QUOTE (njurkowski @ Jun 5 2008, 10:23 PM)

QUOTE
QUOTE
Why even bother with the key signature, when it is possible to just write the #/b where needed? It will sure clutter the script, so the key signature is simply a convinience tool, not having any theoretical or practical significance.
And why can't we just all learn to read MIDI notation, which is completely unambiguous. Because we don't want to play like computers. We want to play like human beings.
I do understand what he's saying, though. One of the habits we have to break first year theory students out of is over-reliance on the key signatures, for exactly the reasons that have come to light with this example (as well as modulations, etc.) The most important thing is for a musician to hear the function behind the key signature. Leading tones, dominant-tonic relationships, and so forth.
Fascinating examples here of "little learning being a dangerous thing"!
As a lingist, I see the confusion arising from a somewhat loose usage of the term "key signature". The arrangement of sharps or flats at the beginning of a score does help you to identify the key, but does not do so unambiguously. Even in "nomal" classical music, two sharps doesn't mean G major. It means G major OR E minor. My mother (a trained pianist who - significantly - could also accompany familiar tunes by ear) told be to look at the key signature and then at the last note in the bass line. Signature '##' + last bass note G = G major; signature '##' + last bass note E = E minor. And when you get into traditional Irish songs, with their (nowadays so-called) Dorian and Mixolydian modes, things get even more complex.
What the so-called "key signature" does define unambiguously is the scale the piece uses. With no sharps or flats, the notes A, B, C, D, E, F and G are all we need; with two sharps, we'll need only A, B, C#, D, E, F#, G. If we need any more - like when we modulate to a related key - we write them as accidentals.
So it would be better to call it a "scale signature".
There are practical reasons for putting this up front, especially for players of diatonic instruments, like Celtic harps, keyless flutes or 20-button anglos. It tells the harpist which strings to sharp, and it tells the 20k angloist whether he can manage the piece in this key or not. If there's a C# or a Bb, he has to pass.
But on chromatic instruments, like English and duet concertinas, it tells the player at what points he must take one of the outer buttons, which I'm sure is more comfortable than having no key signature and sharping or flatting the notes when they appear. A signature up front gives more structure.
For computers and purely mechanical musicians, this may suffice. But by-ear players and singers need a tonality to hold on to. It makes a helluva difference to them whether the two sharps mean G major or E minor.
There are two ways of distinguishing modes. One way is to take the major scale, and specify which of the notes are flatted. Flat the 7th, and its Mixolydian; Flat the 3rd and the 7th, and it's Dorian. This is how I as a singer do it.
Or you can take the major scale and change the tonality, i.e. "home in" your tune on a note other than the "doh" of the major scale. Take the 2nd step of the major scale as your tonic, and you're in Dorian; take the 4th step, and you're in Mixolydian. This is how I think of it when playing tin whistle or Anglo.
In short, the scale with one sharp is sufficient to play tunes in G major, A Dorian, D Mixolydian, E minor (or Aeolian) and a couple of other rarely used modes. Each mode, including the classical major and minor, has a completely different feel to it, even from a melodic pooint of view. When harmonisation is added, the feel gets even more different, because the chord sequences are leading to different destinations, and the available intervals are different.
For instance, with one sharp, you can't build an A major triad - so the mode beginning on A has to have a minor triad as its tonic chord. So Dorian is a "minor mode", though distinct from the classical minor scale.
Ironically, the naive, illiterate musician has little trouble with these modalities. If he's grown up with them, he handles them quite naturally, and if he hasn't grown up with them - well, they don't impinge on him.
The people who do have problems with modality are those who are not "native" to it, but have tried to approach it via sheet music. Sight reading is usually taught in the context of mainstream European classical music, so its application to ethnic or ancient musics is not dealt with.
Having learned the use of the alphabet in English, you're not going to be able to read French or German such that a Frenchman or German will understand you (and not laugh), although these languages use the same alphabet. Same with music - many musics use staff notation, but meaningful interpretation of the notation requires knowledge of the "vocabulary" and "grammar" (i.e. theory) of the music involved.
One thing I've noticed when teaching English and German as foreign languages to adults: people who have a good grasp of the grammar of their mother tongue can learn the structure of the foreign language much more easily than those who have not. The ideal pupils (of either language) are the ones who did Latin at school, and are "theoretically overqualified".
Make your musical life easier - don't just learn to "read misic", get a good grounding in musical theory!
Cheers,
John