I am working in a grade 4 melody writing exercise textbook and it says that you should not have parallel 5ths and octaves on the upbeat before the new bar and the 1st beat of the new bar. It was my understanding that parallels dont sound good only when they are on consecutive downbeats and that they can be used on upbeats because of their weaker metric positioning. Can someone please clarify?

enter image description here


I wasn't going to answer as @Phoog's simple reply was enough, but then I read the comment about the Piston Counterpoint example.

Don't cherry pick examples, especially taken out of context, to rationalize mistakes in exercises.

The fig. 126 from Counterpoint you gave as an example in comments is presented in the text as: "Following are examples of such consecutive octaves and fifths." The operative word being "such", referring back to fig. 124, which is a table of 9 examples of consecutive perfect intervals in various rhythmic and melodic contexts.

enter image description here

Fig. 126 is given as either an example of good or bad. It's just a real example of many ways consecutive intervals can be encountered. It is an example of (g) from the table in fig. 124. Piston's comments on that are:

...On the other hand, the formula at g seems to be open to question. It is used by Bach and others, and its acceptability appears to depend upon the rhythmic importance of the off beats.

Compare (g) which places the octaves on the weak beats of 2 and 4 with (f) which places the octaves on the strong beats of 1 and 3. Piston is saying weak metrical accent is a mitigating factor in the acceptability of parallel octaves. You can probably extend that to any perfect, parallel intervals.

Now that we have digested the Piston text we can return to your question.

Are parallel 5ths 8tvs allowed between beat 4 and beat 1 of new bar?

Piston's table in fig. 124 has only one example crossing a bar line with (i). We can broaden the picture a bit and say categorically crossing the barline is a movement from a weak to a strong beat. Fig. (e) sort of gives us that, as the first octave is on weak beat 2. But, both (i) and (e) put the second octave not on the strong beat and instead after the strong beat. Unfortunately we don't have a perfect example. But, Piston says (e) does not excuse the parallel movement while (i) does, because (i) present more rhythmic difference between the two parts.

IMO, from Piston's examples, parallel, perfect intervals crossing the barline is generally not acceptable. Certainly if the question is asked without any qualification it is not acceptable. If the idea is "crossing the bar line mitigates parallel, perfect intervals", that is certainly false. You can probably generalize that an say "any parallel movement to a strong beat will accentuate the parallel movement." And that will undermine the contrapuntal sense of melodic independence. If you do it, make sure something other than melodic contour accentuates melodic independence, the obvious choice being to make the line rhythmically independent.

Another point should be made about rhythm, meter, and these examples. In fig. 124, g the weak beats are full beat durations, while fig. 126 put the parallel octaves after both strong and weak beats with only a sixteenth note duration. Longer rhythm values are a basic way of creating metrical accent. So, teaching example (g) accentuates the parallelism more than the real example in fig. 126. The OP question is a bit unclear on this nuance of duration.

Back to my point about cherry picking examples. If you are teaching yourself, you need more intellectual rigor when reading texts, not less. Read carefully and be a harsh critic. That doesn't mean you can't write parallel harmony any way you want. Do it, but own it. Don't pull out Piston and Bach to justify it out of context. If you want to write like Bach or Haydn, etc. be very careful about the voice leading conventions they used.


In the traditional rules of voice leading and counterpoint, parallel fifths and octaves are forbidden regardless of where they occur.

  • That is not true. In traditional strict rules, parallel 5ths are allowed on upbeats. music.stackexchange.com/questions/114978/…
    – armani
    Jun 23 at 9:07
  • here is an example where Bach uses 5ths on upbeats ibb.co/XxZhFy1
    – armani
    Jun 23 at 9:17
  • 3
    @armani there are no parallel fifths in that example. The right hand alternately ascends by thirds and descends by steps. The left hand ascends by steps. To have a parallel fifth, you would need both parts moving in the same direction at the same time by the same interval. As the question you link to notes, this contrapuntal technique avoids the parallel fifths that would be present if the right hand had two voices in parallel thirds.
    – phoog
    Jun 23 at 10:33
  • Do you see all the intervals marked with a 5? are you saying those are not 5ths? The 5th is on the upbeat but against the bass it is still a 5th. This is what the teacher says in the video. He says that Bach chose to use them there because of their weak metric positions and had they been on the down beat they would be unacceptable
    – armani
    Jun 23 at 13:59
  • 1
    @armani Just wanted to chime in. I think that the rule of parallel 5ths is more of a spectrum vs a black and white "yes this is okay" or "no, now you will be punished." In this case, Bach does have several 5ths in a row which is generally bad practice, BUT he also broke them up so that the pitches aren't convective and moved them away from the strong beats, so it's not so bad. In school, it's helpful to learn all the rules, but when you're writing on your own time you can break them and do whatever you want. :)
    – Giovanni
    Jun 23 at 16:36

I'd like to point out that that's not quite what the textbook says. It's offering some general guidelines, not hard rules (you will certainly find "tunes" that end in other ways than ^7 -> ^1). It says that first and last beats should not both be 5ths or octaves, regardless of parallelism. In the printed example, the downbeat of the third bar is an octave between the Fs in the upper and lower voice, but the preceding beat is a 9th. They're saying that it would not have been ideal if that preceding beat had been a fifth, or another octave even if the motion were not parallel.


One of the main reasons for avoiding parallel 5ths and octaves in 4-part writing is that it causes the two voices in question to temporarily lose independence and sound like one voice. The 5th and octave are strong fundamentals, so when you add parallel 5ths/octaves, the upper voice sounds like it just disappeared. But that's not to say they were never used.

Parallel 5th/octaves are generally avoided no matter where they are, but if you can break up the parallelism even momentarily, in a suspension for example, you regain voice independence, and is generally accepted.


An excerpt from The Craft of Tonal Counterpoint (p. 45) by Thomas Benjamin: "Parallel perfect fifths are not found in [Bach's] music, nor are fifths by contrary motion. Further, fifths on successive strong beats (or strong parts of beats) are not used, though fifths falling on successive weak beats (or parts) are possible." According to Benjamin, parallel 5ths falling on weak beats are allowed; however, it is a rare occurrence in counterpoint. In general, he says that successive perfect consonances should be avoided.


  • Note that these rules are very old and rarely followed 100% anymore. Parallel 5ths and especially octaves are used all over the place even in most classical music (generally because it sounds like one voice, that's kind of the point). It's really only Baroque counterpoint and earlier that made such a point of avoiding them. Jun 23 at 16:33
  • @DarrelHoffman the prohibition persisted well into the 19th century.
    – phoog
    Jun 24 at 0:55
  • @phoog And yet you still see it all over the place, Beethoven, Mozart, Haydn, etc. used parallel octaves all the time. 5ths were less common, but still regularly seen. Jun 24 at 13:20
  • The rules are just meant for pedagogical purposes, but their principles can still apply to modern music.
    – Giovanni
    Jun 24 at 13:32
  • 1
    @DarrelHoffman: Parallel octaves as a doubling were never a "rule" violation. You can see it occasionally in Bach and earlier too when it's deliberate. Textures and ensembles were generally smaller back then, which is the main reason they were rare. As for the assertion that parallel fifths are "regularly seen" in Beethoven, Mozart, Haydn, etc., I'd like some source to back that up. Aside from certain stereotypical figures (e.g., resolution of Ger+6 chord), they are quite rare (setting aside some of Beethoven's crazier late music and deliberate evocation of "rustic" music on occasion).
    – Athanasius
    Jun 24 at 15:35

It's important to start with an understanding of timbre. If you have access to an old organ with multiple stops, you know that by reinforcing overtones, you can change the quality of a sound. Octaves make a kind of flutier sound, and 5ths make a slightly buzzier sound, and so on.

When you play multiple perfect 5ths in a row, the 5th acts as a temporary timbre of the note below it, rather than as fully functional part of its own. It doesn't matter whether they are on an upbeat or a downbeat.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.