On Consecutive 5ths and 8ths

‹-- PreviousNext --›

One of those penny-drop moments came to me when, as an undergraduate playing through some music I had written as a younger teenager (and finding it both better and worse than I remembered - does this happen to everyone when revisiting the efforts of their younger selves?), I came across a bit I'd always had to play quite carefully to make it sound okay. It could work all right, but if you didn't place it just so, it could sound a bit naff.

In the time between writing this music and revisiting it, I had been taught the concept of consecutive 5ths and 8ths, and the importance of avoiding them. This concept now revealed to me what the problem had been with my piece of juvenilia, and simultaneously made me grasp, emotionally, why I should care about this bit of theory.

There are other schools of thought on consecutives of course. Take Noel Coward:

[My instructor] went on to explain that a gentleman called Ebeneezer Prout had announced many years ago that consecutive fifths were wrong and must in no circumstances be employed. At that time Ebeneezer Prout was merely a name to me (as a matter of fact he still is, and a very funny one at that) and I was unimpressed by his Victorian dicta. I argued back that Debussy and Ravel used consecutive fifths like mad. My instructor waved aside this triviality with a pudgy hand, and I left his presence for ever with the parting shot that what was good enough for Debussy and Ravel was good enough for me.

One of the best explanations I have read of why traditional harmony has an issue with consecutives is this on from Graham Redwood, based on the acoustic closeness of notes an octave or fifth apart:

It stands to reason that, if two voices are singing an octave or a fifth in a particular chord, they momentarily "blend in" to one another and their independence is threatened. That in itself is not a problem, but if they move to another octave or fifth in the next chord, the ear hears a progression of only three voices rather than four, since the two voices an octave or fifth apart are heard as one. If this happens indiscriminately throughout a piece, the effect is haphazard and inconsistent.

The reason I like it is because it also explains why sometimes consecutives can work just fine. Which I will come onto in a moment.

Now, close-harmony arranging has tended not to be quite so punctilious about this as classical harmony, because the focus is much more on the sonority of the chords than on the independence of the lines (as I discuss in my deconstruction of barbershop harmonic theory in Chapter 3 of my first book).

Having said that, in practice, it's consecutives between bass and baritone that are most forgiven, as the baritone's role in the texture is primarily sonic fill. Those between lead and bass do tend to sound naff, as those are the two lines that have the strongest linear role. But in general, if I'm arranging in a mode that is (a) homophonic and (b) harmonically rich, I will generally be pretty rigorous about avoiding consecutives, because it just produces a more coherent and consistent texture to my ears.

(To what extent my ears are over-sensitised by my education is something I don't have the perspective to guess. But I reckon that even if nobody else was going to be distracted by the consecutives, neither are they going to be distracted by their absence, so I may as well err on the safe side here.)

The situation becomes less clear-cut when you're arranging music that originates in the primarily triadic world of rock. You get a few four-note chords colouring rock harmonies, but you also get a lot of music built round pure I, IV and V, with colour coming from instrumental textures rather than extending chords. Harmonic variety is achieved by root movement - surprising us by the juxtaposition of unexpected triads.

In this context, consecutives could arguably be seen as a fair representation of that harmonic world. If consecutives result in the ear conflating two sounding parts into one logical one, as in Redwood's argument, then that's an appropriate thing to happen when rendering a triadic chord vocabulary onto four parts. Indeed, any chord sequence that involved a guitar shift that bridges with the forefinger whilst moving the whole chord shape up or down builds consecutives right in.

At this point, I should say that even if you take this argument as a carte blanche to fill a cappella arrangements of rock tunes with consecutives, do note that they can still sound naff between lead and bass. But if you're building your bassline from the original, you're unlikely to find too many of them anyway - it doesn't sound great on a rock band either.

But I still hesitate to use them, even once I have given myself this stylistic permission. And I think the issue is that, while consecutives are very authentic to some guitar-based idioms, they are still at least somewhat problematic on voices. We need to respect both source and adapted idioms if an arrangement is to work.

The problem I suspect consecutives of present to a cappella voices is that of tuning. I have often observed that simple, triadic songs, that in theory should be easier to sing than chromatic, harmonically rich pieces, can actually come out in performance a bit wobbly. There's less wriggle-room in a triad than a four-note chord; the dazzle of added notes can hide any momentary imperfections in the lock of root and 5th, and opens up more options as to where exactly the 3rd can sit.

And if a triad needs a certain amount of care to tune, it is rather asking for trouble to ask singers to move in parallel exactly the same distance. It's not like a guitarist just sliding the same handshape along the fingerboard.

So, in practice, what do you do instead?


In the example (a) is the obvious way to harmonise a IV-V progression. But it includes both consecutive 5ths and consecutive octaves. Examples (b) through (h) are possible alternatives (one of which I have used recently when faced with this progression). They have different advantages and drawbacks, and which I would choose in any particular circumstance would depend on the context (shape of lyric, level of emotional intensity at that point in the song, who was going to sing it).

(b) is cheeky in that it changes the melody note, but is possibly my favourite in the way it increases tension via voicing on the IV before approaching the perfect intervals in the V chord from the outside.

(c) and (d) are the most minimal tweaks, basically like (a) but moving one voice early to give the other parts something to hang onto. (c) retains the consecutive octaves, but the bari move is more subtle than the bass anticipation in (d).

(e), (f) and (g) all remove the lead-bass consecutives by introducing a leap in the bass. This leads to a doubling in (e), which is avoided in (f) by introducing a bari leap too. Although both leaps obey rules of good form by moving back into the space opened up, you'd need to be expressing something pretty surprising for this to be believable. (g) has very good voice-leading, but moving back into the voicing at a 10th after starting the bar voiced at a 12th feels to me like it is dissipating some of the harmonic drive of the progression.

(h) is interesting because if you played it on the piano, it would sound like an exact parallel, but the switching of tenor and bari means that singers on individual lines aren't all trying to move the same distance at the same time. So you get the illusion of the rock parallelisms while the parts have the greater security of melodic shape to keep them on track.

I'll stop there, because this post is far longer, and has taken far longer to write, than I anticipated when I started!

Archive by date

Syndicate content