This is a letter I sent to Nature (rejected as usual, but I never lose hope!) regarding the article series on music and science, and then some further comments on the interview at the 02/07/08 nature podcast.

Greetings. The ongoing essays regarding music and science are great, but I would like to express some opinions.

In the nice article “Raising the roof”, NATURE|Vol 453|12 June 2008 pg 859, the author mentions that “Research moved into the realm of experimental psychology and away from the area of physics”. It seems to me that this sentence did not transmit faithfully the actual interdisciplinarity that happens in the study of hall acoustics. Was the “away” strictly necessary?

The study of music, from its composition and to techniques for its production and execution, is full of opportunities to make interdisciplinary research, or perhaps “extra-disciplinarity” research. I would have liked to see this more stressed in the text, which also pictured architects and acousticians as in a bit of competition instead of cooperation.

I would also like to make a comment on the articles regarding statistical analysis of compositions, and the related automatic composing programs. These programs are based on the idea that one can study the musics from certain composer and replicate their measured characteristics with an algorithm. While I enjoy very much to follow this kind of work, it bothers me that there is a fundamental question not receiving much attention: who said that the scores available from a certain composer or style contain enough data if compared to the complexity of the algorithm needed? In other words, it could be the case that the best programs would be so much complex that the size of one music, or a small number of them, would not make justice to this complexity. In that case, music created from extrapolations from an input set would be much more related to the algorithms used and to unconsciously introduced bias, than to the actual data. This is an important issue that I believe is not receiving much attention because people still amaze themselves just with the concept of making objective analysis of musics, and even more with the idea of a computer program composing a song. These are ideas that never seem to become finally accepted as anything but new and radical. It’s about time.

*

There are many things said at that interview at the 02/07/2008 nature podcast that let me very unquiet. (I haven’t heard the full version yet, and I am sorry I don’t remember the folks names, I will check later and update here.)

I have studied music for many years, and I am an electrical engineer. The dissertation I wrote for my master degree was about electric guitars and distortion. I have also read a bit about composing music, and I also had a couple of bands. During this personal voyage I have read many texts, some are good, some are not. Some researchers have opinions I like, some I dislike, just as in every other area of study; except that music being this controversial entity, capable of making people have the strangest and strongest reactions, the number of articles I disagree with or simply don’t understand are quite large.

There are a number of things that are often said, and that I disagree with. For example, the enduring feeling that an objective study of music is something strange an “counter-intuitive” (I also dislike intuition intuition itself, and the anglo-saxon reliance on it).

I also dislike when people say different areas of study are finally trying to get along together to see if they can try to understand music, or some other controversial topic… This division of science and of human affairs into well-defined compartments is nothing but an illusion we create when young to try to understand the world. It is something artificial, and more than that, these distinctions are often modern. We talk today as if interdisciplinarity, or multidisciplinarity or whatever were something new, a revolution waiting to happen, but if you study the history of science you will see that the collaborations between the said distinct research areas has always existed. Interdisciplinarity is here already, we just have to embrace it, we just have to stop fighting towards “disciplinarity”. I was very happy in my research when I saw that I could very naturally start telling the story of the topic with Galileo Galilei. Yes, the physical analysis of musical instruments is that old… (perhaps we both find it so natural because of our shared sun sign, aquarius?)

That being said, something else I dislike is when people suggest that “science” is something separate from other fields such as psychology, sociology, music, literature and whatever. AFAIK, science is all that, all the research areas inside a university… We must stop saying that “science” is strictly the said “exact” or “natural” sciences like physics biology and electrical engineering (whcih some people also don’t consider to be science BTW). Science is all that, it is only different from, for example, the activities of a painter, writes, lawyer… A lawyer is a great example of someone whose work is non-scientific in nature, as during a judgment, for examle, rhetoric is often used. (Inside the university, this non-scientific activity is studied in a scientific way.)

And all that being said, I dislike when people say someone discovered that there was a relation between a certain something, and with math. This is somewhat of a redundancy, for a “relation” is necessarily a “mathematical relation”! People don’t discover relations between a certain area of study and math, they discover relations inside that area, using math as a tool.

Mathematics is everywhere. It’s in music, sex, psychology, painting, death, life, humor, drama, sculptures, history, law. It’s never a surprise to find math somewhere, and whenever that happens, it’s nothing but we “understanding” something. “Understanding” can often be understood in mathematical terms somehow.

That all whole being said, I love to study the said quantification of subjective feelings. It’s a specific form of application of math in fields such as the arts that allows some very interesting analysis to be done. One thing we often find out when we do such studies is that the perception of people about something can vary widely. Subjects can say that they are feeling quite different things, and can even contradict themselves from one experiment to another. It can be even worse, they can contradict themselves during the experiment! It’s very interesting to study these experiments because they are a great challenge to our skills of probability theory, model identification, pattern recognition and even epistemology itself.

So, I dislike when someone try to reduce the experience of music to something extremely simple as “surprising” the listener. It’s even worse when the feeling being studied is characterized by literary words such as “moving”, or “life changing”. Life changing in my opinion is a bullet in the head.

I don’t mean a lead bullet, of course, I mean the Rage Against the Machine song called Bullet in the Head . The songs by RATM and many others are characterized by repetitive patterns, but of course, also have unrepeated events here and there.

Something interesting to notice is that while the musics might have these unrepeated events, they become repetitive when you start to listen the music, then rewind the tape and play it again and again and again. One of these unrepeated events might surprise you at first… Actually it might come unnoticed at first, but they you listen to the music and notice it, you get “surprised”, but then you listen to the music oven and over, and it is not a “surprise”. But you might like it just the same. I can say that there are a number of musics with unrepeated events (like screams by the singer or pauses) that I consider some of the greatest moments of these musics…

But on the other hand, there are repeating events I do like too, that “changed my life”. Some James Brown songs are quite periodic, they do have some variations inside it, but the periodic part can be as much, if not more “life-changing” and satisfying than the rest.

There are some musical styles (specially Brazilian chorinho) where while the performer is playing, listeners start to yell “don’t stop!… Keep going!” They like what is happening, and fell like it must keep that way, that a change must not occur at that moment.

If surprises were so important in music, why would people ever feel the need to listen to a certain song another time? I am not saying different interpretations, not even different live performances, I am saying the same song, played to your ear with the same electronic device with the same headphone set. I know I do it, e.g. I love to listen to Pyosisified (Carcass) again and again…

By the way, I mentioned silence… One thing said already too often about music is that the silence can be as important than the sound (or “noise” if you will). Couldn’t it be the case that the “surprises” in music are just as important as the “periodic” parts? After all, something is only surprising if we have a non-surprising event happening at the same time to compare. This is easily appreciated by someone who has studied Claude Shannon’s Information Theory.

I believe this idea that the “surprising” would the the place where the “genius” or the “human element” hides is directly related to the fear / respect we have for the unnameable, the unimaginable… It’s related to the fear that science can kill god, and stuff like that. People who defend this are implying that good music would be inherently those that a computer program cannot produce, or that cannot be grasped by a mathematical equation “predicting” it. It’s just like that dragon-in-the garage tale…

I am sorry if my wild claims has offended anyone, I wish I could make this text a bit more subtle, but again, It’s literally more interesting to have surprising harsh criticisms from time to time. Subtle texts are boring, aren’t they?

And on it goes…

*

The same day I wrote this I read at Slashdot that someone else came up with another revolutionary theory that would “at last” explain why the legendary Stradivarius violins would be so much much better than others.

This is a great example of how people don’t take music and its scientific study seriously. Although the subject has already been extensively studied for decades, by physicists and engineers and instrument builders and musicians, the public (even specialized public) doesn’t know of these studies, and even consider it ludicrous to perform these studies. There is so much misinformation and misconceptions around that from time to time you can see a new “genius” having ideas, and people around marveling at it just because “it’s science mixed with music, how fun!”…

Please, everyone, go look for any of the numerous books on the subject in the best bookstores… At least look for this great Scientific American article on the subject:

Carleen Maley Hutchins. The physics of violins. Scientific American, 207(5):78­93, November 1962.

(it’s the good old SciAm, of course…)