Scientists tell us about music, again…
A friend on Facebook posted this article from Scientific American about statistical analysis of music. If you’re curious about my little rant below reacting to it, read the article here first then come back:
As usual when scientists try to deal with music, I have a whole host of issues:
1) Music is treated as a phenomenon, as though it descended from the aether, rather than something that is constructed at specific times and places by individuals or groups within a larger system of musical practice within an environment of multiple musical practices. In other words, music, though created by human beings, is approached as though it is “natural”.
2) The culturally situated nature of musical understanding is ignored. Saying that a piece of music can't be reversed in time only makes sense if you assume there is one "correct" way to hear and interpret music. For example, Western tonality is such a system and performing some Mozart aria "in reverse" would not make "sense" in that tonal system, but it would still be a sonic experience from which musical meaning can be created. Reversing melodies of any musical style (whether literally reversing the soundwave or instead performing the notation in reverse order) does not produce non-music; it creates a new arrangement of sounds that can be perceived by anybody however they happen to perceive it. Assuming that reversing a piece of music makes it non-music is to assume that there is only one correct way to hear music, that which is already in place to readily interpret that very kind of music.
3) Using statistical models to determine the amount of "entropy" in different composers' styles to create a scientific basis for these composers' styles' distinctiveness assumes without question that distinctiveness is an important aspect of music, an assumption that underlies the focus on "great men" and "genius" in music history and appreciation. Applying scientific concepts like entropy or time to stylistic and aesthetic aspects of music can lead, once again, to certain musics, probably Western music, being "demonstrated" to be superior to other musics through circular reasoning. This can be seen when the article talks about how "[t]he most compelling compositions, then, would be those that balance between breaking those expectations and fulfilling them—a sentiment with which anyone anticipating a catchy tune’s “hook” would agree." While this idea may seem to be "common sense", should scientists be basing experiments on "common sense", especially when they then go on to make highly generalized claims about such a culturally diverse activity like music?
4) Scientists assume that there is an ideal listener to music (there isn't), and then generalize their findings assuming that everyone is an ideal listener, everyone is identical. An ideal listener is someone whose musical knowledge matches the biases and assumptions of the scientists when they designed their approaches. It creates a circular logic that pretends to be scientific and descriptive when each and every person has a different musical knowledge set that necessarily includes a whole host of extra-musical associations.
5) Entropy only exists within closed systems, such as the entire universe or a sealed gas canister. Open systems that receive energy from an external source, such as the Earth gaining energy from the Sun, do not follow the same behaviors as long as there is energy entering the system. Musical compositions are also not closed systems; they are social interactions that are created only when a performance is heard. If the composition can be considered the "sounds" and their interrelationships, it might seem like a set-in-stone "work" like a Mozart Symphony is a closed system, but it does not exist outside of the work (i.e., energetic engagement) of musicians and audiences. Musical compositions, even recordings, bear a closer analogy to an open system like the Earth due to the necessity of listener involvement, which is analogous to external energy input to the "system". Normally it is physicists themselves who object to sloppy usage of physics concepts like entropy by the lay people!
6) The article mentions that a composition with truly random sounds arranged in time would be high entropy. This is not true. Despite the apparent disorder, such a composition would be fixed in its content and structure, and such a fixed nature allows one to become familiar with it over exposure, such that what may appear highly random at first later becomes its own set of expectations for those with acquired knowledge of that particular piece. Moreover, experience with multiple works that are highly “random” will also generally produce a cultural knowledge in a listener for them to “anticipate” a new “disordered” composition they encounter later on.