In a piece that is in C-major, the chords in Roman numerals would be I for C, ii for dm, iii for em, IV for F, V for G, etc.

Now let's say the piece also features a (rare) Eb chord. How would I represent this with a Roman numeral? It sits between the ii and iii, so how would I represent it using Roman numerals?

  • 1
    The answer is given in the wiki page you cited: ♭III May 31, 2021 at 21:39

3 Answers 3


E is the third scale degree in C, and Eb is a flat three, and if such a chord is playing, we must assume that the third degree is chromatically altered, "flattened". As is the seventh, B --> Bb, which is in the Eb major chord (Eb, G, Bb). By calling it Eb you declared that the third degree is made flat. If you had said "D#", it would have meant that the second degree is made sharp.


In popular music and rock music, "borrowing" of chords from the parallel minor of a major key ... in the key of E major, chords such as D major (or ♭VII), G major (♭III) and C major (♭VI) are commonly used.

So, you call your flat three major chord a ♭III.


In key C major, E♭ will be represented by ♭III. Any non-diatonic chords could be represented thus. So B♭ could be ♭VII. Something like D7, though, may be labelled V7/V - as it's the dominant of the dominant. Likewise, E would be V/vi, the secondary dominant of A minor, which is a diatonic chord.


Roman numerals do not name literal chords. They label chord functions. As such, the appropriate Roman numeral would depend on the context: the chords both before and after. An E♭ major chord appearing in the key of C could be a borrowed chord from minor, an applied chord in some other key, a modulation/tonicization in itself, or a passing/ornamental chord that wouldn't receive a Roman numeral.

The further complication is that it depends on the "in between" chord. E♭ minor, augmented, and diminished — not to mention the various seventh chords rooted on E♭ — would also be "in between" chords subject to the same context-dependency described above.

In fact, even the idea that D minor is ii, for example, makes quite a few assumptions about context. It would be somewhat more accurate to say "In the key of C major, a D minor chord commonly functions as a ii chord".

  • I have a misunderstanding! Often, a secondary dominant chord , for instance, V/V, whilst actually being the dominant of the dominant, doesn't actually *lead to that dominant. therefore, to me, V/V is not functioning as such, but still gets the label. Is there some ambiguity here, or is it just me - who used to think secondary dominants must lead to their own dominant chord?
    – Tim
    May 31, 2021 at 12:55
  • @Tim To be a dominant chord, it needs to serve a dominant function. However, dominant functions include, for example, half cadences and deceptive cadences. So it V/V might not lead to V; it just needs to serve a dominant function within the key of V. However, not all dominant seventh chords are V chords. D7 could be V/V, but it could also be bVII7/iii, say.
    – Aaron
    May 31, 2021 at 13:01
  • If it's called a dominant chord, then in some regard it "is" a dominant chord. It seems that the word dominant is sometimes used by some people to describe any sort of dominant seventh chord. You see "C7" you say "C dominant", and might play something like Edim7 or some completely weird stuff that you feel could work for some purpose. Regardless of what, if anything, is before and after the C7. There is no official standards institute of correct meanings, or if there is, people don't respect it. May 31, 2021 at 13:21
  • Isn't the fact that it's built from the dominant note (5) from that key relevant? If V/V 'serves a dominant function within key V, then, yes, that ought to be its function, and describable as such. But if it goes elsewhere next, it's not actually fulfilling that function - which is supposed to be the concept of RN. Not seen bVII7/iii yet. And that would only be used if the next chord was iii, otherwise it gets called, as usual, V/V.
    – Tim
    May 31, 2021 at 13:31
  • 1
    I have the feeling that the first sentence doesn't say precisely what you're trying to say. I mean, the fact that ii or V has a particular function isn't inherent in the Roman numeral, and in fact these signs do (literally) denote a particular chord relative to a given tonic note.
    – phoog
    May 31, 2021 at 13:49

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.