This well-written and well-designed book aptly signifies that three to four decades of hard work by a dedicated group of researchers who combined the skills of musicianship and computer programming has resulted in a field that is mature enough that a textbook/tutorial can be written for use by college-level classes as well as by qualified individuals. Throughout the book, the author is generous in his acknowledgments to others in the field whose pioneering work makes this possible. Needless to say, it is also true that many of the procedures described in the book have moved from theoretical possibility to reality, thanks to the vast increase in computing speed and storage achieved over the last decade.
The opening paragraph clearly outlines the aim of the author and the audiences for whom he is writing:
Machine musicianship is both an exploration of the theoretical foundations of analyzing, performing, and composing music with computers, and a tutorial in writing software to pursue those goals. The theoretical foundations are derived from the fields of music theory, computer music, music cognition, and artificial intelligence. The intended audience includes practitioners in those fields as well as composers and interested performers.
The author attempts to meet the needs and interests of readers with different background skills, especially in computer programming. In every chapter there are several pages of programs in C++, and while these data are not essential to the flow of the information, there is no question that skill in C++ is a distinct advantage in using the text. The author also makes frequent reference to the graphic programming language MAX, which is currently available only for the Mac but is supposed to be available for the PC in the near future. Finally, it is assumed that readers have a working knowledge of Musical Instrument Digital Interface (MIDI), which allows musical instruments and/or computers to communicate with each other.
In the ordering of his chapters the author uses traditional musicianship texts as a model. One is devoted to interval and chord recognition, another to the description of rhythm patterns, a third to phrase recognition, and so on. All of these present artificial intelligence problems of enormous complexity--more complex than is often appreciated by laymen in either music or computing. An instrumentalist may read a phrase containing triplets and dotted figures, but perform it (deliberately or carelessly) in such a way that the borderline between the two figures is blurred, and yet the “listening” computer is expected to sort them out in order to “compose” a duet (in real time!) with the live performer. Another example: the computer performer is expected to determine the key of a performance on the basis of a few notes of a melody, but as any musician knows, the key of a melody can be sharply defined rather quickly or can remain quite ambiguous for much of a phrase. As an example of the latter, the author cites the opening measures of Schubert’s song “Gretchen am Spinnrade” (“Gretchen at the Spinning Wheel”), where the melody implies one key while the accompaniment implies a tension between two keys. The author’s detailed description of how the computer program copes with these ambiguities is an effective demonstration of the level of sophistication available in present-day programs.
In the introductory chapter, the author makes it clear that he is not one who (as critics charge) looks forward to the day when “machines will listen only to machines.” He wishes, instead, to expand the horizons of traditional musical experiences. He writes: “I feel compelled to observe that many of us are motivated by a much different vision of the computer’s potential connection to the community of human musicians. … I believe that if computers interact with people in a musically meaningful way, that experience will bolster and extend the musicianship already fostered by traditional forms of music education.” These reassuring words are supported by the approaches, examples, and results demonstrated throughout the book.
In support of the tutorial features, the book includes a CD-ROM for Macintosh and Windows computers. The various C++ programs described in the text (where only part of the code is printed) are available in fully compiled versions.
Some of this material is more accessible to Mac owners, and the author’s introduction to the CD invites users to consult the author’s Web site for up-to-date material. It would have been better if the publisher had delayed publication of the book until this material could be presented directly on the included CD.
Finally, there are 19 tracks of examples of computer music that can be played on an ordinary audio CD player. In their variety of styles and their subtlety, these examples demonstrate, at least to me, the progress that has been made in computer music since the rather crude attempts at tape splicing that began over 40 years ago.
