Do Music and Language Obey the Same Rules?

Emre Sevinc writes "Ever felt as though a piece of music is speaking to you? You could be right: musical notes are strung together in the same patterns as words in a piece of literature, according to an Argentinian physicist. This article in Nature states that Damián H. Zanette's analysis also reveals a key difference between tonal compositions, which are written in a particular key, and atonal ones, which are not. This sheds light on why many people find it so hard to make sense of atonal works. In both written text and speech, the frequency with which different words are used follows a striking pattern. In the 1930s, American social scientist George Kingsley Zipf discovered that if he ranked words in literary texts according to the number of times they appeared, a word's rank was roughly proportional to the inverse of the its frequency squared. Herbert Simon later offered an explanation for this mathematical relationship. He argued that as a text progresses, it creates a meaningful context within which words that have been used already are more likely to appear than other, random words. For example, it is more likely that the rest of this article will contain the word 'music' than the word 'sausage'. Physicist Damian Zanette of the Balseiro Institute in Bariloche, Argentina, used this idea to test whether different types of music create a semantic context in a similar fashion."

Comments

music as a language

[miles zarathustra]

Learning music at the age when the mind is open to acquiring language skills seems to make a difference. The same part of the brain processes both. I read once that people who learn music at an early age tend to have more connections between the right/left brain.

In my opinion, music has taught me way more about programming than the other way around. (and music is more difficult to do effectively -- it's all real-time -- even though the pay is much better for programming)

As a piano player for 37 years now, I always get a kick out of when I can play stuff that's just notes, and it makes people laugh. It's all about expectation and fulfillment.

Partly, my ability to do so springs from my experience playing musical underscore for melodrama shows (e.g. the Gaslighter theatre in Campbell back in the '80's), which is a lot of fun -- translating dramatic dialog into musical themes.

The funny thing is how artificial the harmonic language we think of as natural is. The urge our ears feel to resolve along the cycle of 5ths evolved over centuries, and only seems natural because we grew up hearing music that spoke in it.

Nominally, it's based on the overtone series, but the actual scale we use is based on exponents of the twelfth root of two. A chromatic scale is defined mathematically as the frequencies:

F * 2^(1/12); F * 2^(2/12); F * 2^(3/12)...

Whereas the overtones are simply multiples :

F 2F 3F 4F ...

One is rational integers, the other irrational exponents.

And when you look at how neatly the key signatures and the cycle of 5ths fit together, it's quite amazing ... and the fact that it works emotionally is remarkable when you understand how entirely artificial it is.

I heard once (from my analytic geometry teacher) that Chopin objected to people's emotional reaction to some of his pieces. The semantic world that he lived in, of advanced harmonic modulation, didn't entirely connect with the emotional content he was conveying.

Fractal Math

[serutan]

I'm surprised there is no mention of fractal mathematics in all this. Back in the 80s there was a big article in Scientific American trying to explain why music sounds good. Music doesn't sound like anything in nature. Individual notes might, but melodies don't. So what does it sound like? Popular music, whether classical, jazz, rock or whatever, tends to have a fractal mathematical property. It's in the middle between brown noise, in which each sound is highly dependent on the preceding sound, and white noise, in which there is no relationship. This pattern seems to mimic something about the way we perceive changes in the world around us. If you take two radar scans of an organic landscape -- trees waving, people walking around -- and subtract one from the other, the difference is fractal. If you measure nerve activity with electrical probes you will get white noise on the peripheral nerves, but the closer you get to the central nervous system the more fractal the signal becomes, as if our nervous systems filter out random noise and let the fractal component of our perceptions pass through. Patterns in music might mimic the patterns used by our brains store memories and emotions. This would explain why a piece of music can make you feel a certain way.

Re:Smallest unit of musical meaning

[Muttley]

From the article at arXiv.org the author states [included below] some reasoning for choosing single notes, or at least shows he thought about it. After the passage quoted below he goes on to mention that from a statistical point of view it makes more sense to use notes, seeing as each composition will have thousands of notes. I would argue that these compositions would probably not have all 156 (13x12, within one octave) intervals possible in them at least once.

I agree completely however, saying a piece has 572 As in it says nothing about the music. But it might say something about the statistical correlation between note frequency and tonal vs atonal composition.

An obvious difficulty in modelling the creation of musical context along the lines discussed in Section 2 for language, which are based on the statistics of word usage, resides in the fact that the notion of word cannot be unambiguously extended to music (Boroda and Polikarkov, 1988). In language, words -or short combinations of words- stand for the units of semantic contents, with (almost) unequivocal correspondence with objects and concepts. Moreover, in the symbolic representation of language as a chain of characters, i.e. as a written text, words are separated by blank spaces and punctuation marks, which facilitates their identification -in particular, by automatic means. Music, on the other hand, does not possess any conventionally defined units of meaning. The notion of word is however conceivable in music by comparison with the linguistic role of words as "units of context," namely, as the perceptual elements whose collective function yields coherence and comprehensibility to a message. In music, the role of "units of context" is played by the building blocks of the patterns which, at different time scales, make the musical message intelligible. Yet, the identification of such units in a specific work may constitute a controversial task.

In the quantitative investigation of context creation in music, I have chosen as "units of context" the building blocks of the smallest-scale patterns, namely, single notes. A note is here characterised by its pitch (i.e. its position on the clef-endowed staff) and type (i.e. its duration relative to the tempo mark), and its volume, timbre, and actual frequency and duration are disregarded. The contribution of notes to the creation of musical context, determining tonality and the basis for rhythm, is particularly transparent. In addition, the choice of single notes has several operational advantages. In the first place, the collection of notes available to all musical compositions -or, at least, to all those compositions that can be written on a staff using the standard note types- is the same. This collection of notes plays the role of the lexicon out of which the message is generated. Secondly, single notes are well-defined entities in any symbolic representation of music, either printed on a staff or in standardised digital formats, such as the Musical Instrument Digital Interface (MIDI). This makes possible their automatic identification, which, as described later, constitutes a crucial step in the analysis. Moreover, in order to extract any meaningful information from a statistical approach, it is necessary to work with relatively large corpora. The compositions used in the present investigation contain, typically, several thousand single notes. This figure remains well below the number of words in any literary corpus, which usually reaches a few hundred thousands (cf. figure 1), but is already suited for statistical manipulations.

Re:Hmm

[flyneye]

nice.
This is all old news though.Herman Helmholtz noted that musical scales and their intervals tend to mimic the mother languages rises and falls in pitch and make them available to the musician for phrasing.
A good example of this would be Indian Raga and its 23 note octaves with rules on bending and sliding notes.