ILLIAC II: A Companion of Composer and Compositional Tool
Repurposing computing technology and turning it into a means for creative practice was common among the 1950s-1960s experimental musicians and electronic music instrument designers.1 The famous algorithmic composition Illiac Suite (1957) by Lejaren Hiller (1924-1994) laid the groundwork for using computers as a tool for composing. But Hiller represents only one type of early compositional use of computers. Hiller’s colleague in the Experimental Music Studios (EMS) at University of Illinois Urbana-Champaign, Salvatore Martirano (1927-1995), pioneered another approach—using the computer as a collaborator in composing music. Let’s take a look into the details of two approaches to algorithmic composition developed by Hiller and Martirano by using ILLIAC II.
The revolutionary super-computer ILLIAC II was finished in 1962, a decade after the first ILLIAC (Illinois Automated Computer) came out. The latter, employed by Hiller for his thesis project in 1956-57, is associated with the beginning of using computation both for research means and for creative practice. In this context, creativity implies the generative nature of computing operations and the possibilities of its application in music composing.
Thinking further of music technology, it can be noted that – just as the early electronic music instruments, ILLIAC II recorded information on a magnetic tape and had the pipelines’ names "Advanced Control, Delayed Control, and Interplay," reminding commands built in the early synthesizers. Notably, the envelope abbreviated as ADSR (Attack, Decay, Sustain, Release) was developed by the synth pioneer Robert Moog around the same time.
Though the advent of a transistorized and 100 times faster computer was promising, within a few years ILLIAC II lost relevance due to the development of ILLIAC III (1966), a computer that enabled pattern recognition. As soon as in 1972, the completion of ILLIAC IV, a new super-computer offering parallel computation tools, made the two previous versions obsolete among researchers – who were interested in the most advanced techniques.
Yet, in the history of algorithmic and experimental music, ILLIAC II left a significant trace thanks to composers who viewed and treated technology differently than scientists and engineers. A former chemist, Lejaren Hiller, with the encouragement of John Cage, enrolled in a master’s program in the School of Music in 1954.2 Assisted by his colleague in the Chemistry Department, Leonard Isacson, Hiller came up with an entirely new musical practice. Inspired by the information theory of Claude Shannon and Warren Weaver (1948), also known as the mathematical theory of communication, Hiller and Isaacson successfully attempted to program computer for composing music. In his thesis, Hiller explains the inter-relatedness of computing and composing:
“The present study of potentialities of a computer for the purpose of musical composition is based upon the observation that the process of musical composition can be understood as involving a series of choices of musical elements from an essentially limitless variety of musical raw materials. The act of composing can be thought of as the formation of order out of a chaotic multitude of available possibilities. Thus, it is conceivable that an analogy between the process of musical composition and the Monte Carlo technique can be established provided that the laws governing the organization of a musical composition can be codified either in the form of precise rules of in the form of probability factors or as a combination of the two.” (Hiller, 1958, p. 3)3
In 1958, Hiller formally established the first academic experimental music studio in the US, called EMS for “Experimental Music Studio.” Despite initial animosity from the conservative music department, EMS became a hub for the formation of the so-called “Urbana School” of experimental music—a competitor of Columbia-Princeton Electronic Music Center in its niche. As a director of EMS, Hiller brought many guest composers and music faculty to Urbana. One of these associates was Salvatore Martirano.
Without any background in engineering, in 1969 Martirano acquired parts of the decommissioned ILLIAC II to create an early digital analog musical instrument that he called Sal-Mar Construction. Completed in 1972 with the help of engineering students, Sal-Mar generated compositions as if it had intelligence of its own and was contributing to the compositional process. Combining the generative capacity of a machine and synchronous control of its operation by a human, Sal-Mar served as a tool for real-time, human-machine musical improvisation.
A notable piece composed by Martirano with Sal-Mar is Look At The Back Of My Head For Awhile (1970s) published in 1995 by CDCM Computer Music Series.
“The title Martirano often used for SalMar Construction performances, “Look at the Back of My Head for Awhile,” expresses his sentiment that what is to be seen by the audience is a view of the instrument and the composer-performer’s back, and what is to be heard is what happens in the concert venue between the time the power is turned on and the first sound is made and when the power is turned off—nothing more, nothing less.” (Rosenboom, 2020, p. 120)4
Another well-documented composition B.C.–A.D. II published by Tzadik in 2012 came to life as a collaboration of Martirano with his former student, experimental composer David Rosenboom. It was recorded in December of 1977, while at Martirano's studio in Urbana. In the duet, Martirano played the Sal-Mar Construction and Rosenboom produced sounds with homemade electronic modules driven by Music Easel and Buchla 200 Series. The track was captured on the “Life Field” CD of Rosenboom.
Hiller used the ILLIAC II to create another musical piece that was based on a computer algorithm. In collaboration with John Cage, Hiller completed the HPSCHD, which premiered at the Assembly Hall (State Farm Center) in Urbana in 1969, and then toured in Europe and the US. As with his first algorithmic piece, Hiller used material generated by computer to create a score, essentially working with randomness analogous to the Cageian use of I-Ching. Yet, instead of the Chinese symbol system, Hiller used numbers generated by computer (a physical embodiment of rules of formal logic that can be used to manipulate symbols) to identify the order of what Cage called “chance events” associated with tonality and melodic profile.
Because they employed computers differently in music composition, Hiller’s and Martirano’s use of musical mediums and performance capabilities are distinct. Emerging earlier than Martirano, Hiller’s approach doesn’t assume a live performance with a gigantic computing facility. Instead, it hides the compositional process from the audience. Generated with the help of a computer, the score needs to be transcribed so it can be performed with conventional musical instruments.
In Martirano’s case, Sal-Mar can be used in the performance, generating compositional material through live interaction with the composer-performer.
Despite these differences, in both cases, a human makes compositional decisions, choosing how much control to give to a computer and what algorithm the machine should follow to create a comprehensive and aesthetically pleasing musical structure.
In the context of today’s discussion of AI substituting humans as originators of music material, basically washing away creativity as a driving force of art, it’s important to historicize this discourse. Looking back, we can see that the use of AI in music-making counts for about 70 years. Over this timespan computing technology served rather as an augmentation of human creativity – being used for the development of new compositional techniques, as in the case of Hiller’s work, or for the design of “intelligent” instruments, such as Sal-Mar.
Schwartz, E. (1975). Electronic Music: A Listener's Guide. Greenwood.
Bohn, James M. An Overview of the Music of Lejaren Hiller and an Examination of His Early Works Involving Technology. DMA Dissertation. University of Illinois at Urbana-Champaign, 1997.
Hiller, L. A. (1958). On the use of a high-speed electronic digital computer for musical composition (Doctoral dissertation, University of Illinois).
Rosenboom, D. (2020). Exploring Compositional Choice in the SalMar Construction and Related Early Works by Salvatore Martirano. Between the Tracks: Musicians on Selected Electronic Music, 15.