Marking key signatures.

[gregh]

I'm looking at the arrangement of Scarborough Fair in Alfred's. (If anyone has noticed my progress, I haven't been moving along very quickly lately. My efforts have been diluted by many things.)

Anyway, it's in harmonic D minor. There's a B-flat in the key signature, and every single B is naturalled out, so really it's all white keys.

If the key signature was nothing but a short-hand for notes to play on black keys, the B-flat would have no business there. But I suppose it tells you that a D minor of some kind is coming up (unless it's an F major). It tells you something about the scale that is to be used, but it's only a clue, it's incomplete information.

What about a D blues scale? That doesn't have a B-flat, but it has both an A-flat and an A-natural. Would you ever see a key signature with a G-sharp to indicate D blues? What about the key of D in other modes, like dorian? Would it be two sharps and accidentals as needed? Or a C whole-tone scale--that would never be indicated by an F-sharp and a B-flat in the key signature, would it?

I guess I'm hoping that people have opinions they want to share about the use of key signatures to indicate the notes that are coming up.

[iansinclair]

Quite true -- as soon as you get away from main stream western classical and popular music, the key signature can sort of make one wonder why bother.  I like to think of it, in those cases, not so much an indication of which notes should have certain flats or sharps, but via convention an indication of what the tonic (and hence entire harmonic structure) of the piece is.  Thus if the key signature is D minor, then we know that the tonic is D, and the overall structure of the scale will most likely have a diminished first third.

As soon as one moves away from main stream western classical and popular music, the question comes up again and again -- and the choice of the exact scale divisions and the harmonic structure goes a long way to setting the piece in a proper context (think Celtic music, for instance -- a very distinct scale and harmonic structure, not like main stream at all).

[Bob]

There's no harmonic d minor as a key.  It would just be d minor.  Harmonic minor is more for thinking about chords (harmonies).  In terms of scales, it's more like ascending minor and descending minor really.

But Scarborough Fair is modal.  Dorian.  D Dorian for this piece.  I cheated and looked it up. I don't have the music in front of my or the ears to notate it. 

The key signatures are really just pointing at the tonal center.  It is kind of interesting that they wrote it that way.  I heard one reason for key signatures was to save ink costs for printers.  Doesn't sound like it would save anything in this case.  I bet there are other arrangements without a key signature or possibly the 'wrong' key signature.  No need alter anything that way.  It would be obvious from the bass, although using a key signature does make it easier to understand.

[gregh]

Dorian? I'm not sure I'd recognize that. But, looking at the chords, I think Alfred's arrangement, at least, really is D minor.

I'd been fascinated by other modes, and still am. But then I thought of what is done with good ol' major and minor. It's not as simple as happy and sad, you know. "Stormy Weather" is major, for instance, and sounds like a wailing lament. And when the swingers get their hands on minor it sounds positively naughty. So I thought that the mode probably wouldn't do as much for a song as the composer just applying his artistic sense.

If I ever put out music that I had composed or arranged (this is a little like "When I'm rich and famous...") I suppose I would use the key signature to define the tonal center (or the tonal center for the relative major, as seems to be the custom for minor) and explicitly write out what I'm working with, like "D Dorian" or "F Blues Major", for instance. And putting both F-sharp and B-flat up there for C whole tone is right out--not that there isn't some sense to it, but it would just confuse or irritate people.

[mjedwards]

     I think people use key signatures differently in regards to modal influences like this.  To my ear, "Scarborough Fair" is definitely modal, and definitely Dorian.  If there is an arrangement that makes it sound otherwise, it strikes me that it must be a rather odd arrangement, and it might have to introduce harmonic B-flats to contradict the B-naturals in the melody, which is unambiguously Dorian.
     Personally, in arranging or composing a modal piece (or section or theme within a piece), I would choose the key signature that eliminates accidentals, not the one that would be suggested by the plain major or minor equivalent of the modal scale.  So no sharps or flats for D Dorian - 3 sharps for E Mixolydian - 5 flats for C Locrian - and so on.  I wouldn't want to falsely suggest D minor (for example) when in fact it wasn't that, but D Dorian (Phrygian, or anything else).
     I greatly support literalness and truth in notation, so I guess that's where I come from.  Considering that most of the modes have one or more notes altered from the corresponding major or minor scales, altering the key signatures to reflect that would actually be correct for those particular modes.
     Also, composers sometimes choose wrong key signatures for even traditionally tonal passages - even great composers do.  For example, in Beethoven's Piano Sonata no. 31 in Ab major, Op. 110, the first "Arioso" section in the last movement is in Ab minor, but incorrectly given a key signature of 6 flats - but the correct one should be 7 flats.  (Of course all the actual notes are correct, and the many F-flats which occur are always given the required accidental.)
     (Beethoven also makes serious time-signature errors (at least according to current conventions about time signatures) in his last Piano Sonata, no. 32 in C minor, Op. 111.  In the Arietta with variations which forms the second (final) movement, Variations 2 and 3 are notated respectively in 6/16 time and 12/32 time.  These are both incorrect - in both variations a simple 3/8 is the correct time signature - assuming that the fastest notes are considered to be triplet groups but with the triplet brackets not actually marked in.  (The 9/16 used for the remaining parts of that movement are correct, though.)  Perhaps the now-quite-clear rules governing the choice of time signature had not yet quite solidified at the time Beethoven wrote the sonata.)

Regards, Michael.

[mjedwards]

And putting both F-sharp and B-flat up there for C whole tone is right out--not that there isn't some sense to it, but it would just confuse or irritate people.

     Apart from the fact that F# and Bb together don't define the whole-tone scale, but the rather different scale variously known as the "Overtone" or the "Lydian Dominant" scale, doing this actually isn't right out of court for all composers.  Composers have on occasion used irregular key signatures to allow the notation of synthetic or irregular scales without accidentals.
     Bartok did it quite a few times - and you might like to check out the astonishing series of five piano sonatas by U.S. composer Andrew Violette, available on imslp.org in the composer's manuscripts, which are very readable by the usual standards of composer manuscripts.  There are numerous passages in those sonatas where Violette is using odd scales, and he uses many different odd key signatures for notating these.
     Violette's works are well worth exploring - I just found them weeks ago - and remind me of Sorabji in places: in complexity, original concepts and harmonies, and in extreme length, too - fascinating stuff.  (Oddly, I hear the ghost of Beethoven occasionally in them, too.)  He has posted many of his scores on imslp.org recently; and you can also hear them all on Andrew Violette's own YouTube channel, and follow with the score as it is displayed to keep pace with the performance (by the composer himself - an amazingly precise and virtuosic pianist).)
     Somewhere in my massive piles of sheet music I have an old copy of a piano sonata by someone called Alexander Jemnitz - and the entire first movement is written with a key signature of just G-sharp - nothing else.  Every G in the piece is sharp, all other notes are natural, and there is not a single accidental in the piece.  The movement is not in A minor, the chords are not ordinary triads, and the harmony is very strange and original - but entirely within that scale which would correspond to the harmonic A-minor scale.
     Some may find this sort of thing with unorthodox key signatures too disorienting, and prefer to use no key signature, but mark all black notes with accidentals - but odd key signatures are certainly something composers also do - and one I would at least consider doing in some situations myself.

Regards, Michael.

[gregh]

     I think people use key signatures differently in regards to modal influences like this.  To my ear, "Scarborough Fair" is definitely modal, and definitely Dorian.

It is. I had to think about it, but you're right. I was looking at the arrangement in Alfred's beginner's book where he introduces D minor and harmonic minor. D minor natural, of course, has a B-flat, while the harmonic minor has a raised seventh, meaning both a B-flat and a C-sharp. Scarborough Fair, though, has C natural, and every B-flat is turned to a natural. I was used to doing that for C minor harmonic, so I was off by a step. Alfred introduced the natural and harmonic minors, and then presented something in Dorian.