“Through Music to the Brain”

Spring, 2004
Presbyterian College,
Clinton, South Carolina

Ann Stoddard, Curator

Piece #1: 

“Please help us understand insect brains by expressing your musical tastes.” 

(No insult to your musical tastes is implied at all!)

Visitors to the gallery help us understand neural events in insect brains through indications of music appreciation.

The story:

Collaborations with neuroscientists have kept me on my toes for some years.  One of them, Christoph von der Marsburg, proposed many years ago that synchronous rhythmic firing was a principle information- bearing phenomenon in a neural network.  Meanwhile another of my long term collaborators, Jim Bower, one of the fathers of computational neuroscience, was building ever better computer programs to represent biologically realistic neural networks.  While more and more is known about the brain, no one knows how to interpret the brain’s internal patterns, the internal coding scheme.

As the great physicist Freeman Dyson has observed, science is most often driven by the availability of new instruments to measure reality in new ways.   A new instrument did appear about a decade ago, in this case in the form of tiny electrodes that are made using technology  similar to that used to make computer chips.  These can be placed inside the brains of living insects, allowing measurements of patterns of neural firing events deep inside a brain.

So what to make of the new data?  The two most common approaches are to analyze data using mathematical techniques and to make graphical representations of it.  In late 2003, Jim Bower asked me to try a third technique.  Since the working hypothesis is that synchronization and periodicity are important, why not turn the data into music, which emphasizes these qualities?  Maybe something surprising will become audible.

We sought out Kevin Daly, who had actually tried this already.  He gathers data from the olfactory bulbs of moths that are being exposed to a series of odorants.   The music he made from this data was interesting, but did not seem to clarify what core structure was actually present in the data.  You can make even random numbers sound like music.  A musical rendering of data ought to reflect something in the data that’s really there even though we didn’t notice it before.

Over the recent holidays I connected some of Kevin’s moth data to a software system that can turn it into music in a myriad of ways.  I have been listening for the past weeks to music derived from the firing patterns in the olfactory bulb of a particular moth that was exposed to a sequence of several dozen odorants.

So far the music hasn’t displayed any overwhelmingly obvious structure, although there are interesting moments.  But the music might very well become informative if I find an appropriate mapping of data to music.

I need your help to do that.  You are listening to music that changes every few seconds.  What is happening is that the mapping of moth brain data into music is changing over time. 

Please press on the big silver button whenever the music sounds “good.”  

I will go over the favorably marked sections to find out what visitors were hearing.  In this way I might accelerate the search for a useful musical mapping of the data.

Please be honest and don’t press the button when the music sounds random.  To correct for this possibility, I am occasionally letting the system play music that really IS random.  So if someone hits the button during one of those periods, I’ll throw out the surrounding button hit records, on the assumption you were just fooling around.

Science is how we can see more of nature, the most precious and lovely stuff we are so fortunate to have inherited in this world.  Please join in me in the use of esthetics to both further science and celebrate it.

Piece #2)

“Face the music”

The story:

One of the pleasures of working with neuroscientists is that once in a while research leads to new computer capabilities.  This happened in the case of Chistoph von der Marsburg and his former student Hartmut Neven.  We have worked on ways to use ideas about how the brain is able to interpret images in order to create programs on computers that are surprisingly good at recognizing objects placed in front of a camera.

In particular, we’ve been interested in the human face.

In this exhibit, one visitor at a time will stride up to a camera and make funny faces.  These result in immediate and wonderful visual and musical effects.

Piece #3:

“Motion into Shape”

The story:

The long term potential of virtual reality in my fondest expectations includes the expansion of the range of means of expression between people.  Specifically, I hope culture will expand to include what I call “Post-symbolic communication.”  This means that people would be able to create content and events within a shared virtual world with a level of speed, ease, and fluency comparable to that of spoken language in the current era.  One could create the world instead of, or in addition to symbolizing it.  The result would be open in possibility, like a dream, but taking place in an intentional waking state, while also being collaborative, like talking.  It would be concrete, but without the limitations we normally associate with concreteness, so abstraction would become an option, rather than a necessity.

The primary obstacle to achieving this potential future is the user interface to support this kind of fluency in world improvisation.

As a tiny step in the direction of imagining this interface, I have created an installation in which body motions are instantly turned into shapes and activities in a virtual world.  This installation has appeared in various forms over time, in places like the Brooklyn Bridge Anchorage, the Roskilde Museum of Modern Art, and the Exit Art Gallery in NYC.


Go back to Jaron's home page.