Today I learned that the composer David Cope died on May 4, at the age of 83.
Cope was a long-time professor of music at UC Santa Cruz, and is known for his pioneering work in computer generated music, especially with his “Experiments in Musical Intelligence” (EMI or “Emmy”) project, carried out in the 1980s and 1990s. EMI was a program that could generate pieces in the style of various composers. Here’s an example of a fugue in the style of J. S. Bach, generated by EMI:
(You can listen to many more EMI pieces on Cope’s YouTube channel.)
I first learned about Cope and EMI from my former PhD advisor Doug Hofstadter, and I wrote about this project (and how much Doug was disturbed by it) in my 2019 book, Artificial Intelligence: A Guide for Thinking Humans. Here are some excerpts on Cope’s work (and Doug’s reactions to it) from that book:
Excerpt from the Prologue:
In the mid-1990s, Hofstadter’s confidence in his assessment of AI was again shaken, this time quite profoundly, when he encountered a program written by a musician, David Cope. The program was called Experiments in Musical Intelligence, or EMI (pronounced “Emmy”). Cope, a composer and music professor, had originally developed EMI to aid him in his own composing process by automatically creating pieces in Cope’s specific style. However, EMI became famous for creating pieces in the style of classical composers such as Bach and Chopin. EMI composes by following a large set of rules, developed by Cope, that are meant to capture a general syntax of composition. These rules are applied to copious examples from a particular composer’s opus in order to produce a new piece “in the style” of that composer.
…Hofstadter spoke with extraordinary emotion about his encounters with EMI:
I sat down at my piano and I played one of EMI’s mazurkas “in the style of Chopin.” It didn’t sound exactly like Chopin, but it sounded enough like Chopin, and like coherent music, that I just felt deeply troubled.
Ever since I was a child, music has thrilled me and moved me to the very core. And every piece that I love feels like it’s a direct message from the emotional heart of the human being who composed it. It feels like it is giving me access to their innermost soul. And it feels like there is nothing more human in the world than that expression of music. Nothing. The idea that pattern manipulation of the most superficial sort can yield things that sound as if they are coming from a human being’s heart is very, very troubling. I was just completely thrown by this.
Hofstadter then recounted a lecture he gave at the prestigious Eastman School of Music, in Rochester, New York. After describing EMI, Hofstadter had asked the Eastman audience—including several music theory and composition faculty—to guess which of two pieces a pianist played for them was a (little-known) mazurka by Chopin and which had been composed by EMI. As one audience member described later, “The first mazurka had grace and charm, but not ‘true-Chopin’ degrees of invention and large-scale fluidity . . . The second was clearly the genuine Chopin, with a lyrical melody; large-scale, graceful chromatic modulations; and a natural, balanced form.” Many of the faculty agreed and, to Hofstadter’s shock, voted EMI for the first piece and “real-Chopin” for the second piece. The correct answers were the reverse.
…[Hofstadter] went on:
I was terrified by EMI. Terrified. I hated it, and was extremely threatened by it. It was threatening to destroy what I most cherished about humanity. I think EMI was the most quintessential example of the fears that I have about artificial intelligence.
Excerpt from Chapter 16, discussing “creativity” in AI systems:
There have been similar examples with music generation, in which a computer is able to generate beautiful (or at least pleasing) music, but in my view the creativity comes about only through collaboration with a human who lends the ability to understand what makes music good and thus provides judgment on the computer’s output.
The most famous computer program that generated music in this way was the Experiments in Musical Intelligence (EMI) program, which I mentioned in the prologue. EMI was designed to generate music in the style of various classical composers, and some of its pieces managed to fool even professional musicians into believing they had been written by the actual composer.
EMI was created by the composer David Cope, originally to serve as a kind of personal “composer’s assistant.” Cope had been intrigued by the long tradition of employing randomness to generate music. A famous example is the so-called musical dice game, played by Mozart and other eighteenth-century composers, in which a composer cut up a piece of music into small segments (for example, individual measures) and then rolled dice to choose where the segments were placed in the new piece.
EMI, it could be said, was a musical dice game on steroids. To get EMI to create pieces in the style of, say, Mozart, Cope first selected from Mozart’s works a large collection of short musical segments and applied a computer program he had written that identified key musical patterns he called “signatures”—patterns that help define the composer’s unique style. Cope wrote another program that classified each signature as to the particular musical roles it could play in a piece. These signatures were stored in a database corresponding to the composer (Mozart, in our example). Cope also developed in EMI a set of rules—a kind of musical “grammar”—that captured constraints for how variations of signatures could be recombined to create a coherent piece of music in a particular style. EMI employed a random-number generator (the computer equivalent of throwing dice) to select signatures and create musical segments from them; the program then used its musical grammar to help decide how to order the segments.
In this way, EMI could generate a limitless number of new compositions “in the style” of Mozart or any other composer for whom a database of musical signatures had been constructed. Cope carefully chose the best of EMI’s compositions to release publicly. I’ve listened to several of them; to my ear, they range from mediocre to amazingly good, with some beautiful passages, though none have the depth of the original composer’s work. (Of course, I say this knowing ahead of time that the pieces are by EMI, so I might be prejudiced.) The longer pieces often contain lovely passages, but also have a non-humanlike tendency to lose the thread of a musical idea. But overall, the published works of EMI were very successful in capturing the style of several different classical composers.
In 2005, in a decision that I find bewildering, Cope destroyed EMI’s entire database of musical signatures. The main reason he gave was that EMI’s compositions, being so easily and infinitely producible, were devalued by critics. Cope felt that EMI would be valued as a composer only if it had, as the philosopher Margaret Boden wrote, a “finite oeuvre—as all human composers, beset by mortality, do.”
I don’t know if my opinion will offer any consolation to Douglas Hofstadter, who was so upset by EMI’s most impressive compositions and their ability to fool professional musicians. I understand Hofstadter’s worry. As the literary scholar Jonathan Gottschall has observed, “Art is arguably what most distinguishes humans from the rest of creation. It’s the thing that makes us proudest of ourselves.” But I would add that what makes us proud is not only the generation of art but also our ability to appreciate it, to understand what makes it moving, and to comprehend what it communicates. This appreciation and understanding are essential for both the audience and the artist; without this, I can’t call a creation “creative.”
All this aside, Cope’s work on EMI and related systems was truly original, visionary, and enormously impactful. May he rest in peace and may his memory be a blessing.
I would like to offer a few comforts to anyone inclined to be terrified by programs like EMI.
First is the simple and obvious point that EMI could not have composed like Chopin if Chopin had not existed. Chopin, on the other hand, wasn't copying anyone.
Second, Hofstadter's experiment contains a crucial flaw. He had to use a very obscure Chopin piece, lest the professional musicologists in the audience recognize it. But surely the piece deserved its obscurity! So the comparison was between EMI's mashup of Chopin's best work — probably not even including lesser pieces in its database — and a throwaway that Chopin probably tossed off one day when he had a hangover. Not fair at all.
And thirdly, reading your description of how EMI worked, it's clear that this is not AI composing music from scratch on its own; this is David Cope writing a program incorporating a considerable amount of his own knowledge about composition, along with stylistic elements from great composers, to generate music. In short, the agency and the genius, if it was genius, lay within Cope, not the machine. If computers didn't exist, Cope could in principle have done the same thing with pencil and paper and dice; would that have meant the dice were composing music?
Perhaps these observations will help people ... uh ... cope :-)
I imagine generative AI, suitably trained on sheet music, can make something that sounds much more like bach, most of the time (with a few bars of pure absurdity now and then). But it sounds like copes code is a lot more useful if we want to actually understand Bach. I miss the days when it seemed like these goals of reproducing and understanding seemed inextricably linked. Spectacular thougn they are, interesting though they are, sometimes LLMs feel like part of the plot of "idiocracy".