Credit: veer images
The elusive quality that makes a piece of music, a ballet, a painting or a face seem beautiful or interesting isn't all that elusive.
It has to do with geometrical patterns that can be described by relatively few bits of information, according to Jürgen Schmidhuber, a computer scientist at the University of Lugano in Switzerland and an expert on topics ranging from theories of beauty and complexity to artificial intelligence.
The successful coding and decoding of compressible patterns is at the core of not only aesthetics and art, but also science and philosophy. This is superbly exemplified in the laws of physics, which explain complex observations with concise mathematical expressions.
In fact, Schmidhuber says artists and scientists are very much alike, in that both try to create or discover surprising information. Surprising in the sense that it contains previously unknown regularities which, when decoded in the mind of the observer, elicit pleasure.
Compression, a central concept in the branch of mathematics called information theory, also lies at the heart of the biological origin of human appreciation for music, according to a hypothesis put forward by Nick Hudson at Australia's national science agency, CSIRO.
"Without knowing it, most of us actually encounter the principles of information theory every day in things like picture and music files where data are compressed into neat, small packages," Hudson says.
A muscle and systems biologist researching red meat production in his day job, Hudson spent many nights pondering the biological rationale for why we experience some genres or pieces of music as more beautiful and pleasing to the ear than others. "I wanted to explore further how this was connected with pattern recognition and data compression."
Undertaking a computer analysis of various pieces of music, Hudson found the degree of compressibility of the music correlated with its perceived beauty. He hypothesised that highly compressible music - typically classical masterpieces - contain patterns that appear more complex to the ears than the patterns in simple, catchy pop songs, but that this first impression is reversed when the brain deciphers the patterns.
"From an information theory perspective, classical music is apparently complex, but really simple, whereas popular music is apparently simple but really complex," Hudson says. "And what we sense as beauty is the unexpected compressibility of music."
Hudson believes this feeling is a 'parasite' on a cognitive information compression skill originally evolved to help us store and make sense of the endless data we absorb during our lifetime.
