Phrygian is a very distinct sounding mode. I've been trying to better understand the harmony of the mode so I can use it more. Obviously the i chord is a tonic chord and iv is a subdominant chord, but the other chords in the mode seem not to line up right. For example, there is a vo chord, which does not lead back to i, and instead leads to VI, and the II chord seems to act much more like a dominant chord then any other chord in the mode.

So what chords are tonic, subdominant, and dominant in Phrygian?

  • 5
    I love this question and wonder what an empirical approach might uncover. That is, we could look at songs written in phrygian, find common progressions (e.g., III-II-i), and then assign chord functions based on how the chords are most frequently used (e.g., III is subdominant, II is dominant, and i is tonic). Along those lines, is there a particular genre you're thinking of? Or better yet, do you have any specific songs in phrygian in mind? – jdjazz Jan 18 '18 at 22:44
  • 3
    I'm not sure if the terms apply to modes. Here is an answer discussing Persichetti's concept of primary chords from T̲w̲e̲n̲t̲i̲e̲t̲h̲ ̲C̲e̲n̲t̲u̲r̲y ̲H̲a̲r̲m̲o̲n̲y. music.stackexchange.com/questions/72137/… – Chris Strickland Dec 27 '18 at 6:09
  • Modal music is not about functional harmony. Functional harmony is tonal, Phrygian is modal. Unanswerable. – user53472 Jun 19 '19 at 3:30

You can't really apply these functional terms to modes. A mode refers to a specific scale and its characteristic melodic phrases. The problem with Phrygian, when you try to build classical harmony on its scale, is in fact constructing the "dominant" harmony, i.e. the triad on the 5. step. It has a diminished fifth (in "White Key Phrygian" on E this would be B-F), so you can't get a major or minor triad.

But let's ignore that for a moment and try to construct a chord from the Phrygian scale which is "dominant" in that it leads back to the "tonic". We'll have to look at Gregorian Chant and pre-Baroque polyphony:

A typical Phrygian chant melody reaches its final note from the second step (a semitone above the final) downwards to the final. This is unique, because all other modes have whole tones above the final as penultimate notes. This typical Phrygian semitone can also be described as a downward leading-tone.

Now moving on to two-voice counterpoint. If your main voice (in medieval terms, the "tenor") has the typical closing phrase F->E, you can make a two-voice cadence by adding an upper voice (a "discantus") which forms a major sixth with the tenor, resolving to an octave: D->E. Now you have the two main parts of a polyphonic Phrygian cadence.

You can now add more parts ("contratenores"). For example, if you insert a middle voice between your tenor and discantus (a "contratenor altus"), you could let it move in parrallel with the discantus: A->B. That would be a typical three-part Phrygian cadence, ending on a perfect harmony:

D -> E

A -> B

F -> E

So, if you translate that to classical harmony, in this example a d-minor chord (in first inversion) functions as the "dominant". But of course, as said above, these terms don't really fit.

  • 2
    I like how you worked out this answer. B->B wouldn't be the middle voice, as it would mean b d f an unstable diminished chord for a dominant, right? I've always thought Dm or F could be the "dominant" for E Phrygian, more or less in line with your answer. – Michael Curtis Apr 12 '19 at 20:52

I haven't ever encountered this topic in a theory book, but I think an empirical approach would produce good answers. The method I'm envisioning is: find a bunch of songs written with a phrygian tonal center, pick out common chord progressions, and determine the chord functions based on how they're used.

Here are chord progressions I'm most familiar with (I'm thinking wide enough to encompass Spanish phrygian/phrygian dominant tonal centers too):

  • III-II-i
  • iv-II-i
  • vii-i or VI-vii-i

From those progressions, we could deduce these functions:

  • subdominant chords: III, iv, VI
  • dominant chords: II, vii
  • tonic chords: i

The only chord missing from this list is the vo chord (or vø7 if the 7th is included). I can't think of any progressions involving the vo at the moment, but my intuition is that the vo chord plays more of a subdominant function than a dominant function. For example, I can imagine a vo-vii-i progression sounding pretty straightforward in a song written with a phrygian tonic.

As a side note, I do think this question makes sense to ask. Many songs exist that have phrygian tonal centers but are non-modal. In these songs, the chord progressions include the sort of resolution that modal tunes actively try to avoid. For all such songs, I think it's useful to consider a phrygian tonal center and figure out the traditional function that each degree plays.

  • 2
    @MaikaSakuranomiya, normally, I would agree. But in phrygian, I don't think the typical functions you cite apply. – jdjazz Apr 2 '19 at 20:16
  • 1
    @MaikaSakuranomiya, that's a good question, but I think it only makes sense to ask in the context of traditional harmony. It seems like you might still be thinking about this question from the perspective of the "normal rules" of functional harmony. If we move away from traditional major/minor tonal centers and instead allow a phrygian tonal center, then does the tritone substitution even make sense as a concept? – jdjazz May 22 '19 at 12:31
  • Also, I also would like to say the vø would rather have dominant function, because it contains the characteristic ^2 tone like II and vii. Furthermore, I can think of progressions like i-iv-vø-i or i-VI-vø-i, or even i-III-vø-i. – user53472 Jul 5 '19 at 4:24
  • 1
    @MaikaSakuranomiya, I haven't heard those progressions but I'm very interested in learning more. Could you mention a couple songs with those progressions you cite? Thanks! – jdjazz Jul 5 '19 at 20:11
  • 1
    Thanks for this great answer. But you write, “Many songs exist that have phrygian tonal centers but are non-modal. In these songs, the chord progressions include the sort of resolution that modal tunes actively try to avoid.” Could you name some examples? – symplectomorphic Sep 23 '19 at 21:25

Dominant, subdominant etc. are terms from functional harmony. You can give a modal flavour to functional harmony (throw in some ♭VII or #4 notes) or you can give a functional flavour to modal harmony (stick the sharpened leading note into the cadences of 'Greensleeves'). Music (as all art) thrives on cross-breeding, there's no special virtue in being pedantically functional or pedantically modal! But when you do this, it becomes pointless to impose the strict theoretical labels of one scheme or another.

You can build a triad on any note of any mode. If the 5th note of the mode happens to be a perfect 5th above the tonic (I think we can allow every mode to at least have a tonic? Unless you count the whole-tone scale as a mode?) it will have some of the character of what functional harmony calls 'dominant', a tendency to resolve to the tonic. If it also contains a major 3rd, we've pretty well got a dominant, if we choose to use it as such.


Again we are mixing styles and terms of analysis of different epoques. OP question is interesting but it won’t help much to juggle a lot with the function terms or the roman letters or chords.

However - it isn’t forbidden to try to do so.

jdjazz is right: only an empirical approach will bring clearness but even then we wouldn’t know wether this investigated songs were based on rules, laws, misunderstandings, errors, ignorance or original inventions.

Anyway, we wouldn’t be surprised to find music in Phrygian mode with a dominant V7: right the one degree and function that should contain the elements of this specific mode. (Why not? analogically to the major V in harmonic minor ...) or should we say then: This tune is in phrygian mode with a harmonic close?

It is like the orthography in language: Music and languages are living culture. And what is living is changing.


In classic music theory, when you see that the 7th note of your minor scale is not augmented, then you use the phrygian chord progression. The chord functionality never changes.

I-VI are T(tonic)

II and IV SD (subdominant)

V-VII are Dominant

The III is considered to belong to T and D team.

What changes in modal harmony, is the tonal centers. Since you begin with the 3rd note of your scale

E is the I

A is IV and

B is the V.

Now, what happens if a phrygian mode appears somewhere in a diatonic work?

For example in Amin if the 7th tone is G and not G#, then you have to follow some different rules.

Because of the fact that the V is not major anymore, progressions such V-IV are allowed, for we don't have the G# to A resolution. One easy way to approach this type of harmony is this:

Try to discover the second 4-chord of your minor scale descending. For instance,


Then apply the following harmony:

  • in A: I
  • in G: φV6 or φVII6 (phrygian V of VII in first inversion)
  • in F: IV6
  • in E: V

If the second 4-chord is in the Soprano voice, you can also apply

I-φIII-IV-V respectively.

Note that in the final chord the V must be major. This is how you declare the end of the modal harmony. If you have some more notes before this very last E, then you approach all these notes with phrygian mode harmony and in the last E you augment the 7th note.

  • Why doesn’t have this answer any votes? Is there an error I don’t see? – Albrecht Hügli Apr 12 '19 at 5:22
  • Because it's massively and unnecessarily complicated? – Laurence Payne Apr 12 '19 at 12:22
  • @AlbrechtHügli Because the minor seventh definitely doesn't always imply phrygian? – user45266 May 22 '19 at 16:29

True modal music is not functional harmony , if you would play modal some tips:

Emphasize the tonic ( pedal, ostinato )

play the distinctive note of the mode ( 9 b for phrygian )

do not use tritone in chord ( it sounds tonal ! )

  • 2
    I'd just add that "functional harmony" is a historical term, and has nothing to do with whether harmony functions or not in modal music. – Scott Wallace Nov 14 '17 at 11:25

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.