Page  00000403 The Notation of Interactive Music: Limitations and Solutions Margaret Anne Schedel Peabody Conservatory fax: (410) 783-8592 Johns Hopkins University gem@peabody.jhu.edu Baltimore, MD USA 21202 www.peabody.jhu.edu/~gem Performers are eager to play interactive music, but the performance of much interactive music is difficult, if not impossible, without the presence of the composer. While some composers consider this advantageous, others struggle to find ways to make their attendance unnecessary. One solution to this problem is improved notation. Precise notation not only encourages performers to play pieces without composers present, it also promotes faster learning, and helps players diagnose performance problems. Thorough notation can also enable performers to give a more intelligent performance. The notation of interactive music will be discussed from the perspectives of composer and performer based on the results of a questionnaire. While there will always be limitations imposed by visually notating aural events, this paper will propose several solutions to the unique problems faced by composers of interactive music. Introduction Performers are often left out of the notational process; they are presented with a score and asked to play it, or they find a score and wish to perform it. Although most performers can play from any type of notation, when asked they do have specific preferences that are remarkably consistent. All agree that notation can make the difference between just playing an interactive work and an intelligent performance of the work. As shown by this survey, composers and performers have very different ideas about the notation of interactive music. Definitions For the purposes of this paper a modified version of Todd Winkler's definition of interactive music will be used. Interactive music is music composition or improvisation in which software or hardware interprets live performances to produce music generated or modified by electronics (Winkler, 1998). This definition allows the inclusion of pieces for performers with effects units or electronics other than computers. Musical notation is the written symbols by which musical ideas are represented and preserved for future performance or study (Rastall 1982). Notation is not just a means of communication between composers and performers; it also serves as a tool for theorists. It is no coincidence that most of the classic works of electronic music are notated. Notation is even helpful for composers; it serves as a reminder of original intentions and can also be an organizational or conceptual tool. Survey In 1970 the Rockefeller Foundation granted funds to find out if some form of notational standardization of twentieth century music was possible. An eight page 'Notation Questionnaire' was sent out to 1,000 composers, performers, theorists, and music editors to evoke opinions and suggestions concerning notational preferences. About 300 responses (30%) were received. The program culminated in 1974 with a three-day International Conference on New Musical Notation held in Ghent, Belgium attended by over eighty participants. Only 9 (11.25%) of the participants listed performer as their primary occupation. The notation of electronic music was not discussed because it was considered unable to be standardized. (Stone, 1980) In 1998 the author sent a questionnaire on the notation of interactive music to performers and composers of electronic music. This survey was not based on the 1970 survey, but interesting comparisons can be made between the results. 17 questionnaires (33%) were completed and returned. The other surveys were not completed for a variety of reasons. 4 contacts preferred to be interviewed rather than filling out a survey, 17 composers of electronic works composed no interactive instrumental/vocal music, I cited a language barrier, 7 did not use notation, and 5 offered no response. Of the questionnaires received, 12 (71%) were from composers and 5 (29%) were from performers. In addition to the survey, composers were asked to send two copies of one of their scores, one 'clean' and one with performer's markings, but most composers did not have a score with performer's markings. Performers were asked to send a copy of an interactive music score with their own markings. It was hoped that performers' markings would be valuable clues in clarifying the notation of interactive music, but the author was unable to draw any conclusions from the data gathered. In all 15 scores were collected showing the great variety in notational practices of interactive music. ICMC Proceedings 1999 -403 -

Page  00000404 Results The following table shows the results of the 1998 Notation of Interactive Music Questionnaire. Comments are quoted directly from responses; they were included to illustrate the.musicians' diverse opinions. This survey is not a scientific sampling; it is an informal survey of personal contacts in the field of electronic music. Question italicized text indicates performers' response Composers % Performers % Comments:Yes. No, Yes No 1 Is notating (playing) an interactive piece more difficult for you 58 25 20 80 than notating (playing) a traditional or tape work? a) No..why notate it [interaction] at all if you don't need it for i. 33% of composers performer cues. surveyed do not notate b) Yes...graphic representation of the electronics that make sense to the interaction performer and don't take too. much time to figure out. c) No... Any score I write for interactive electronics: is by definition nonspecific.; d) No...the reaction is not thesame;for each performance, there is really no purpose in notating this. e) Yes...things get more complicated when one tries to notate the electronic part as well to show the:whole: "score." f) No... but the performer must rely onr the direction of the composer/engineer without really understanding the function; this may lead to unexpected performance problems; g) Yes... There is miore etqitpmenii etup,.: and mo.re:things that can go wrong., 2. How do you approach the problem of representing the timbre of _ 100% answered the electronics? Do you think timbre should be represented in a "It depends" score? a)...graphic representation iitheii outnd.:!: b)...my main concern isridingcle taton th control" of computer parts be they synthesis or processing... 3. Do you think sound samples should be included with the score 75 25 100% answered as an aid in notation of electronic timbre? ___"It depends" a) Yes...the performers can translate: the composer's intentions to their i. 33%.of composers think a own shorthand notation symbols. sample performance b) Possibly...for rehearsal purposes itmay be useful. should be included. 4. Do you think a musician should be able to imagine or hear your 17 83 80 20 (an) interactive work just by looking at the score? a) No.. Graphics fail in conveying most timbre information. i. 25% of composers think it b) No... There would be no reason in the world that my silly signs is in spirit of their pieces would be interpretable in someone else's imagination. not to know about the electronics. c) No... Interaction will vary from performance to performance. You might try to imagine what you. will see the next time you walk out ii. 100% of performers your front door, but your imagined scenario will probably not fit your qualified positive response actual experience, with "rough idea" d) Yes. If there was a way to have two scores (violin part and full score) then one should be able to have at least an idea about the electronics. e) No, but the performer should have a reasonable idea of how the piece would work in a performance environment...notation of the electronics should be directed towards explaining the effect desired, and perhaps the method for creating sounds, rather than a series of device-specific instructions which can not be generalized to work S with alternative equipment. _ -404 - ICMC Proceedings 1999

Page  00000405 Question (continued) Composers % Performers % Comments Yes No Yes No 5. Should a performer (you) be able to perform your (an) 100 0 60 40% interactive piece without interacting with you (the composer) replied personally? _ _'la) Yes... the idea is to provide aiguide for the performing musicians to i. 25% of composers think realize the piece- the fact that a piece requires computers and personal contact preferable interactivity should not affecttthis goal. ii. 33% of composers b) Yes...this: simplyreqiesstable a user friedly interactive qualified positive response software. 'with "ideally" c) Yes, if they areinot ridc anidbeautiful. d) It depends...I am certain that astecrhnology advances and the tools for interactive music become betteroit will be easier to perform without the:composer.iBetfter notation wil-l help, butnot solve this problrri. e) * Itdepends~.. h ^ npno eforeri dio s essential to the realizatin f: a piec.. 6. What issues have you faced in notating (performing) interactive - -. computer music? a)...one hlas to be asevisuallyciieati asoonei usically creatie i. 1.7%. of composers b)..after they play piece lectroics re apparent.. mentioned limitations of commercial notation c)...the electronipt is bt out f the actal software forget "prenotation... d). ea:e.ii4 readb iin oid i iei'eti tgdey nemoi.while nnfonaion. e)- The most difficuilt paf noiating intecitive ieces hasinotbeen notatingjparts tforthe ^per ert,tii er( ntig a f:ll score:that represents all of the elements of the piece. f) The performer should knowtat much as possible about the technologly side. It really helps to interpret and to help the composer/engineer to realize the work. g) In.most of the works I have performed, the lack of notation has forced me to listen more to the sound.of the piece and get my cues aurally. This is not necessarily a bad thing.but it: depends on either a high degree of performer freedom in interpretation, or a higher degree of interaction with. the composer. Analysis By comparing composer and performer responses one can easily see that composers and performers have very different views on the subject of the notation of interactive music. The author believes composers answered the questions based on their own music while performers took a more general approach and often couldn't come up with one answer. A surprising number of composers thought a sample performance should be included (3i) After asking performers about this issue it was found that most performers do not want one recording bundled with the score. They would feel too much pressure to recreate the performance on the recording. Performers would prefer selections of the music, or multiple recordings of the work to be bundled with the score. A huge discrepancy was found in the results of question 4; 17% of composers verses 80% of performers felt musicians should be able to imagine the work based on the score alone. Even if the 25% of composers who think it is the spirit of their pieces not to know about the electronics are discarded it leaves 58% of composers who think their works cannot be represented on paper.. All performers thought they should have a rough idea of the music or how it worked. ICMC Proceedings 1999 - 405 -

Page  00000406 Another disagreement occurred in the responses to question 5. All corriposers that filled out the survey felt that their works should be able to be performed without their presence, but 25% qualified their response with ideally, and 40% of performers thought it depended on the piece. It should be noted that two of the composers who were interviewed felt their pieces could not be performed without their input. Composers and performers did agree on one point; the best performances are the result of performer-composer interaction. Solutions There is no one correct way to notate interactive music, but there are a few solutions to common problems. Composers often struggle with putting enough information on the page without crowding the page and distracting the player in performance (6d). A separate score and part would enable the performer to read off the part, and study the complete score (4d). Composers also struggle with graphically representing timbre; especially if (s)he is not artistically talented or the timbre is complex and changes based on performer input (6a). If performers were provided with a "sonic key," a recording that gave sample sounds and their corresponding notations it would be easier for them to decode the composers' scribblings (4b). Most performers would like more information about the technological side of interactive music to aid in debugging.(6f, If). Composers could reference specific functions or subpatches of their program within the score. Performers had many other suggestions. Composers should keep notation consistent and specific even at the cost of redundancy. Do not use dashes to indicate continuation of a graphic; use the graphic. If using a pedal for a performer other than a pianist don't use the %. marking. Most performers have difficulty reading the delicate script. It is better to use a graphic such as Ie. Boxes are helpful for calling attention to instructions, or separating items from the rest of the score. The best advice is to ask performers about the notation. Most feel hesitant about giving suggestions unless the score is unreadable. Limitations Composers of interactive music already have more to consider than traditional acoustic composers. They not only have to compose the piece, but also have to conceptualize the interaction, and most program their own computers. The notation of interactive music has the same limitations as notating tape music, problems notating timbres and unmetered rhythms, with added limitations imposed in describing complex interaction and timbres that change during performance. Many of the solutions offered in this paper take time composers do not feel they have; writing separate parts and scores is very time consuming. Even though part/score notation is now easier because of notation programs, these programs are not advanced graphically, and it is often easier to notate non standard music characters by hand (61). Sometimes the interactivity is too complex to notate on a two dimensional score; it is possible to compose a piece that changes so drastically depending on the performers' input that a single score could not encompass all the possibilities. Conclusion The notation of music is an imperfect art; visually representing aural events is difficult even for acoustic composers. With the added complexities of interactive music, most composers find it difficult to notate their interactive music. Performers are an important resource for composers who are struggling to clarify their notation. Clear notation helps performers learn pieces faster, diagnose performance problems and give more intelligent performances, but as this survey showed, composers and performers have very different opinions on what constitutes clear notation of interactive music. There will always be limitations in notating complex interaction, but in the future, the author would like to see interactive music represented with interactive notation; notation on a computer screen could respond to performer input. Truly interactive notation could solve many of the problems that exists in notating interactive music. Acknowledgments The author would like to thank everyone who participated in this study. References Stone, Kurt. 1980. Musical Notation in the Twventieth Century. New York: W.W. Norton & Company. Rastall, Richard. 1982. The Notation of Western Music an Introduction. New York: St. Martin's Press. Winkler, Todd. 1988. Composing Interactive Music Techniques and Ideas Using Max. Cambridge: MIT Press. -406 - ICMC Proceedings 1999