Theory Experts, Liszt Nocturne No3 - Progression?

[steve jones]

Hi all,

Just been having a listen to the wonderful 3rd Nocturn by Liszt. Love this piece! But something occured to me just that I dont fully understand.

The progression of chords in the main theme (in the opening key) proceeds with a series of dominant chords. I make it:

Ab Major, C Dominant, F Dominant, Bb, Dominant, Eb, Dominant, Ab Major


The Ab is clearly the tonic, with the Eb being the diatonic dominant (resulting in the authentic cadence). But I dont get how the progression of C, F and Bb dominants can be rationalised. The C maybe could be considered a chromatic mediant. And the Bb could be a secondary dominant, tonicizing the Eb. But where does the F dom come from???

SJ

[allchopin]

The progression appears to be a simple walk around the circle of fifths, beginning on C.  C is the dominant of F, which in turn is dominant of Bb, then -> Eb - > Ab.  If you attached a snippet of the music that might be helpful too.

[steve jones]

What Im unsure about is the justification of the dominant chords (or substitutes) in each of these chords. From what Iv been taught, the dominant chord is to be used functionally yet I dont see the function here (even though it does sound sweet!).

https://www.yousendit.com/transfer.php?action=download&ufid=4C6DB314249288BC

SJ

[steve jones]

I mean, root wise, you have something pretty simple:

I, III, VI, II, V, I

But ofcourse, in the diatonic system it would be:

I, iii, vi, ii, V, I


There would be ways of explaining these chords, if they were resolving properly, by interpreting them as secondary dominants:

I, V/vi, V/ii, V/V, V, I

However, my knowledge of chromatic harmony isnt good enough to judge whether or not this is correct. I was under the impression that secondary dominants had to resolve to diatonic key chords (be it in the home key or a modulation)?

SJ

[jlh]

I don't have the score, so could you please tell me why you say the chords are dominant?  Just because something is major doesn't make it dominant.

From what you're saying so far, yes, it is a circle of 5ths progression.  The chords may also be borrowed from the parallel minor key, so that iii becomes III, vi becomes VI and ii becomes II.

[jlh]

OH!! This is the Liebestraum Nocturne No3!  Yes I have this one...

Ok, here's what's happening:

I - V7/vi (borrowed from III/i) - V7/ii (borrowed from VI/i) - V7/V - V7 - I

The reason this works is because Liszt is borrowing from the parallel minor in order to keep the sound major while also descending chromatically.

[steve jones]

Hmmm, wouldnt III/i be Cb Major?

Again, I can account for the chords by way of secondary dominants. Its the fact that they are resolving incorrectly that is confusing me. I had considered the idea that they might be resolving to borrowed chords, but that does seem to fit as the roots dont match - III/i would be Cb I believe.

Is it possible that the C7 is infact a chromatic mediant, with the F and Bb being secondary dominants?

Thanks for the input btw! Appreciate the help 

SJ

[jlh]

Hmmm, wouldnt III/i be Cb Major?

Again, I can account for the chords by way of secondary dominants. Its the fact that they are resolving incorrectly that is confusing me. I had considered the idea that they might be resolving to borrowed chords, but that does seem to fit as the roots dont match - III/i would be Cb I believe.

Is it possible that the C7 is infact a chromatic mediant, with the F and Bb being secondary dominants?

Thanks for the input btw! Appreciate the help 

SJ

You're right, that could be a chromatic mediant.  Then again, you can analyze it as #III, meaning that the root is raised (I meant to make that clear, but forgot the #III). 

The sec doms do resolve correctly if you borrow the resolutions:

The V7/vi resolves to V7/ii (VI/i) and the V7/ii resolves to V7/V (II/i)

[steve jones]

I wonder if Liszt considered this when he penned the piece? From what I gather, he was a big forward thinker and major contributor towards the eventual breakdown of tonality. So is it possible that he simply disregarded the traditional 'rules' and did what sounded good? Or do you think there was method in this progression?

SJ

[pianiststrongbad]

I think Liszt and Chopin and many others played what sounds good.  That doesn't mean they don't know what tonal harmony progressions tradionally are.  I have seen in some Mozart sonatas the progression goes V-IV-I.  Not exactly typical for that type of music.

[steve jones]

Yeah, when you look at alot of Chopin's music you see alot of chromatic harmony, yet it sounds so free and hardly contrived. Take his prelude No9 for example, as its a classic case.

SJ

[cjp_piano]

Everyone writes what they think sounds good don't they?  Music theory tries to explain WHY it sounds good. 

For the Liszt:

The progression IS a chain of secondary dominants.   Although it's true that secondary dominants TYPICALLY resolve to their obvious tonics, it is also common that the resolution is actually the next secondary dominant (the root staying the same but the chord being major and likely having the 7th).

So I would explain it just as you did:

I   V7/vi   V7/ii    V7/V   V7   I

Even though V7/vi would resolve to vi which is F minor, it actually goes to F major/dominant, so it's a V7/ii,  etc. 


C7 (V7/vi) -------- F minor (vi)
                           F7 (V7/ii)--------- Bb minor (ii)
                                                     Bb7 (V7/V)---------- Eb (V)

The secondary dominants are substituted for the regular resolutions.
                                                                      

[allchopin]

Again, I can account for the chords by way of secondary dominants. Its the fact that they are resolving incorrectly that is confusing me. I had considered the idea that they might be resolving to borrowed chords, but that does seem to fit as the roots dont match - III/i would be Cb I believe.

Is it possible that the C7 is infact a chromatic mediant, with the F and Bb being secondary dominants?
SJ

What Liszt is doing is actually quite simple (as are most of his progressions, actually) - I wouldn't try to over-complicate things.  If it were any other composer, such as Bach or Schoenberg, you could go into a more in-depth analysis, but Liszt all in all is a fairly straight forward, in-your-face type of composer.  The progression used actually is a staple in Vivaldi, where the chords walk through the circle of 5ths until rearriving back on tonic.  But instead of using diatonic chords in the progression, like Vivaldi would (I->ii->V->I6->IV->ii->V->I) he is using secondary dominants, which almost resolve, but leave you hanging on the next chord, because it too becomes a dominant.  In the strictest sense of how theory 'works', the progression shouldn't really be 'allowed' but Liszt pretty much violated most codes of conduct of his day, did he not?

[steve jones]

Excellent, thanks all for the assistence!

SJ

[allchopin]

Everyone writes what they think sounds good don't they?  Music theory tries to explain WHY it sounds good. 

For the Liszt:

The progression IS a chain of secondary dominants.   Although it's true that secondary dominants TYPICALLY resolve to their obvious tonics, it is also common that the resolution is actually the next secondary dominant (the root staying the same but the chord being major and likely having the 7th).

So I would explain it just as you did:

I   V7/vi   V7/ii    V7/V   V7   I

Even though V7/vi would resolve to vi which is F minor, it actually goes to F major/dominant, so it's a V7/ii,  etc. 


C7 (V7/vi) -------- F minor (vi)
                           F7 (V7/ii)--------- Bb minor (ii)
                                                     Bb7 (V7/V)---------- Eb (V)

The secondary dominants are substituted for the regular resolutions.
                                                                      

You stole my words cjp - it's this darn lag in the forum.

[jlh]

Everyone writes what they think sounds good don't they?  Music theory tries to explain WHY it sounds good. 

For the Liszt:

The progression IS a chain of secondary dominants.   Although it's true that secondary dominants TYPICALLY resolve to their obvious tonics, it is also common that the resolution is actually the next secondary dominant (the root staying the same but the chord being major and likely having the 7th).

So I would explain it just as you did:

I   V7/vi   V7/ii    V7/V   V7   I

Even though V7/vi would resolve to vi which is F minor, it actually goes to F major/dominant, so it's a V7/ii,  etc. 


C7 (V7/vi) -------- F minor (vi)
                           F7 (V7/ii)--------- Bb minor (ii)
                                                     Bb7 (V7/V)---------- Eb (V)

The secondary dominants are substituted for the regular resolutions.
                                                                      

Which is basically what my post attempted to explain.  The major chords the sec doms resolve to are actually borrowed from the minor.

[steve jones]

Indeed, but I dont see that these chords are being borrowed from the relative minor mode. Both the mediant and submediants are flat in the relative minor.

I interpret this as a series of short tonicizations leading upto the authentic cadence. Each secondary dominant appears to be resolving to the next, probably as a way 'pushing' the phrase along. Its an interesting technique, as I didnt know you could do that and retain somekind of tonal bearing.

But secondary dom's are not borrowed from the relative minor, I dont think. They are borrowed from the tonicized key, which obviously changes as each chord progresses. For example, V7/ii borrows the dominant (F) from the ii key (Bb minor) in order to reinforce the Bbm tonic (i/ii). This is where I got confused, as I did not know that is was acceptable to resolve a seconardy dom to another secondary dom.

SJ

[jlh]

Indeed, but I dont see that these chords are being borrowed from the relative minor mode. Both the mediant and submediants are flat in the relative minor.

I interpret this as a series of short tonicizations leading upto the authentic cadence. Each secondary dominant appears to be resolving to the next, probably as a way 'pushing' the phrase along. Its an interesting technique, as I didnt know you could do that and retain somekind of tonal bearing.

But secondary dom's are not borrowed from the relative minor, I dont think. They are borrowed from the tonicized key, which obviously changes as each chord progresses. For example, V7/ii borrows the dominant (F) from the ii key (Bb minor) in order to reinforce the Bbm tonic (i/ii). This is where I got confused, as I did not know that is was acceptable to resolve a seconardy dom to another secondary dom.

SJ

It's not the relative minor, but the parallel minor from which you borrow chords.   

I was attempting to analyze this without actually modulating the key for every chord, which I'm not sure that it does.  Using the borrowed chord theory, I tried to give justification for the quality of the chords in their progression from tonic and back.  Yes, the chords change, and yes the sec doms resolve to other sec doms, but does the piece really modulate the key for every chord change?  The key is still Ab...

For the question of the mediant being flat in the parallel minor (not the relative minor), yes it's flat, but you can also # the mediant or submediant to say that the root is raised a semitone.

[steve jones]

Sorry, the parallel minor - C, Cm.

Yeah, I do see the thinking behind your analysis. As usual, there is more than one way to skin a cat - quite often passages can be analysed numerous ways and still be correct.

Personally, I prefer the secondary dominant method, as it adequately decribes the passage without over complicating things - in the progression of 5ths, each chord is clearly resolving the previous without need for altering roots.

But yeah, Im with you now, and your analysis does make perfect sense.

SJ

[cjp_piano]

Yeah, I don't like the borrowed idea, but that's just me.  If you raise the root of the mediant or submediant, they become dimished chords, not major.  There's too much altering for me.  F A C Eb isn't from Ab minor, it's from Bb minor.  It's V7/ii 

[jlh]

Yeah, I don't like the borrowed idea, but that's just me.  If you raise the root of the mediant or submediant, they become dimished chords, not major. 

The conventional roman numeral identification of borrowed chords indicates the scale degree upon which the chord is built and the quality of the chord.  For example, If the chord were a C major triad, borrowed from the parallel minor key of Ab minor, the roman numeral would be #III.  Here, the three means Cb, the sharp means that the root is raised (C), and the upper case numeral means a major triad.

That said, you have to be careful when analyzing a piece to be sure the chords are not actually secondary dominants, which they are.  So I have to ask again -- if a secondart dominant resolves to another secondary dominant and you're not analyzing that the resolutions are borrowed in origin, wouldn't you also have to be modulating on every chord?  Meaning... the resolutions of the secondary dominant resolutions are not diatonic to Ab Major.

Personally, I prefer the secondary dominant method, as it adequately decribes the passage without over complicating things - in the progression of 5ths, each chord is clearly resolving the previous without need for altering roots.

Both "methods" discussed here make use of the same secondary dominants, so to say that one method is the "borrowed chord method" and the other is the "secondary dominant method" is an incorrect conclusion.  The method is the same, differing only in the origin of the analysis.  That's why I mentioned that what cjp-piano posted was in essence the same as my post.

Anyone out there know how to clarify this issue?

[steve jones]

if a secondart dominant resolves to another secondary dominant and you're not analyzing that the resolutions are borrowed in origin, wouldn't you also have to be modulating on every chord?  Meaning... the resolutions of the secondary dominant resolutions are not diatonic to Ab Major.

I believe the term is 'tonicization', meaning to emphasize a chord by means of its dominant (or sustitute).

The progression here works, as the C7 tonicizes to F:

C,    C7,    F,   G7,  C

I     V/IV   IV    V     I


This Im sure you already understand fully.

What confused me about the Liszt progression was the fact that he appeared to be resolving not a diatonic chord but the next secondary dominant following the progression of fifths!

In diatonic music, this would not be allowed Im sure. But as Liszt was a driving force in the move to chromatic harmony and eventually atonality, Im guessing he thought it would be cool to delay resolution while exloiting chromatisism in the process.

Whether you interpret these as brief modulations or not is subject. Personally, I dont. I prefer to see this as Liszt's effective use of chromatic harmony, where the bounds of the diatonic system need not apply. Here he is free to progress to none diatonic chords and back again via these 'semi resolving' secondary dominants. What makes this even more engenius is the use of the descending 5th roots, which means that the progression never seems to lose its tonal bearings (ie, it is moving towards an eventual resolution) in the perceptive sense.

Atleast thats I interpret this passage.

SJ

[jlh]

Works for me...   (yes it's tonicization)

I guess I've just studied 20th century theory so much that I forget that Liszt also introduced some genius in the 19th century.  I think I also tend to over-complicate things and hate admitting I'm wrong...

Great discussion though!  I had to use muscles that have not been used in a few weeks [i.e. you made me think]. 

Have a good one!
Josh

[steve jones]

Hmmm, 20th century theory.... 

Didnt every 20th century composer design their own theory, write a book on it, then decide it was crap and revert back to romanticism? 

SJ

[jlh]

 

[cjp_piano]

Ha!  I like your thought about 20th century theory  

Just to affirm the secondary dominant thing, here's a quote from my harmony book:

"Secondary dominants are often used in harmonic sequences.  The most common progression is a circle of 5ths sequence in which secondary dominants are substituted for diatonic chords.  Often, the whole progression is altered chromatically so that each chord becomes the dominant of the next."

Example from Chopin's Mazurka (op 33 no. 3):

E7         A7        D7         G7       C  . . which is analyzed as
V7/vi     V7/ii      V7/V      V7       I

There is another example in Mendelssohn's Song without Words (op 102 no 3 - tarantella), and others from Schumann and Schubert.  

So it wasn't just Liszt!

[jlh]

I wish I wouldn't have sold my harmony textbook 3 years ago...

[steve jones]

I have the AB Harmony book, and to be honest, its alright but you could pretty much condense it into a much shorter text.

Personally, I believe Iv learned far more by analysing pieces as I learn (to play) them. Text books are good for giving you the facts, but unfortunately they arent able to give you natural instincts to recognise the techniques, let alone use them musically.

Tell you what though, I had a look at some Rachmaninov the other day and it seemed very complicated. More so than the typical Chopin, Liszt etc. He is constantly pulling off modulations and it can get confusing!

Ironically, my text book doesnt feature much by the big man 


cjp_piano,

Quality, cheers for that.

Yeah, when you think about it, that progression does sound pretty familiar in the romantic rep. Id just never considered it before in all its glory. Kind of like those old jazz guys who play the most insane progressions, yet have no clue why they sound good. Im beginning to think that is probably the best way to do business - concentrate on what sounds good and leave the 'whys' to the theorists!

SJ

[cjp_piano]

Yeah, when you think about it, that progression does sound pretty familiar in the romantic rep. Id just never considered it before in all its glory. Kind of like those old jazz guys who play the most insane progressions, yet have no clue why they sound good. Im beginning to think that is probably the best way to do business - concentrate on what sounds good and leave the 'whys' to the theorists!

SJ

Exactly.  Unfortunately, I'm cursed with a theorist's mind.  I always have to figure out why pieces work and I get frustrated if I can't.   

[steve jones]

Haha, yes, I fear we were cut from the same mold!

SJ