In the video How To Analyze Songs from the channel 12 tone, the guy wisely separates the task of labeling from analyzing. There's a comment where he lists some possible analysis options of the label bIII:

  • Modal interchange

  • Chromatic mediant of the root

  • Passing chord between ii and iii
  • New root in a direct modulation
  • Chromatic mediant of the V chord
  • Tritone substitution resolving to ii
  • Secondary-axis substitute for V chord
  • Secondary dominant of tritone substitution of V chord
  • It's all just atonal
  • Someone also says: part of an extended plagal cadence bIII bVII IV I

I'd like to know, how a bIII (the Eb chord in the key of C) functions different when analyzed as a modal interchange chord and a chromatic mediant chord? And also how would you differentiate the chromatic mediant of V and of I. I thought chromatic mediant was just a label of a chord distant a minor/major third apart from a degree (not being diatonic), but aparently there's also a semantic issue involved because a chord would "function" as a chromatic mediant and not as a modal borrowing. What should I look for to determine this classification? Thanks.


I believe the "Chromatic mediant" wikipedia article shed some light on my question.

"Some chromatic mediants are equivalent to altered chords, for example ♭VI is also a borrowed chord from the parallel minor (...), with context and analysis revealing the distinction."

And that was exactly my question: What to look for in the context to distinguish both? The article quotes:

"Chromatic mediants (...) provide color and interest while prolonging the tonic harmony, proceed from and to the tonic or less often the dominant"

So, is "being surrounded by the same chords" a necessary condition for a chord to receive the chromatic mediant terminology?

  • I feel like there are multiple questions in this question. I think if you asked one at a time in separate questions it would be more appropriate.
    – coconochao
    Commented Dec 7, 2018 at 16:38
  • I edited and deleted the implicit last one :) I believe everything else is related. Commented Dec 7, 2018 at 19:16

2 Answers 2


'Chromatic mediant' isn't an analysis, just a description. 'Modal interchange' is often merely an excuse. (Yes, a chord that isn't diatonic in this key will probably be diatonic in some other key. So what? We're in THIS key, the chord needs a reason to exist in THIS key!)

The sort of analysis that wants every chord to have a dominant relationship to the next one can run into problems with much commonplace modern harmony. Look carefully at what a chord DOES, in context, before assigning it a functional description.

C, F, Fm7, Bb7, Eb. A journey from C major to Eb major. Fm7 and Bb could be usefully analysed as 'borrowed' from Eb major. (But note that we can get to ANY new tonal centre by putting its ii7, V7 in front of it - there's nothing particularly special about this being a trip to the 'chromatic mediant')

Or let's just go C, Eb, with a common melody note of G. Yes, that works. It works particularly well when the common note is the tonic or dominant of the home key. C, Ab with the common note C is nice too. (Are we beginning to sniff a reason why 'chromatic mediant' is 'a thing'? But be careful, we could come up with something just as plausible for just about any other chord.)

C, E, Eb, D, (G7, C). Let's be honest, a guitarist was just sliding a full barre chord shape along the fingerboard. Sounds fine, but analysis of the Eb as a 'chromatic mediant' or a 'b5 substitute dominant of D' would be ludicrous. (The excuse for this is 'planing'.)

  • Ok, tks. I believe "'Chromatic mediant' isn't an analysis, just a description." answers the question. You believe the author of the video is wrong when he considers a modal borrowing or a chromatic mediant analysis to be in the same level. Commented Dec 9, 2018 at 21:07
  • I'm not too sure that 'modal borrowing' is much of an analysis either in many cases. As explained in my answer.
    – Laurence
    Commented Dec 10, 2018 at 2:36
  • Thanks, @LaurencePayne, for the giving a response that does away with the overt sophistry that can plague analysis! Commented Dec 11, 2018 at 9:57

V and I do not share all their chromatic mediants, so there are chords that are chromatic mediants of V but not I. For example, in C major, B major and B flat major chords are chromatic mediants of V (G major) but not I.

As for bIII as chromatic mediant vs. modal interchange, I generally only classify bIII as a chromatic mediant when it is tonicized/modulated into, it resolves immediately to I, or it resolves directly from a dominant-function chord. Otherwise, it sounds more like it was borrowed from the tonic minor (one of the prime cases of modal interchange).

  • V and I share the bIII as chromatic mediant. Then I think you'd say that if it resolves immediately to V it would be a "chromatic mediant of the V", is that right? Commented Dec 9, 2018 at 21:27
  • @Allan Felipe - Not unless it transitions directly from V (e.g. I V bIII). If all it does is resolve immediately to V (e.g. I bIII V), I'd call it a modal borrowing.
    – Dekkadeci
    Commented Dec 10, 2018 at 8:43
  • Your answer is close to the kind of reasoning I expected, but maybe not thorough enough. I made an edit to my post. Commented Dec 14, 2018 at 23:08
  • By the way, It seems that your comment contradicts your post. If the mediant M appears in A->M->B your post says that you pay attention to B ("M resolves to..." is the important thing) and not A, but the comment says the opposite ("M transitions from..." is the important thing). Commented Dec 14, 2018 at 23:17
  • @AllanFelipe - I edited my answer just now to make it clearer. In general, I only classify bIII as a chromatic mediant at all if it is tonicized/modulated into or it is not used like a predominant chord.
    – Dekkadeci
    Commented Dec 15, 2018 at 0:50

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.