extent does the solo performer (here: Hilary Hahn) influence the outcome of the performance? How much control does she have over the sound and shape of the composition? 

1 Andrew Gilbert, “New Work Brings DJ into Orchestral Mix,” San Jose Mercury News (March 18, 2004), E3.

2 Quoted in Gilbert.

3 Quoted in Gilbert. 

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s