Prioritizing Part-Writing Errors in a Freshman Harmonization
In a college freshman theory class, students submit a four-part SATB harmonization of a given soprano in G major. Dev's Roman numerals and chord choices are otherwise sound, and he has left the soprano untouched, but three problems appear in mm. 3–5. In m. 3 the alto moves A4 to B4 while the tenor moves D4 to E4, producing parallel fifths. In m. 4 the soprano's leading tone F#5 is doubled in the tenor over a V chord. In m. 5 a V7 has its seventh in the soprano, and it resolves up by step instead of down. When you point out the parallel fifths, Dev says, "But it sounds fine to me." You have one 50-minute class before the next assignment, which introduces seventh-chord resolution in inversions.
1. Rank the three errors for instructional priority and justify the ranking on pedagogical rather than purely rule-based grounds.
Read the worked answer
Rank the unresolved seventh first, the doubled leading tone second, and the parallel fifths third.
The seventh is first because it is conceptual, not local. Dev is treating the seventh as an ordinary chord tone, and the next unit builds directly on seventh-chord behavior in inversions; leaving this uncorrected means the error propagates across an entire unit. It also has an unambiguous aural signature, the tendency tone failing to resolve, that Dev can hear immediately once it is pointed out.
The doubled leading tone is second because the fix is small and mechanical, respell the tenor as scale degree 5 or 2, and because the doubling rule is attached to a function Dev can articulate rather than memorize.
Parallel fifths come third not because they are unimportant but because Dev cannot yet hear them, so a rule citation will not land, and the fix may cascade through adjacent voices. Sequencing by what the student can perceive and what recurs is better than sequencing by how many textbook rules each error violates.
2. Describe how Dev should revise each problem given that the soprano is fixed, and what he should check after making each change.
Read the worked answer
The seventh is the binding constraint. Since the soprano must keep the seventh and the seventh must resolve down, the chord assignment has to change: move the seventh into an inner voice, alto or tenor, and give the soprano the root or third, so the resolution can occur without altering the given line. Dev should then verify that the new inner-voice seventh resolves down by step and that no new spacing or doubling fault appears.
For the doubled leading tone, respell the tenor as scale degree 5 or 2. Because the tenor also participates in the parallel-fifth pair, this change may solve both problems at once, which is why Dev should make it before patching m. 3.
For the parallel fifths, the smallest fix is usually to hold the tenor or move it by step in contrary motion to the alto, but the real method is to identify the chord pair first, choose the voice with the least harmonic responsibility, and then re-scan the entire progression rather than repairing one vertical slice. After each change, Dev should check the four standard scans: parallels, doubling, spacing, and tendency-tone resolution.
3. Dev says the parallel fifths "sound fine." How would you use that comment productively within a single 50-minute class?
Read the worked answer
Do not argue with the perception; test it. Play the passage twice, once as written and once with the fifths corrected, without telling Dev which is which, and ask which version he prefers and why. If he cannot hear a difference, that is diagnostic information: his error is perceptual, not conceptual, and no amount of rule repetition will fix it.
Then isolate the sound. Have Dev sing the alto line while you play the tenor, so the two voices lock into the hollow fifth sonority. In isolation the effect is much easier to hear than inside four voices. Follow with a two-voice reduction of the whole phrase sung by the class, then restore the full texture.
Close the loop with a self-monitoring routine rather than more correction: before submitting, Dev marks every pair of adjacent voices moving in similar motion with a colored pencil and checks each one. The goal of the 50 minutes is not that Dev accepts the rule but that Dev can hear the rule, because that is what transfers to the next assignment.