Very interesting from you both !

I have a Romanian 'Elements of Tonal Harmony' book (I think 1st year university level at a private univ., not the State Conservatory; meant it was easier than Conservatory) that has lots of exercises like that. One line plus a few numbers here and there (esp. the some 6's; other 5's are just assumed) to produce a full 4-voice chorus-like sound.

Only that it doesn't have explicit the principles behind it. Guess that's where the seminar instructor closely works with students showing them more hands-on examples ?

Deleted - I forgot to transcribe some of the accidentals. Corrected version coming shortly. (sorry )

(link works again)
Fourth try - with corrections pointed out by Harmosis, and worked on measures 5 - 8.




KS

So much better! Only a couple things to correct here:

1] The viiº6 chord, in m6 beat 1, needs to be resolved correctly - the E in the alto should resolve down to D (it can still move to E in the next beat without any problems).

2] The ii chord, in m7 beat 1, is in root position but with a doubled 5th. Make the tenor a C# instead of G# (double the root). Move the bass to F# above, instead of down, in the next beat to avoid parallel octaves (tenor and bass will be at the same pitch on beat 2). You can make the voice-leading even a little smoother by keeping the alto on E, making the V chord a V7 (this is optional, though).

Good job

Thank you once again!
For:

I would then want to make the tenor in that chord an F#. Otherwise the chord would sound too thin having only D's and B's. Is there any problem doing that?

No, the F# (5th) is not needed. It's better to have the doubled root instead of the inclusion of the 5th. You could go to F# in the tenor, as you propose, but the strongest resolution, resolving the tendency tones, is the way I described. There are times when it's better for an inner voice to resolve from the leading tone down to the 5th, but this is not one of them.

I think I understand. The B in the tenor is part of a resolution in that A# goes to B in the tenor. If I use that F# as I first proposed, I lose that resolution of the tendency tone. Do I have that correct?

I think I understand the meaning of tendency tone.

Yes, that's right.

Final version? And one question if I may:
I researched tendency tones this morning and learned about the tendency tones of 4=>3, 6=>5 and of course 7=>8. It's not totally new since "fa-mi" goes that way in solfege, and the minor harmonic has the tight semitone between ^6 and ^5 which I would also sing as "fa-mi" (tonic as "la").

The ^4^3 in major scales has an interval of a semitone, as do ^7^8, but in minor scales there is a whole tone interval. I'm not sure that I feel as much of a pull. Same for 6^5^ in major scales (whole tone).

So is there a difference? Is the idea of tendency tone more "forgiving" among the degrees of scales where the two notes are separated by a whole tone? And then I wondered for minor scales if you don't also have a pull from ^2 into ^3 ascending?

I must admit that the term was new to me, although I knew it (without the term) for the leading note. That was a big hole in my knowledge. Many thanks once again.

It looks good but you should resolve the A# in the tenor (m6, beat 1) to B in order to resolve the tritone, A# to E, properly. Again, there are cases where you want to move from the leading tone to the 5th but this is not one of them.

As for tendency tones, yes, the whole step does not pull as much as the half step. However, voice leading considerations (i.e., resolving a tritone) are still present.

Thank you.

You're welcome

Harmosis, I'd like to tie up some loose ends, since my purpose is not just to produce the right answer, but learn the principles behind what I'm doing.

In your last correction you proposed B in the tenor (m. 7) in order to resolve the tritone. I could see the tritone, which is tension, that gets resolved. I would have thought the B must be placed there because the preceding A# is a leading note which must be followed by the tonic that it "leads into". You will tell me that both views apply: in voice leading we have the leading note moving into the tonic, and in harmony we have a dissonance that must be resolved. The two work together simultaneously. correct? I can resolve the dissonance simply through a consonant chord, but the choice of B involves the leading note. Similarly ^4 leads into ^3, and ^6 leads into ^5; and ^1, ^5 and ^3 are "strong" degrees. I think that I have a good understanding of that part now (hopefully).

I'd like to double check about this part since I'm applying what I've learned:

Is the tritone an example taken from the present exercise, or do tendency tones always involve tritones? ^7 (if maj7) always does, of course. But we also have ^4 and ^6. I want to make sure that I understand what you have said correctly.

In regards to suggestions you made a while back, I've begun studying the Riemenschneider chorales which I happen to own. I will be looking for the types of things I have learned in this thread while doing the rest of the chapter. the next chapter involves passing notes, which I imagine rests on the foundation of the present work and I suspect would go well as long as this one is mastered first.

Many thanks as always.

KS

Yes, you have to keep in mind the melodic aspect of each voice, and the harmonic implications thereof (since harmony is really multiple melodies occuring simultaneously).


No, but tritones, as dissonances, are generally resolved a certain way: Resolve diminished 5ths "inward," and resolve augmented 4ths "outward" (one is just the inversion of the other). Part of the reason for this is to create contrary motion, which gives the voices more independence (melodic reason), and gives a more satisfying resolution. Also, harmonically, you wouldn't want a dissonce "resolving" to another dissonance (i.e., an aug 4th moving to a perfect 4th, like E-A# moving to F#-B).The more voices you have to work with, the more you can go around some rules (which is why there are situations where you can resolve the tritone in similar motion). But if you can follow these rules, then you should - it really will sound better.

Thank you, this is very helpful. I have done 6 questions like this altogether including this one. I will go through them with a fine tooth comb applying what I have learned here.

Yes - both vertically and horizontally. I have been playing and listening as I went along. It really does make a difference.

It always seem strange that a P4 should be dissonant since it sounds so ... er .... harmoniously pleasant.

Right - in organum, it's certainly not a dissonance!

This is a new one that I had a bit of a struggle with so I'm hoping to double check again. I'm hoping to have learned something from the last time in doing this one.

The ending of the proposed melody seems odd: does this kind of ending happen in music of any particular kind?

Edit: The tenor looks weird now in bar 7: E D# / E ?
(I kept getting tritones or parallel octaves and was ready to throw in the towel last night)

[edit: there is an error in this one which I've fixed in a post below it. The error is in the A# in m. which I transcribed wrongly from my rough copy. It should be G#)

I'm pretty sure the melody is incorrect at the end. Either that, or the harmony that's specified is incorrect. The A# suggests a secondary dominant, but none occurs, so either it's wrong or incomplete.

Kreisler, the exercise consists of solely the melody, and I am to provide the harmony. Would you suggest that I skip this particular 8-bar exercise, in the sense of it being a poorly crafted exercise?

Having said that - is there anything incorrect in what I've done? I am thinking of changing the tenor in bars 7 to 8 to E D# E because as voice leading what I've done doesn't work (I think).

Oops, I made an error in transcribing. That A# in the melody in bar 8 should have been a G#. I've also changed the tenor in bars 7 & 8 to what I wrote before. The original is in (brackets)

Hi keystring,
You've got several problems here:


1] Tripled 5th, no root (m5, beat 2) - the only note you should ever triple is the root, and then only in cadential situations. Also, you need to have a root here.

2] Parallel octaves (m5, beat 2 to m6, beat 1) between the bass and alto

3] Spacing (m6, beats 1 & 2) between the alto and soprano

4] Augmented 2nd (m7, beat 1 to 2) in the alto

5] Doubled leading tone (m7, beat 2) in the tenor and alto - never double a sensitive tone or dissonant tone (like the leading tone or the 7th of a chord).

You've got the right idea with the chords you're going for, but you need to clean up the voice leading. Also, watch your doubling - you should be looking to double the root in 1st inversion chords.

Harmosis,

Using the melody provided, can you do something with those last four bars that breaks no "rules" but also stays within the limitation of no 2nd inversion chords, etc.? I absolutely suck at this kind of voice leading, but I could not come up with anything that worked myself. After KS is through with this, I would be curious to see a workable solution.

I think the melody is a good exercise, it just has some tricky spots. I'll play with it and see what I can come up with. I'd be interested in seeing Harmosis' version, too.

Kreisler, if you're duplicating my task, I am still limited to root position and first inversions. I understand that it gets easier when there are more choices. I would be very interested in seeing what the experts come up with and think I would learn something by studying your approaches.

Of course I will have another go at it myself. I seem to get hung up every time the melody has a melodic minor. I was constantly getting parallel octaves and fifths or tritones. Every time I managed to fix one part, the other part collapsed. Reminded me of an old tent we used to have - the minute the fourth leg was up, the other three would collapse. The answer was to buy another tent.

I appreciate the help.

Okay, here's my version. The first is the bass line - I always do bass lines first. A good bass line just makes the rest easier. Next, I fill in the inner voices. You're right, the last 4 measures are tricky - it's always a bit of an issue when you have a scale in the melody and you're limited in the inversions you can use.

I think what I have is legal and sounds okay. (We'll let Harmoisis check my work to make sure I didn't miss anything.)

Only root position and 1st inversion? That's pretty restrictive but I'll do one.

Kreisler: Yours looks good except the doubling on the iiº in m3, beat 1. The root and 5th are dissonant to each other, so double the 3rd here (Also, your iiø in the next beat breaks the inversion restriction! .) Another thing just to be aware of is the direct 5ths between the alto and soprano in m1 beat 2 - but since the harmony is NOT changing, this is not a big deal.

I'll get mine posted as soon as I get it done.

Yes, I haven't been given a lot of tools yet, so those are my limitations.

It may take me a few days: I'm a freelancer and something large has come up taking every free moment. I'll post and do mine when I can. But I'll keep popping in briefly to look.

Harmosis, what does this symbol mean? iiø (the part after ii)?

KS

OK, here's mine. I also broke an inversion restriction with V42/iv (it's ridiculous be restricted to 6/3 chords without being able to use 4/2 chords!).

The ø sign means half-diminished.

Ah! I just found parallel octaves in mine (tenor & soprano m3, beat 2 to m4, beat 1)! This is what these inversion restrictions lead to!