mental representation of music
When musicians learn a piece of music, they don’t commit a detailed image of the score to memory, in order to visualize the piece. So what the heck is stored in the brain instead?
David Huron has a lecture (available on the web) where he asks the following yes/no question: “Does the word ‘but’ occur in the lyrics to the song Row, Row, Row Your Boat?” Go ahead and figure it out for yourself, I’ll wait for you…
This is designed to prove something. If you don’t know the song, then regrettably it won’t work–sorry for the cultural bias, but it’s possible to substitute any childhood song and choose a word accordingly. Whay most of us do is “play back” the song internally. We access an auditory image of the song, words and melody. And most people will buzz through the song faster than it’s possible to reasonably sing it, because the task was set to look for a particular word.
Professor Huron Then goes on to make three observations about this.
1. We are able to access mental representations for music. In this case, I had you focus on the lyrics, but the same can be done for melody alone.
2. We can access music-related representations in the total absence of sound.
3. We can manipulate these mental representations in certain ways (such as speeding up the rendition beyond what would be musically acceptable). But we cannot manipulate these mental representations in any way we wish. For example, you might have been able to answer my question much more quickly if you had random access to all of the words of the lyrics. Similarly, it would have been faster if you could start at the end of the lyrics and work your way forward. Either of these two strategies would have generated a faster answer to my question, but as far as we know, people are unable to do this. It is as though the mental representation for Row, Row, Row Your Boat is a linear recording that we must play from the beginning (or from a handful of possible starting points). Once again, my third point here is that we can access and manipulate musical representations only in certain ways.
Wow. You didn’t know how heavy Row, Row, Row was did you? What he doesn’t say, but I am pretty convinced is accurate, is that this isn’t something that is acquired through musical training. Everyone can access a mental representation (auditory image is the term I used, not Prof. Huron) of a song, not just trained musicians.
So if we can all (making suitable allowances for cultural factors–not everyone knows this particular song) do this stunt, what does that mean?
This is the part that interests me most. I’m deeply intrigued by the differences in how musicians and casual listeners perceive music. Since it’s tough to get inside someone else’s head, stunts like this help shine light on the discussion.
It sure seems to indicate that humans have some kind of a “music processing unit” inside the brain. All of us have the basic equipment to answer the question. So what happens for musicians? I tend to think that this facility gets trained through practice and study of music. We learn ways to manipulate the representations of music inside our brains. Some of these learned manipulations include transposition, transcription, reharmonization, or change to a minor mode. Dare I say, playing the representation on an instrument is one of these learned manipulations.
Try it–access your auditory image of “Row, Row, Row,” but “play” it in a minor mode instead.
Cool–you can do it easily, even though I just asked you to take a mental representation, and gave you fairly abstract instructions on how to modify it. But your brain can do it on the fly. I could ask you to count how many notes were different between the minor and major versions, you could do that, and so on.
So it begins to become more clear that musicians gain skills for manipulating mental representaions of music. Performance is one of these manipulations. And it’s a critical one. When we perform, practice, or otherwise realize our mental representations (here I’m allowing for scenarios like the guy sitting at a laptop “producing” music with software as well) we are connecting a kinesthetic image to the auditory one.
Kinesthetic images are your brain’s stored instructions for realizing the auditory image. If I play you a recording of a pianist playing a trill, your brain can perceive it, identify it as a trill, and then access the stored kinesthetic image of how to play a trill. (heh, you’re probably accessing the image right now just from my suggestion). With practice, the connections between hearing, perception, stored imagery, and performance are strengthened, and even made largely automatic. I bet if I play the hypothetical trill recording for most keyboardists, they will access their kinesthetic trill imagery, and do it beneath a conscious level, more like an instinctive response.
But its a learned response. It’s often said you learn all the scales and theory so that you can forget them. (Well, I say that often anyway.) What that’s trying to convey is that you learn them to a degree that it no longer takes a conscious act to access and realize the kinesthetic image. Your consciousness is then free to deal with higher levels of musical abstraction–the factors that add up to what’s known as expressiveness.
Just maybe, this helps us account for the tendency of popular music toward simplicity. Everyone’s heard of a tune with a good hook, or an advertising jingle that’s catchy. To be catchy requires accessibility, which amounts to having a structure that lends itself to the efficient creation and maintenance of the mental representation. The “can’t get this song out of my head” experience is testament to the power of simplicity in this regard.
Rob, these recent posts have been great. As a relative newcomer to the music scene, I’ve been struggling with my ear training more than anything else. Right now, when I’m listening to music or learning a new piece, I’ve been trying to improve my ability to hear a note’s relation to the key center. This has a been a monumental task for me as I’ve been very used to simply playing by rote, or sight-reading without really using my ear to pay attention to what was going on. When I’m trying to ‘remember’ a song without a keyboard onhand, I often still find myself having to visualize the keyboard and recalling actual kinesthetic experience of striking the keys in order to recall which notes were in the song. I suppose it means that I’m still not really hearing the notes clearly in my head. It’s frustrating but I think that learning a song in terms of its scale degrees is the only way to really get the music into my ears. Is this similar to how you would approach learning a new head? Or have you already internalized the sounds that you don’t even really process it on such a conscious level?
Comment by whitecraneboxing — October 23, 2006 @ 6:02 pm
I don’t take any issue with accessing a kinesthetic image–if you can reliably do that now you should be real happy about it. Eventually you wnat to get those so connected to your muscles that you just “intend” the image and it happens without conscious initiative.
But, you asked about how to get to that point. Ear training is a whole different subject, and I have some concrete tips there (which I’ll save for a future post). Right now, what I’ll offer you is that you are absolutely confronting the right skill. It’s hard at this point, but it’s really productive when you get it. Here’s why: if you get the ear trained well enough to do transcriptions with some degree of ease, you suddenly have access to a lot more teachers. Plus, it’s a really commercially valuable skill. Dudes who can whip out charts are “music directors” not sidemen.
So, I would urge you to try for hearing the relationships between consecutive bass notes first. It’s a little easier than hearing the relation to the key center, because key centers can move. But if you can notate the bass line, you can then derive the functional relationships (dominant, supertonic, whatever) easily enough. And you’ll quickly fall into that habit. You’ll get to be able to spot a ii-V-I before it even completes.
I suppose you know about ii-V-I. It’s a fundamental building block, and it gets a lot of love from writers/composers. If you can spot it, you’re well on the way.
Now you also asked how I approach learning a new head. In most cases, I’ve heard the tune enough to have a well-developed mental representation. I have tried lately to work just from that, not a chart at all. Charts are loaded with mistakes and unfortunate notation choices. I’ll usually go back to a source recording if I have doubts, and I try to do it early enough that I haven’t cemented a mistake if I have a doubt. By doing it like this I feel a lot more comfortable playing subs or reharmonizing passages.
Give me an example of a song you’re working up.
Comment by Rob — October 23, 2006 @ 7:49 pm
When learning new heads, what do you mean by having a ‘well-developed mental representation’ ? Are you referring to the abstract structure?
ie: (I )(iii7#5)(IV)(VI7#5)(ii)(VI7#5)(ii)(V)
[first 8 bars of 'Someday My Prince Will Come']
Do you see some sort of image of the ‘chart’ in your head? Lately, when I’m listening to or playing a song, I’ve been experimenting with organizing the charts in a 4 or 8 bar/ line image system in my head so that I can have some sort of visual picture to follow when I’m counting along and/or trying to keep up with the harmonic changes. Interesting how you commented on how some fake books don’t always print them out in a way that makes it easier to see the parallelism of the form; I was just starting to notice that.
Another thing I’ve been experimenting with, is when I’m listening to someone play a song I’ve ‘memorized’ the changes to, I try to deliberately lose track of time and then ‘feel’ my way back into the structure of the song by listening to the bass notes that are being played on the 1-beat of the each measure. This is somewhat easier for song forms like the 12-bar blues where there are only 3 major chords most of the time and I can ‘feel’ a shift to and from the V or IV chord or back to the I. RIght now it’s almost impossible for me to do with any consistency on any other kind of song. In the case of Someday My Prince Will Come, sometimes I can hear the Tonic note being played (Bb) and then I can listen for something that might sound like the 3rd and then I can tell that I might be on the 2nd measure of one of the sections, but then I have to decide if its the A section or the B section which means I need to listen for the (#IV) root. When I’m listening for the root movement, I’m listening to the bass note’s relationship to the Bb, not really listening for the interval relationship between the previous note and the new one. So my ‘method’ for ‘memorizing’ a song’s root movement is really just singing the root notes along with the melody and just trying to get a ‘feel’ for what sounds right, similar to how one might learn a the melody to a children’s song: just keep repeating it until the sound is in my head. The next trick is getting my fingers in the left hand (playing the bass notes) to ‘feel’ their way along with the root movement in my head without requiring an actual intellectualized command from my brain (ie: “the next root sounds like it’s the 6th of the of tonic, therefore its going to be the G and so I need to move my index finger from the Eb up a major third or down a minor 6th.” etc.)
Comment by whitecraneboxing — October 23, 2006 @ 8:57 pm
Yeah, I don’t see anything like an image of the chart. Instad, I access a representation of the functional units of the song. So a nice compact one is the second four-bar phrase there, it’s a prolonged dominant. Writing it as (ii)(VI7#5)(ii)(V) is correct, but also is being slavish to the notation. Inside my head, it’s stored more like ii-V of Bb, but you walk chromatically down to V from ii. And the words are just a description of the pattern, I’ve seen that pattern so many times that I have a “handle” for it.
So if you can get to where you see the larger structural pieces like that, you learn a tune way faster. At best, in on elistening. If you were playing this tune at me, and I’d never heard it before, by the time you got the ii chord with the major 7 (Bar 6) I’d be anticipating at least the next three bars.
And yeah, a great way to learn is to copy charts. When I do it, I use 4 bar phrases (except those rare cases where something else is “right”), and I don’t use first and second endings. I always recopy the chart so you can play it straight down. Then the form reaches out and smacks you. In this case, the form is 32 bars, ABAC.
Your experiment sounds a lot like playing “drop the needle.” That’s an awesome training technique that is a relic of the days of the phonograph. Just like you said, you enter in the middle and try to orient as fast as you can to where you are in the form. A fantastic drill. Jamey Aebersold used to teach that.
Here’s the “big idea” for you to take home from this: Once you’ve learned a few dozen songs, you’ll start to see more similarities than differences. It’s like a compression algorithm for your brain. Right now, you just can’t access the most efficient internal auditory image for what I described above. But it’s not far out of reach either.
Just the act of going through your favorite fake book and documenting the form (32 bar ABAC) is a good start. That’s kind of the root node of the hierarchy of cognitive stuff that is the mental representation. It decomposes into four parts, three of which are unique (wow, already we’re achieving compression). A is built of two four-bar phrases. The first is a fairly idiosyncratic progression (you won’t see it reused a lot) that builds tension by ascending root motion: up a Major third, up a semitone, up a Major third. The second phrase is the prolonged V chord mentioned above. Notice how the last chord of the first phrase is the dominiant of the first chord of the second phrase.
Now, that’s analysis, but it really will help you to understand the progression in terms of what’s going on, rather than what symbols the guy who engraved the chart selects. It’s also a hierarchy. Wholes are built of parts, which are built of parts. Those parts are quite well-used! You don’t run into many chord progressions that are truly distinctive. At the phrase level,it’s all been done.
Look for voice leading opportunities in the harmony. The descending chromatic line in the second phrase of A is reasonably obvious, but there’s a more subtle ascending one in the first phrase of A. (F-> F#-> G-> G).
Okay, your job is to analyze B and C for me in similar terms to how I presented A for you. I did half the tune, now you do the other half. :D Big hint: turnarounds.
Comment by Rob — October 23, 2006 @ 9:25 pm
B Section:
(iii)(ii#o7)(ii)(V)(iii)(ii#o7)(ii)(V)
Chromatic movement for the first three measures, then goes up to the fifth. Is it a giant v chord that leads back to the tonic chord? Could I think of the whole thing as a ii-V?
C Section:
(I)(III7#5)(IV)(#IVo7)(V)(VI)(ii)(V)
More chromatic bassline movement from iii7#5 to the VI. And then goes to the ii V, Could I think of the whole thing as a giant I-VI-II-V ? If this is the case, how would change the way that one approached playing the song?
Why is it so important to look for voice leading opportunities? For that matter, why does good voice leading matter? And what defines it a ‘good’ voice lead? Just a the least movement possible in the highest voice, ie: smooth voice leading?
Comment by whitecraneboxing — October 24, 2006 @ 9:18 pm
Let me come back to the voice leading question–that deserves its own post really.
The issue I have with your B analysis is that ii#o7 isn’t a functionally meaningful symbol. change that to VI and we’re on to something. iii-VI-ii-V is a turnaround. This is a fundamental building block of music, and jazz standards especially. It is another form of dominant prolongation, yes. You can look at iii-VI as the (ii-V) of ii: ii/ii-V/ii. Note how these relate to the circle of fifths–they are adjacent. Each root movement is by a fifth.
I have two issues with your analysis of C. Here’s where I remind you of my earlier comment that it’s not as important to relate every chord to the tonic. We are driving toward a landing point, which is the IV chord. The prior two chords are just its ii-V. The chord that follows IV functions as a secondary dominant (V of V) so the first phrase of C is (ii/IV)-(V/IV)-(IV)-(V/V). Then a slightly modified turnaround: (V)(VI)(ii)(V).
The takeaway here is that the dominant-functioning chords are setting up your anticipation of the subsequent chord. That’s why I’m going to such trouble to point them out and make the distinction between V of something and the various symbols like ii#o7 or #IVo7. Diminished 7 chords are dominant chords. You need to interpret them in the context of how they resolve.
When you do that, they will stick in your brain a lot better.
The other thing that’s nice about reducing things to these bite-size chunks is that it makes it a lot easier to do the all 12 keys thing. If I have you work on a iii-IV-ii-V in all 12 keys you wouldn’t be too upset about it. But look, that’s more than half of this tune. There are really only two non-turnaround phrases–the first phrase of A and the first phrase of C. So really–there’s only 8 bars of material out of 32 in this tune that isn’t a stock turnaround.
Comment by Rob — October 24, 2006 @ 10:05 pm