I believe this melody well conforms with the harmony, and does not offense my ear and the rules about counterpoint and chord progression. However, when trying to fill in the inner part, I realize the 4-th degree in IV6 will either go to 5, which forms a parallel fifth with soprano, or go to 7, an awkward tritone leap. If going to 2, the fifth of V will be doubled (soprano already has a 2), which is incorrect as V is in root position. As a result, some exception of rules must be made. But which?

  • It's not so easy to recommend anything if you don't give use the surrounding chords. Please post a pic of the relevant part of the score. Jun 16, 2014 at 0:00
  • If the question is, which rule should I offend first then it's easy. parallel fifth is simply a no-no. Tritone leap should be shunned. So the doubling seems the least evil. Not sure what you mean by "...which is incorrect as V is in root position"? What rule would that break? Jun 16, 2014 at 0:03
  • Yes, although my expression of question is not clear enough, but you got it exactly. I say "which is incorrect as V is in root position" refers to that, in root position, doubling fifth is a wrong doubling (but it is correct in 1st/2nd inversions). Jun 16, 2014 at 4:04
  • 1
    Although doubling the root of a root position chord is generally preferred, it is definitely not a rule. The only doubling of a V chord that fully violates a voice-leading rule would be doubling the third, since that's the leading tone. Doubling the 5th is not ideal, but it isn't terribly uncommon either. Jun 17, 2014 at 1:18
  • If I am putting the V in a half cadence, which should be given enough harmonic consideration, is the doubling fifth still the best choice? Jun 17, 2014 at 5:55

2 Answers 2


Changing the doubling of IV6 will not help, as the issue concerns the progression of the root, which must appear in at least one inner voice. The doubling is definitely the lowest-priority rule, so 4 should proceed to 2, possibly jumping up to 5 if necessary to connect to the next harmony. (Taking advantage of the fact that the original asker did not specify a key, I note that G major yields comfortable ranges.)

    B - G - A
    G - G - F#
    D - C - A -(D)
    G - E - D

Alternatively, if only the bass and soprano are fixed, the progression I-vi-V yields nice root doublings:

    B - G - A
    D - E - F#
    G - B - D
    G - E - D

The convention is "Roots a second apart":

"Roots a Second Apart (no common tone): Move the upper three voices in contrary motion to the bass, making sure each voice moves to the nearest chord tone of the next chord; the roots of both chords should be doubled."

Though the first chord is an inversion:

"When one of the two triads is in inversion, write to or from the doubled note first, using oblique or contrary motion if possible, and then fill in the remaining voices."

I am working a picture of the score for your situation, but hopefully this will help


enter image description here

A. I see your problem, even with doubling the tonic in IV6

B. If you are able, going to a V7 would help avoid issues, but I understand if you can't/don't want to

C. This is risky, but you could go in similar motion with the bass, doubling the fifth of V consequently

D. I guess if you doubled the third of IV6 you could escape, but I completely understand if that is not acceptable able for you. Plus it kinda sounds terrible. The tenors have that tritone you were trying to avoid, but they're hopping to "ti," so it's not that bad.

I believe the best option is for the alto to go to B (going to "ti" from anywhere shouldn't be hard) and make the tenor go to high G (sorry fellow tenors). Or you could consider my options above or something else. Who knows?

I apologize for any misunderstandings and/or errors. I am no expert. By any means.

  • Thank you, and I see that in any case I have to make a choice about what rules to give up. If we can use V7, I will definitely go with (B). Otherwise, I will choose (C) if IV6 is in metrically stronger beat than V, and (D) if the opposite. (I personally feel voice crossing with both voices leaping stands out too much, like the option that alto goes down to B while tenor goes to high G). Jun 16, 2014 at 4:22
  • Now my case is to harmonize 3-1-2, where 3,1 are quarter notes and 2 is a half note meant to be a half cadence, with all the three notes in the same measure in 4/4 metric. So the harmony of V is important, thus I shall give it a correct doubling, and as a half cadence it cannot be replaced by V7. The remaining choice now is (D), to make a dim5 leap of tenor from 4 down to 7. But I am still worried about this leap, because tritone leap usually occurs within dominant harmony, not between a predominant and dominant. So I wonder if I have to give up my harmonic plan. Jun 16, 2014 at 4:36
  • Maybe you could do a suspended IV, or use a non chord tone. Jun 17, 2014 at 1:09
  • I just realized that a common harmonization of soprano 3-1-2, bass 1-6-5 is I-V43/V-V, but still it is difficult to implement because a d5-P5 will occur. (#4,1 - 5,2). If I intend to keep the rhythm in unison and contain no nonchordal tone, will there be solution? (I could not find perfect one, but it is unbelievable that such a good soprano-bass counterpoint that implies at least two good harmonic progressions do not have a 4-part realization.) Jun 17, 2014 at 6:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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