Music and Language

One of the very first posts I put up on this blog over a year ago was on music and language and in January I put up one on music and meaning. But this is such an interesting and complex question that I will take it up again. Thanks to a friend of mine who is a philosopher, I have a bit more to say about it. He mentioned that "significance relies on some meta-subjective but non-natural correlation of signifier to signified ("smoke" signifies smoke)." Putting that in more ordinary language, we understand the meaning or significance of something if there is some agreed-upon connection that is not simply a natural phenomenon. If we just see some smoke, that isn't necessarily a communication unless it comes in measured puffs! But the word 'smoke', which uses standard letters to form a word we can look up in the dictionary is significant and refers to the phenomenon of smoke. All this is painfully boring to you semioticians out there, but bear with me!

Now, musical notation is language-like in that it signifies things, just like words formed from the Roman alphabet. If you know how to read it, the significance of this is perfectly obvious:

To a guitarist this says "play a C major chord (probably in first position with your third LH finger on the low C, your 2nd finger on E and your first finger on the high C--the G is an open string) and hold it for four beats." Now that is certainly a musical instruction, but it is not at this point, music. Assume the guitarist plays the chord and holds it for four beats, what does a listener hear, musically? A major chord. But there is really not much of a musical effect unless it is followed by a whole bunch of other stuff. This other stuff will consist of various notes forming chords and melodies and they will be articulated in time with rhythms and a meter consisting of groups of measured pulses or beats. Perhaps we will end up with something like this:

That piece starts with the simple C major chord of my example and goes on to be a complete, simple, piece of music. Each measure of the notation signifies to the guitarist instructions such as I described above. The composer, guitarist and listener are all perceiving this as a piece of music, that is to say, a structure of sound in time. The musical work exists in different ways in different places: the composer's mind when he was writing it, the music notation, the guitarist's mind as he learns how to play it, the performance itself and finally in the mind or ear of the listener. Given that, perhaps we can take a stab at  the kind of meaning there is at each stage. In this particular case, the composer was wanting to create something that would teach a guitar student something about different 'voices' and about harmony. The guitarist has to solve a number of technical problems, especially for the left hand, in order to realize the music. If he is successful, the listener hears a serene piece in C major with independent voices and nice suspensions.

Sorry for the lengthy example! We can derive from this the reasons why music has some language-like aspects, but is not a language. When a musician looks at the score to this piece he sees a precise notation of the sound structure. With it, he can perform the piece. But neither the instructions, nor the resulting performance can easily be put into words. You could describe what you do: "first I put my fingers on a C major chord and pluck with the right hand..." but it would be terribly lengthy and pointless. And what can a listener do to describe the performance? "There were a lot of chords and they seemed to fit together well and the guitarist was wearing a white shirt..." Also terribly lengthy and pointless!

So how is the music itself language-like? Not in any specific way, certainly. But it is hard to deny that communication is going on and and very clear communication at that if you are acquainted with this kind of music or with guitar-playing. Which brings me once again to a famous quote by Mendelssohn:

A piece of music that I love expresses thoughts to me that are not too imprecise to be framed in words, but too precise.

 I think that where music fails to be a language is in the correlation of signifier to signified. We have all those notes, but they don't point to something outside themselves like the word 'smoke' points to a plume of smoke. This gives them the freedom to have expressive power without being caught in a web of language. Now songs, of course, are a whole different story because they operate on both levels: musical expression and the significance of language. Here is the first song from Dichterliebe, a masterpiece where a great composer of song, Robert Schumann, sets sixteen poems by Heinrich Heine, a master poet.

Read more »

Music and Meaning

Music can mean a lot to us. Certain pieces or songs remind us of a particular time in our lives, a particular emotional content. But the kind of meaning we find in music is not the kind we find in language. Musical meaning doesn't signify the way language does. Of course, some music has text and the text has linguistic meaning. A song can tell a story. But music without text can also have meaning. Without semantic content, though. You can't say in music "the ball is red"

But on the other hand, you can't say in words

Music notation is both a set of instructions "put tab A in slot B" and a description of a soundscape. The illustration above, converted into instructions, might be something like "play three eighth note Gs followed by an E flat; the length of each eighth note is 0.14 seconds; play the notes loud, etc..." The cumbersomeness of this is why prose does not make for useful music notation. With training, you can read a musical score and "hear" it in your head; that is, you can estimate the resulting sound. As proof of this, a typical test in ear-training is to hand someone musical notation as in the illustration and have them sing it.

Some instrumental music almost seems to have semantic content. I once performed an experiment with the cooperation of a class I was teaching. There were about 100 students, so a good-sized sample. The experiment was I played a movement from an orchestral work and asked everyone to write down a few sentences describing what was going on. I suppose it was like asking them to imagine the movie this was the soundtrack for. I did not tell them anything whatever about the piece and I doubt if any of them had heard it before. If you want to perform the experiment yourself, just listen to the following clip without looking at it.

The vast majority of the students, at least 90%, wrote some variant of the following: "there is a great conflict or battle between two opposing forces, one good, one evil and the battle ends with good triumphing". This is the third movement from Shostakovich's Symphony No. 12, "The Year 1917" and it represents in some way the events of the Russian Revolution. The third movement is titled "Aurora" after the battleship that fired on the Winter Palace signalling the beginning of the conflict. Here is Wikipedia on the symphony.

Music can communicate very powerfully. What it can't do is communicate specifics: "meet me at 7" or "property is theft". What it can do is express how we feel in general, not specific terms. Now that raises some interesting possibilities. Can we use music as a kind of time machine to get a sense of how people felt in the 18th, or 16th or 14th century? Well, possibly. But we have to take into account that we can only hear from our perspective and the melodic, harmonic and rhythmic techniques of music have changed radically over time. The depiction of a battle in a piece from the 16th century is not going to sound much like the depiction of a battle in a piece from the 20th century.

One last thought. Beethoven titled his Symphony No. 3 the Eroica, and dedicated it to Napoleon. But, as the story goes, later on he scratched out that dedication and made it merely to "the memory of a great man". Here is the Wikipedia article with the story. Here is the title page with the scratched out original dedication:

Now here is what is interesting: when he changed his mind about Napoleon after he made himself emperor, Beethoven had to change the dedication but he didn't have to change a single note in the symphony! The expressiveness of music is non-specific, which is ofttimes a good thing. Because of this, Shostakovich was able to write symphonies that were both acceptable to the authorities under Stalin and expressive of the torment and oppression felt by the people. Here is the first movement of the Eroica:

And here is the slow movement of the Symphony No. 5 by Shostakovich:

All this makes one recall a letter Mendelssohn wrote in which he said:

A piece of music that I love expresses thoughts to me that are not too imprecise to be framed in words, but too precise.

This is saying the opposite to what I have been saying here! But oddly enough, I think there is truth in both. What Mendelssohn is saying, as is clear from the context of the passage, is that while most people find language precise and music imprecise, for him it is the opposite. He finds language very imprecise with the meaning often unclear and easily misunderstood. Music on the other hand, to Mendelssohn, is perfectly clear even though inexpressible in words. Just as, for me, I know exactly what the meaning (significance? import?) is of the short theme by Beethoven I quoted at the very beginning of this post. But I really have no idea what someone might mean by saying "the ball is red". Unless we are playing billiards. It might all be some murky metaphor...

Read more »

The 'Language' of Harmony

Somewhere I made the claim that music is not a language despite many assertions that it is. I think there is at least one book titled The Language of Music. But, as with phrases like "the language of love", this is just a metaphor. As someone once said, if music is a language, what are the words and where is the dictionary where I can look up what they mean?

But that being said, I was musing over a comment and my response to the last post on cadences and I realized that, while music is not a language in the sense of being able to communicate specific concrete meaning, it does have language-like aspects. Music has a 'vocabulary' that is learned. Harmony, especially, is something that one only becomes familiar with over time. Counterpoint is another. Melody and rhythm do tend to have a more direct impact, but it is perfectly possible to listen to music having only the sketchiest notion of harmony and counterpoint.

One implication of this is that music is not a 'universal' language. Even Esperanto has to be learned. If you want to really appreciate the drum music of Ghana or the gamelan music of Bali or the mbira music of Zimbabwe, then you need to learn the 'language' just as you need to learn the harmonic 'language' of Haydn to really see what he is up to in a sonata movement such as the one I put up in the previous post.

That being said, it is also true what my commentator was mentioning, the fact that consonance and dissonance are related to certain acoustic properties. The relation between the frequencies of C and G is a simpler one than that between C and F# which is why the interval C-G sounds 'pure' while the interval C-F# sounds harsh. But the fact that harmony is based on certain acoustic properties does not mean that all the possibilities and developments are perfectly obvious. The Prelude to Tristan und Isolde is based on simple acoustic proportions, but it extends and expands the implications to an amazing degree. For many minutes it simultaneously implies a tonic chord, while doing everything it can to avoid it! Enjoying and understanding this takes quite a bit of exposure to harmony as practiced in Western Europe for a few hundred years.

And here is a little mbira music:

Read more »