# What is the clearest way to signify the harmonic function of a polychord?

Chords can be analyzed with roman numerals to indicate function:

`C, Am, Dm, G`

`I, vi, ii, V` (or `I, VI-, II-, V` in jazz theory)

And inversions may be analyzed by adding an arabic numeral indicating the note which takes the lowest voice:

`C/G, Am, Dm, G/B`

`I/5, vi, ii, V/3`

But what about polychords, and their ambiguous cousin, slash chords?

`C, Am, Dm, F/G`

`I, vi, ii, IV/2`

The "/2" seems a bit unclear. You could write "IV/V", but that's also a secondary dominant.

Usually, polychords are written like this: (with a fraction).

So, let's say you have Am {fraction} G7 and we are in the C major scale. You could symbolize that as VIm {fraction} V.

Notice that for the polychords, there is no slash, but a fraction. Slashes are for the slash chords or hybrid chords (inversions).

For the slash chords inversions, if the score you are written isn't really formal, you could write it as `Cmin (1st inv)` or `II(3rd inv)`.

If the score is formal, the first inversion is symbolized with a `6` (actually ^6 -- power of 6) so, you could symbolize Cmin/Eb as I^6, and the second inversion with a `6 4` and the 3rd inversion with a `2`.

If I'm not mistaken, a F/G slash chord, is written as F^9, or IV^9

• Great answer. So for a formal poly chord, it should be written as a fraction. That is consistent with what I've seen too. But what about the "slash" chords? Like F major triad with G in the bass? You could think of it as a G7sus4(9, omit5), but usually it's written "F/C".
– Grey
Aug 22, 2014 at 8:30
• If I'm not mistaken, it is written as F^9, or IV^9 Aug 22, 2014 at 8:34
• Oh I see, yes, indicating the bass note is a tension does kind of get the point across. Good call.
– Grey
Aug 22, 2014 at 8:35

# Polychords, Roman numerals, and Functional Analysis

Roman numerals, in the context of functional analysis, are an analytical language, not a chord-naming symbol. As such, they don't follow the conventions of chord-naming systems. C/F is meaningful as a chord name (whether slash chord or polychord), but its functional meaning is undefined. I/4 and I/IV are not part of the language of functional analysis (except I/IV as a secondary-chord indication). In the context of functional harmony, polychords, specifically, have no meaning as such, rather they would be labeled as a single chord according to their overall harmonic function, and that would depend on context.

Using the F/G example:

## F/G Polychord

### Example 1

Consider a piece of music in C major. We encounter an F major chord, which is tied with a G major chord being added underneath the tied portion, followed by a C major chord.

Analytically, the chords would be labeled IV - V742 (or V974) - I. This notation is understood to mean that the seeming polychord is "just" a V7 chord with suspensions from the previous IV chord.

### Example 2

But now consider a piece in which there is a sustained G chord with a variety of chord changes above it, including an F major chord. This is as close as functional harmony would come to an explicit polychord. The Roman numerals would be given according to the upper-voice chord progression with another notation indicating a G major pedal chord. (Or course the notations might change as in Example 1 should notes from the G major chord actually be functional parts of those upper chords.)

## F/G Slash Chord

In this case, it is likely the Roman numeral will correspond to the F chord, with the G being interpreted either as a pedal tone or a suspension, depending on context. It also could be interpreted as an incomplete G7 chord with suspensions (as in Example 1, above). Less likely it would be considered an F chord with a non-harmonic G. For instance, consider the progression F - F/G - F. In this case, the G would probably just be seen as an ornamental upper neighbor and not as part of the harmony.

# Polychords in Polytonality

One could imagine a polytonal piece in which, say, one set of voices was in F major while another set was, simultaneously, in G major. Polytonal music analysis may borrow ideas and notations from functional analysis, but the stretching of the language's meaning would be clear in that context.

Two analytical options for polytonality — provided each tonal part was well-enough behaved in functional terms:

1. Use two separate analysis, one for each part.
2. In the specific case where two simultaneous chords play the same role at the same time within their respective keys, one could perhaps "invent" a notation where a single, "big" Roman numeral is used. Something along the lines illustrated below.

In this setting, a polychord-type notation would have meaning, but at present there is no such established notation.

• What about likely the most often cited example of a piece that uses polychords in this website, "The Point of No Return" from The Phantom of the Opera? It outright starts with polychords, but settles down into F minor. Aug 27, 2022 at 13:02
• @Dekkadeci Tell me more about it. I'm not familiar. Aug 27, 2022 at 13:52
• music.stackexchange.com/questions/72068/… asks about the polychords in "The Point of No Return". Admittedly, the question implies that the first chord may not be a polychord after all, although there are definitely polychords in the introduction. Aug 27, 2022 at 14:14
• @Dekkadeci I can't give a good answer. The notated example certainly indicates polychords, but in the recording I don't hear polychords, and I don't have access to the orchestral score. Either way, I don't hear it as functional harmony — just a sequence of chords more akin to planing. Aug 27, 2022 at 18:48
• Do you hear the recording's music as using (possibly horrifically) extended chords instead of polychords? I do agree that the introduction's harmony does not really sound functional, but that hasn't stopped me from analyzing heavy metal and video game music with Roman numerals and realizing that they at least use tonic and non-tonic chords. (Heavy metal, or at least extreme metal subgenres, often repeatedly uses the same chord progressions like i-bII-i and i-#IV-i.) Aug 28, 2022 at 3:36