An Engineer Looks at Melody

[fiziwig]

First, a disclaimer. I know nothing about "real" music theory. Now that I'm retired, I'm going to be taking some theory class next year at the local community college, but before I get "contaminated" with other people's ideas, I thought I'd jot down my own thoughts on how melodies are built. If you expect "traditional" music theory from me, you'll be disappointed. But if you'd like to take a look at an engineer's ignorant ramblings on melody, I've started jotting them down on my web site: http://www.fiziwig.com/music/melody01.html

This is all from an engineer's perspective, examining melody the way I would analyze any structured edifice. For what it's worth ...

--gary

[chckn8r]

Have to admit, a very interesting take on melody analysis.

I haven't been able to get through the whole analysis (on vacation right now and kids are in and out...), but it seems as though you're hitting on some key things that have also been touched upon in other posts here too!

More on harmonizing melodies: ive got a melody... now what?

Melodic writing:
Help....

I'll try and get some time to really digest what you've written, but on first read-through, what you've touched on is finding the resting or "kernel" tones of the melody. Although, in your analysis, you find them by mathematically subdividing measures - which works, for the most part, for Yankee Doodle, but is probably only applicable to simple, rhythmically symmetrical melodies.

What you'll need to do next, in order to make your model a bit more scalable and applicable to other, more complex melodies, is apply the harmonic layer. What chords are being played under the melody and determine which notes of the melody are the "core" notes that emphasize and fit into the harmony and which notes are the "ornaments"?

Very impressed by the depth of independent analysis on this!

[ttw]

You may want to look at Goetschius book, "Exercises in Melody Writing," to see how this is handled by some musicians.

[stevel]

Originally Posted by fiziwig

First, a disclaimer. I know nothing about "real" music theory. Now that I'm retired, I'm going to be taking some theory class next year at the local community college, but before I get "contaminated" with other people's ideas, I thought I'd jot down my own thoughts on how melodies are built. If you expect "traditional" music theory from me, you'll be disappointed. But if you'd like to take a look at an engineer's ignorant ramblings on melody, I've started jotting them down on my web site: http://www.fiziwig.com/music/melody01.html

This is all from an engineer's perspective, examining melody the way I would analyze any structured edifice. For what it's worth ...

--gary

Gary, I don't want to poo-poo all your work, but a couple of points:

1. Even though "quoted" "contaminated" this is something common I see. Many people believe that they will be contaminated by learning theory. No. You can choose to use theory or not. However, not learning theory will never enable you to use theory. It's better to know it and choose to use it or not, than to not know it IMHO.

2. You're basically doing what thousands of people have done before you, only you are discussing it with your own terminology (some of which others have used) and without the whole picture. Once you take some theory courses I think you'll be able to form these concepts into a version that uses standardized terminology (which is important for communicating with others in the field) with a deeper understand of the similarities and differences between Art, and Math.

Peace,
Steve

[fiziwig]

Originally Posted by stevel

Gary, I don't want to poo-poo all your work, but a couple of points:

1. Even though "quoted" "contaminated" this is something common I see. Many people believe that they will be contaminated by learning theory.
...

2. You're basically doing what thousands of people have done before you, only you are discussing it with your own terminology

...
Peace,
Steve

Good points. I do plan to take some theory classes next year.

Perhaps "contaminated" was a poor choice of words. I've observed in other fields of study that once a person has been "indoctrinated" (uh-oh, another charged choice of words ) to a certain way of approaching the problem it can make it very difficult to approach the problem from a different direction. (E.g., a logic course I once took was all about giving names like "Doris" or "Barbara" or some such to syllogisms. From the broader perspective of symbolic logic, giving a particular Boolean equation names like "Penelope" is just so much nonsense. As a "hard science" guy, I can't help but wonder how much of traditional music theory is similarly non-systematic and "artsy".)

It doesn't at all surprise me that my thoughts are not original, but I did want to get them down before learning traditional theory.

Your point about speaking the same language is well taken. I started to read another article in these forums that looked interesting, but soon found myself facing terminology that I'm unfamiliar with (not to mention Roman Numerals! Yikes! What's THAT about?). I'll revisit that, as well, after I tackle some good old fashioned traditional theory.

Clearly I need to know proper music theory in order to communicate.

--gary
My compositions: http://www.fiziwig.com/

[ttw]

Just for fun, try your method on a Gregorian Chant or even the Dies Irae.

[stevel]

I can't help but wonder how much of traditional music theory is similarly non-systematic and "artsy"

There are three parts to Music Theory: Terminology and Symbology; Comparative Analysis; and Practical Application (my terms). T/S deals with things like naming chords, identifying scales, understanding note values, etc. CA deals with discussing similarities of pieces and forming conceptual models based on observation, and PA has to do with using the other two to perform and compose music.

So some of it is systematic: A Major scale is pretty well-defined an invariable. However, much of music theory is context driven - for instance, A C Major chord is also well defined and invariable, but in one context it may have a role of being the most important harmony, and in another it may not, and in yet another it may be anachronistic, in yet another it may be astylistic and so on. I know of one internet poster who came from a math/science background and thought that learning music theory would help him compose music. He was very disappointed (and vocal about it) that musicians use the term "theory" so loosely. It's not lat Newtonian Theory or anything. It's like hoe Newtonian Theory might break down in another dimension - everything in music is context-driven - millions of dimensions so to speak.

Your point about speaking the same language is well taken. I started to read another article in these forums that looked interesting, but soon found myself facing terminology that I'm unfamiliar with (not to mention Roman Numerals! Yikes! What's THAT about?). I'll revisit that, as well, after I tackle some good old fashioned traditional theory.

Here's a freebie to elaborate on the point above:

We call C C.
We call C - am - F - G a chord progression.
We also call G - em - C - D a chord progression.
To the uninitiated they seem unrelated, other than the fact that the minor chord is second in both instances.

But when we compare these pieces, we can do things like abstract the letter names of the chords, and look at how they're related. The first, in the key of C is I - vi - IV - V. The second, in the key of G is, I - vi - IV - V - hey, what do you know. Then when we compare hundreds of pieces and find out they all share this same relationship, we can make an observation about them. When we notice that this chord progression occurs frequently in the years 1800-1900 (it doesn't necessarily, I'm just throwing out some numbers) and not prior to 1800 or after 1900, we can make yet another observation. It helps us to classify like and dislike pieces, and discuss their features.

So, here's the freebie:
Given a key, the chords are given Roman Numerals based on the scale degree on which they're built. For example, in C major, the NOTE C is 1.
C D E F G A B
1 2 3 4 5 6 7
And the CHORDS are:
I II III IV V VI VII

Just to let you know what you're getting yourself into, there are actually a bunch of different ways these numbers are used (could you imagine if the Periodic Table was Ag for silver in America, and Si in England!):

The most common system is that the quality (whether the chord is major, minor, diminished, etc.) is shown by the numeral - uppercase is Major, lowercase is minor. So in C major: they go:

I ii iii IV V vi vii^o

Some people don't make this distinction, and use all uppercase, expecting you to know the quality if you know the key - the only make a distinction if it's unusual. Still others who use UC, use IIm to mean minor (others use a figured bass symbol to show the alteration). Inversions of chords are also included with RN. Classical analysts use Arabic numerals from figured bass.
I⁶ is first inversion, etc.
Still others (especially it seems English posters on these boards) use a letter system: Ib means first inversion.

Usually the context is clear enough to determine what is meant (like Row (a boat) in American, and Row (an argument) in English - in addition to other meanings) though there are some cases where it can be very ambiguous (especially when conversing with a person who only knows the system they learned, and not the other systems).

Peace,
Steve

[chckn8r]

Originally Posted by stevel

Many people believe that they will be contaminated by learning theory. No. You can choose to use theory or not. However, not learning theory will never enable you to use theory. It's better to know it and choose to use it or not, than to not know it IMHO.

Very well stated! As Steve mentions in a later post, a lot of people look to theory as a "rulebook" and expect it to give them formulas to create music and when they find it doesn't, they try to shoe-horn creating music using theoretical rules and then get frustrated with the results. And to complicate things further, as Steve eluded to, there are more than a handful of flavors of music theory - different ways to symbolize, analyze and apply.

I liken the study and application of music theory to be less like a science or mathematic theory and more like the study of economics or psychology. They're fields of study that formulate models based upon observation of products of human nature and are not 100% effective when applied to the real world (case in point: economists have accurately predicted 27 of the last 14 recessions).

I like to use my theoretical knowledge as a starting point, but then go off the beaten path to find that "nugget" of a chord progression or modulation or voicing that wasn't really part of the "best practices" or rules, but perked my interest.

Again, as Steve stated above, knowledge is the key here. I applaud the gusto in which you've tackled this problem on your own and your desire to get your head into this some more by taking classes. There are a lot of paths to follow and you could spend a LOT of time getting into all of this once you do!

D

[reith]

An excellent piece of work there, fiziwig - applause for coming up with your own answers!

The problem with theory is that by its nature it concerns rules or attempts to formulate them - fine for analysis and a great help with composition...at least, knowing they're there but not necessarily attempting to apply them. From the compositional viewpoint they are best seen as guidlines toward good musical taste. It's often definace that leads to a new/original sound.

cheers,
reith

[ttw]

Note that rules in music are descriptive not prescriptive. They're more like taxonomy than like developmental biology.