If you were ever a student of music theory at any point in your life, you have no doubt encountered the prohibition against parallel fifths. If you were a student of music in the last 20 years, you have no doubt heard or seen some variation of this parody:


However, no matter when it was that you were a student, chances are that the prohibition itself was difficult, confusing to grasp, or simply unconvincing i.e:

‘… because they sound terrible’

(They don’t)

‘…because they create an accent’

(Not always, and especially in the case of the fifth)

‘…because all composers followed this model. The masters would never write a parallel fifth.

(Not true)

This last point, often promulgated by theory teachers, is especially erroneous.

Brahms was of particular note to use forbidden parallels without a searing conscious, and Schenker tried to stabilize his own theories by amending that within tonic prolongational progressions, which could include a variety of the most prominent harmonies(tonic, subdominant, dominant), consecutive parallel fifths and octaves are not forbidden at all, since the lack of melodic movement in the contrapuntal texture is further supported, indeed justified by the psychological stasis that tonic prolongation induces.

To further obfuscate the issue, theorists have stated that ‘full textures’ can often seem to create forbidden parallels, simply by the byproduct of ‘chord fillers’ i.e. doubling notes to create an amalgamated compound line, which, they assure us, is, in reality,  only one true voice, as in the following example of another masters work; Beethoven’s Waldstein op. 53, I, mm. 37:


It would SEEM to create forbidden parallels because it does! The argument goes that the complexity of the texture is such that if we removed the doublings we would be left with an E major to F#m7 in parallel 63 motion, which is, of course, part of our ancestral fauxbourdon tradition of the 15th century. None the less, the reality is that the ear DOES include these parallel fifths as part of the integral sonic anatomy of the chords. Our aural perception does indeed hear the difference between the 63 chords with supporting parallel fifths vs without those present parallel fifths. They are a fundamentally different texture and only superficially represent 4 voice expansion, despite the attempt to regulate the discrepancy of their appearance. If it did not, Beethoven would have no other practical reason for including it. None the less, a theorist might say that we can boil this down to an integral four-voice texture of which, these parallels would then disappear, and that is absolutely correct; The problem is that the parallel fifth is still there visually, aurally, perceptively and psychologically.

Furthermore, if the stipulation becomes relaxed under the notion of complex textures that when ‘boiled away’ reveal proper voice-leading procedures, then how do we explain the following example, by yet another master, this time Mozart in his Violin Sonata K. 379, I:



Here the parallel fifths are not the byproduct of a doubling. If we remove the fifths, we would ameliorate the matter considerably(according to the rule). Furthermore, this type of 53 triadic movement is the textbook example that many theory teachers use to elucidate the main ‘problems’ with this type of chordal relations. They might cite that Bill Withers can use it in ‘Lean on me’ but it doesn’t have a place in 18th-century music. It will be expedient at this point to introduce the pink elephant duly parading around the room.


We are again assured that the sonority in the Mozart example is simply the reinforcement of the overtone-series which would naturally produce those notes if only the bass was sounded, and so the 3rd and 5th of the G major and E major only represent the bass. And indeed this, again, would be true. Removing the triad and octave doubling in the lower left-hand chords would obliterate any objectional voice-leading. But if this assessment were to hold true, surely a composer could write any damn set of parallel fifths and octaves since their justification of placement is inherent in that most sacred of all-natural order; the overtone series.

The last vestige of hope in which the instructor of voice-leading will turn to is the definitive old-testament of part-writing; Bach’s four-part chorales, of which the instructor will state that

  1. Bach never wrote parallel fifths(When you do, it invokes such wrath that he kills a kitten)
  2. While other composers may have written forbidden parallels, it was always a conscious decision based on the study of Bach’s writing, either in support of replication(sticking to the rules) or conscious and deliberate revocation(purposely abandoning the rules).

The only obstacle preventing us from packing up our toys and going home is this; It’s a lie.


In the Riemenschneider edition of the Chorales, there are at least 46 instances of forbidden parallels over the course of 371 examples. If we were to remain conservative and assume that of those chorales only one mistake per chorale is made, where this is one at all, that would mean that 12 percent of the chorales contain mistakes

In other words, Bach would have received 82% on his work,  if his chorales were submitted for assessment by university panels. Ironically, the criteria by which the university demarcates it’s grading system, including what constitutes 100% , is, no doubt founded on Bach’s chorales.

Bach becomes a ‘B’ student based on the bar set by his own work.

Perhaps we could say that these were mistakes Bach absent-mindedly made. Perhaps he was not as meticulous as we proclaim. After all, he was writing one major cantata per week at one stage in his life. But making this proclamation also forces us to question the insistence and stringency of its application in university and conservatory studies at all! After all, we are nicely situated in the 21st century! Why should we care for a practice of style that very few people are still composing in seriously today? A style that is now 50 years shy of being 400 years old? No one surely writes books in Shakespearian English anymore!

To be sure, I find this last argument severely depressing. I could not imagine a musical landscape in which it’s prophets, technicians, and dreamers, those artists of the future, did not thoroughly study and subsume the voice-leading principles of the 18th century into their very bone marrow. These practices were ‘discovered’ in that century because of a poignant inner-symmetry and intervallic distance. This internal structure has become only increasingly apparent with our undertaking of exploration of the chromatic universe, of which the diatonic/triadic model is only but one partitioning of a profoundly rich codex of symmetrical designs.


And, people do still love Shakespearian English:



Furthermore, contrary to what our conservative assessment posited(1 mistake per chorale), one chorale, BWV 40.8 ‘Cantata, Dazu ist erscheinen der Sohn Gottes’ contains 4 incidents of forbidden parallel movement. That is a high percentage to constitute as mistakes. Clearly, Bach did not find them offensive.


And herein lies the crux of the operation. We do not teach, because we were not taught, how parallel perfect fifths and octaves are tools to the composer.

A carpenter, crafting a bed, might gently tap a mallet against a chisel to groove intricate details into the ornamental ‘wood scroll’ making up the edge of the backboard. He might then go on to use the mallet to attach the backboard to the rest of the frame by hammering nails at their adjacent angles. If his tapping was too heavy during the former process, the ornament would be destroyed and the integrity of the design would be ruined. If the hammering was too light during the latter process, the entire frame would not hold together and the structure of the entire frame would fall apart.

Therefore, the use of parallel fifths is a discipline in intensity and context, and not prohibition and exception as is often currently taught. We should be thinking of when and when not to use perfect fifths.

If we know what perfect fifths are capable of, and why they are endowed with those capabilities, then the question of ‘is this safe to do, according to voice-leading “experts“‘ is nullified based on our ability to hear them correctly in context.

Let’s just address the first myth. Perfect fifths do not sound bad. Far from it. For that matter, tritones don’t sound bad. They sound ‘energetic’ in particular contexts, and at other times completely neutral depending on context. But the Perfect fifth is special because it is the first partial in the overtone series that is not considered equivalent to it’s fundamental, the octave. At the same time, it exhibits such a strong relation to the fundamental that it essentially grounds it. Whereas, G and B can be a G major chord or an E minor chord, or even part of a C major 7th chord, the G and D a perfect fifth apart always solidifies and orients us to the fundamental. We KNOW is some sort of G chord. Therefore the perfect fifth is a very strong pillar of harmonic stability. When we ornament the chord with non-chordal tones(notes that are not part of the triad) it is the perfect fifth that acts as a sextant, firmly asserting the underlying true identity of the chord; i.e. when E and B are present against a G and D related by fifth, the ear does not hear EGBD= E7 wanting to resolve to A minor(or A major if E7 is the supertonic of D), but G major with an E suspension wanting to resolve to D.

We must conclude then the perfect fifths will not sound ‘wrong’ to our ear, they will sound ‘stable.’ Why would we want to loosen stablity? In order that we might give the impression of fluidity. A resultant consequence of parallel perfect fifths sounding wrong is that it assumes that using a lot of parallel thirds and sixths is ‘right.’ And again, it is not. In the same way that a perfect fifth creates too much stability, so too do a continuous line of parallel sixths or thirds ‘fuse’ into one line with an intervalic ‘color’ shadowing the fundamental line.

The observation suffices to solidify the notion that we do not use parallel fifths when we want to create a fluid harmonic texture. This can be observed readily.

In support of this position, Bach’s chorales that exhibit perfect fifths fall under three major categories:

  1. Parallel perfect fifths that are the result of the ending of one Aufgesang(phrase) and the beginning of another. Such as in the case of BWV 78.7 (All images are borrowed from www.bach-chorales.com but annotated by the author.)
  2. When they arise out of a syntactic and fundamental chord as ornamental passing tones, such as BWV 48.7 mm 13This example is possibly harder to detect, but the blue arrows align the voices so that we can see the fifths clearly. Again, when the perfect fifths are the result of an ‘attachment’ to the metrically placed, functionally prescient harmony, the passing chord is not perceived as a chord at all. There is a very easy way to relate this musically. When an artist sings a text and interrupts statements with interactions such as “oh” “mm” and “yeah” our brain simply filters them to the background of context. When we hear “It was a monday… Oh, yeah… It was cold.” We do not try to figure out what ‘oh’ is elucidating or ‘yeah’ is affirming. They’re ornamental, they add to the music, they have a purpose, but they structurally lower, or ‘background’ and thus do not interfere with the syntactical linear aspects of the music. An even more striking example is found in BWV 167.5 at the very onset, using parallel octaves in a very exposed position, by step and not leap which draws greater attention and, at the beginning of a formal structural area. It is difficult to assume this to be an insouciant oversight on Bach’s part.
  3. When they are the result of a delayed arrival of the ^7th in a dominant chord at cadences, such as at the cadence in BWV 146.8In this case, the fifth that is created in the inner voices of the dominant 7 are not structural. Remember when we said that if G and D are present as a perfect fifth, then its hard to interpret an E and B as a seventh chord in inversion, rather the ear hears it as ornamental non-essential tones. And that is exactly what is happening in the numerous case of parallel fifths in this matter, found in the chorales. We do not lose orientation that D is the basis of the triad. The C is a passing tone no to B, ^3 of G major and G is an anticipation of the tonic with the motion ^2-^1.

By no means is this list exhaustive and we must remember that we said that there are AT LEAST 46 examples, but the Riemenschneider edition of the Chorales omits some 60 extra chorales. There are other, more baffling examples of the forbidden parallel perfect motion that are not within the scope to investigate within this article.

What I hope that is expressed, is that parallel fifths(and octaves) constitute a fundamental, stable, and perfectly usable tool for composers, of which the merits of its use must be determined by the ear and context of both student and instructor. There is room for negotiation, just as there is room for the painting instructor and student to negotiate ‘how much blue is too much blue’, or the culinary instructor and student to negotiate ‘how much cinnamon is too much cinnamon?’


We should not think of these as ‘breaking the rule when we are justified’ because they should not be viewed as rules and prohibitions at all, just as a dentist who decides not to use a hammer on your teeth is not ‘following’ the rule of prohibition against hammers on teeth, but is rather exercising common sense.