Sunday, March 08, 2015

Overture: Alien Music

As described in this previous post, the  text below is a draft of one of several "interludes" to be included in a book that I am working on concerned with music and artificial neural networks.

Figure O-1. The musical signal composed by John Williams for Close Encounters of the Third Kind.
Steven Spielberg’s 1977 movie Close Encounters of the Third Kind declares that we are not alone, and we should not be afraid.  The film follows ordinary people after they experience a close encounter with an unidentified flying object.  After this experience, the protagonist of the movie, Roy Neary (played by Richard Dreyfuss), becomes obsessed with seeing the UFO again, as well as with creating sculptures (from mashed potatoes and other materials) of a shape that has a deep meaning that he cannot quite fathom.

Eventually it is revealed that this shape represents a monolith of rock (Devil’s Tower, Wyoming) marking the selected location for first contact between humans and aliens.  The obsessions of Neary and others are produced by an implanted invitation from the aliens to journey with them.  The invitees recognize the shape from television news broadcasts concerning a story about a (hoax) anthrax outbreak in Wyoming.  They spend the latter part of movie traveling to Devil’s Tower and dodging the efforts of the authorities to stop them.  In the final scene of the movie only Neary succeeds; he is welcomed by alien children as he ventures aboard their mother ship.

A simple musical signal plays a key role in the film, and is presented in the musical score above (Figure O-1).  It was composed by John Williams.  He strove to write a signal that it was long enough to be set apart from the simplest musical elements (e.g. a chord or an interval), but not so long to exist on its own as a melody.  He decided that these goals would be achieved by a theme that was only five notes in length, and composed about 350 different five-note permutaitons.  Spielberg liked the one presented in Figure O-1 the best, and it became one of the most famous musical themes in film history. 

The musical theme is first encountered in the film being sung by hundreds of pilgrims at Dharmsala, India.  When asked where they first heard it, they all point, in unison, to the sky.  In a later scene, scientists broadcast the tones into space, and receive the coordinates of Devil’s Tower in return.  At the climax of the film, the signal is performed on an ARP 2500 synthesizer located on a massive runway constructed atop Devil’s Tower.  The music greets, and communicates with, the alien visitors.

In this final scene the keyboard is played by a young, fresh-faced Phillip Dodds.  Dodds was in fact the engineer that ARP sent to set up the synthesizer used in the movie.  When the original actor who was to play it became sick, Spielberg saw Dodds working on the machine, liked his look, and cast him in the role.

As the scene unfolds Dodds is instructed to play the musical signal faster and louder while the chief scientist (Lacombe, played by Francois Truffault) strides out along the runway.  Eventually the enormous mother ship arrives, hovers over the runway, and begins to loudly copy the notes that Dodds plays.  This musical mimicing quickly erupts into an interstellar jam session of increasing tempo and complexity.  Awestruck and wide eyed, Dodds exclaims “What are we saying to each other?”

This is a very deep question indeed.

The intended message in the film is that music is – literally – a universal language, one shared by all intelligent life forms.  That the alien ship generates the same notes, and that it jams with Dodds’ character in the same musical system, supports this optimistic view.  To answer Dodds’ question a scientist standing beside him says “Seems they’re trying to teach us a basic tonal vocabulary.”  Another immediately adds “It’s the first day of school, fellas.”

However, other intriguing scenarios exist; there are alternative musical possibilities that Spielberg did not explore in his film.  Does the fact that both the ARP 2500 and the mother ship generate the same basic tonal vocabulary imply that the humans and the aliens share the same underlying musical theory?  Or instead is it possible that a completely different musical theory – an alien music theory that is dramatically different from our own – is still capable of generating the same patterns of musical notes?  Perhaps all these fellas know they are attending the first day of school, but are not aware of the lecture topic! 

In many respects, my aim in this book is to explore the possibility of such alien music.  However, instead of requiring a close encounter of the third kind for this exploration, I am fortunately able to adopt a much more terrestrial approach.

This method involves teaching a particular type of computer simulation to generate responses that are consistent with Western musical theory.  For instance, the computer simulation can be presented the tones that define a particular scale, and can learn to respond with the tonic note of that scale, or to identify that scale as being major or minor.  From this perspective, similar to the alien encounter at the end of Spielberg’s movie, the computer simulation and I are saying something to each other, and are learning to use a basic tonal vocabulary.  Our inputs and outputs are consistent.

However, the particular type of computer simulation that is being trained in this fashion is called an artificial neural network, and while it learns to generate the correct outputs to various musical inputs, it is not taught – or constrained by – traditional Western music theory.  Many researchers in what is known as connectionist cognitive science argue that the internal workings of artificial neural networks are quite distinct from the clear formal properties found in logic, mathematics, or music theory.  As a result it is possible that the artificial neural network can discover a completely novel approach – and alien music theory – that generates the same input/output relationships that are defined by Western music theory.

In order to determine whether this is possible, after a network is trained we have to examine its internal structure to discover how it is able to generate its musical responses.  As is explained in more detail in Chapters 1 through 3, an artificial neural network is a messy collection of different processors (analogous to neurons) that send signals to one another through a larger and messier collection of weighted connections (analogous to synapses between neurons).  The musical knowledge of a trained network is represented in its internal patterns of connectivity.  The messiness of such patterns – called distributed representations – is what makes the existence of an alien music theory plausible.  We can use a variety of techniques to make sense of a network’s internal structure, and in so doing can reveal the musical regularities that the network has learned, on its own, to exploit.  Of key interest is whether the theory that the network has learned is the same as our own.

Interestingly, and as will be detailed in the chapters that follow, artificial neural networks can discover novel musical theories that seem completely alien.  This has interesting implications for music, insofar as it reveals alternative musical formalisms.  This also has important implications for the study of musical cognition, because it reveals a variety of different kinds of representations that the human brain might use to process music.


No comments:

Post a Comment