Just Friends

The analysis of Just Friends poses a conundrum: on the one hand, the song's long-standing public exposure means that its harmonic language is almost idiomatic; on the other, a close look reveals an odd harmonic ambiguity.

Gustave's harmonic analysis of the song is the starting point for an in-depth discussion, that touches on:

  • mode change
  • “minor plagal”:
  • creating pivot chord:
  • and ‘key’
  • Picardy Third
  • pivot chord

Just Friends

Hi, Gustave;

As I wrote earlier, I spent some time looking at and thinking about some of the harmonic devices used in this tune. First off, however, let's look at your own analysis:

Excellent: you’ve correctly identified each chord according to its function (although I’d like to come back to the Bbm7 – Eb7 – Am7 of bars 7 – 9, later), as well as correctly observed that the Cm7 and F7 chords are derived from the key of G minor – which, of course, is a mode change, and not a modulation. (More on the Cm7 and F7 chords, as well, in a moment.)

Personally, I would identify F#m7 and B7 as being simply “ii of vi” and “V of vi”, rather than “ii of V of ii” and “V of V of ii”. I’ll explain why: although both “vi” and “V of ii” have the same root (the submediant – “E”, in this tune), they have very different functions. While “vi” is simply the chord built upon the 6th degree of the scale, “V of ii” implies a chord that performs the role of dominant in relation to the supertonic. Although dominant chords can fulfill various other roles — eg, functioning as a goal of harmonic movement, or as a contrapuntal chord that simply links and supports other chords — the main ‘dominant’ function is to drive the music back toward the tonic. Similarly, ‘secondary dominants’ – like this ‘V of ii’ – drive the music toward their respective ‘transient tonics’;   eg, ‘V of ii’ drives toward ii.

If we play bars 25-28 (melody too!), we sense that the first note of bar 28 is itself a goal – of both melodic and harmonic movement – and that the Em7 that accompanies it acts more as a chord of resolution, or place of rest (albeit very temporary), than as a chord that is actively driving the music toward the supertonic chord. That is, we don’t get the sense that the Em7 is fulfilling any ‘dominant’ role, but is rather acting as a harmonic goal itself – and that goal is simply ‘vi’. (Of course, ‘V of ii’ would necessarily have to be an E or E7 chord, not an Em or Em7 — but that detail is secondary to the fact of the chord’s function within the given context.)

Well, yes -- and no: from certain perspectives, there’s a lot going on — let’s take a look, beginning with the song’s first ten bars:

If you look at the rectangles I’ve drawn, you can see that the opening of the tune consists of a melodic sequence: note that the two phrases are identical, the second one being a transposition, down a second, of the first. The second phrase’s harmony, though, is not a transposition of the first’s, so this sequence is melodic only.

Sequences are most often both melodic and harmonic, so why did John Klenner (the tune’s composer) not fulfill the harmonic aspect? Well, it turns out that the harmony of its first phrase — IV of G major (Cmaj7) → iv of G minor (Cm7) — isn’t really transposable: if we follow the melodic sequence, and transpose down by a diatonic second, we should begin our second phrase on the iii of G major (Bm7). But how to then reproduce the first phrase’s harmonic content, of one root, that supports first a major and then a minor chord? Bm7 is already ‘minor’ -- it can’t get any more ‘minor’ -- so there’s no where to go, without changing root (which is precisely the path that Klenner chose). While changing root allows the phrase to move forward, it does not create a harmonic sequence. Here’s the dilemma, written out:

We could reproduce the first phrase’s harmonic content by moving to the III of G minor (Bbmaj7):

. . . but that leads us into serious difficulties: first, the song’s departure from G major is meant to be temporary – we are meant to return to the key of G major by the beginning of the second phrase, at bar 5. Beginning that phrase with Bbmaj7 keeps us in the key of G minor, which throws the modality of the whole piece into question. Second, while the change from
C maj7 to Cm7 reflects the typical mode change from G major to G minor, the change from Bb maj7 to Bbm7 reflects a mode change from G minor to
G locrian -- something totally foreign to the era and style of this song.

So let’s take a look at how Klenner did harmonize this phrase, and see if we can trace his musical logic:

The move from Cmaj7 to Cm7 is the very common sudden mode change – from major to parallel minor – that is often called “minor plagal”, and that can take a variety of forms. In the simplest situations, the iv7 returns to the I chord of the major mode; here, though, it is used as a modulatory pivot chord, acting not only as the iv7 of G minor, but also the ii7 of Bb major, which then moves to the V7 of Bb major. But has the song really changed key? In one sense no, but in another yes. Let’s take a look:

As you correctly analysed, the F7 is also acting as a bVII chord in the key of G minor. As such, it is part of the chord progression — bVII – i — that is commonly used to create the equivalent of an ‘authentic cadence’ in music composed in aeolian (‘natural minor’) and dorian modes. As we see here, this formula is frequently ‘borrowed’ from the parallel minor key as a way of extending the “minor plagal” effect. However, the bVII does not move to a i (Gm), but to a I (Gmaj7) – ie, the music undergoes another sudden mode change, this time back to the original major key. This particular kind of last-minute, minor-to-major mode change – always on the cadential tonic -- is hundreds of years old, and even has a name: the Picardy Third. It is a very handy harmonic device, that allows us to make excursions to minor-key-
related chords and still find our way back home, to the original major tonic, at phrase’s end.

In the important sense, then, there has been no modulation – only two mode changes (from G major to G minor, and back again). I believe that there is a modulation happening, though, but on another level. We’ll take a look at that idea later on, once we’ve examined more of the song’s harmony:

You identified the Bbm7 and Eb7 as acting in relation to an Ab7 — the tritone substitution chord for V, which is built on the lowered supertonic, or “bII7”. The Bbm7 is ‘ii of bII’, and the Eb7 is ‘V of bII’. Many jazz musicians would agree with your analysis, for the convention is that one can easily add a “V of . . .”, and a “ii of V of . . .” to anything. I would agree with that practice — as long as it meets one important condition: that the harmonic result make sense within the larger musical context. If we take a look at the present context, however, we encounter some difficulties. First, the chart shows us that the ii of bII – V of bII progression does not lead to the expected bII (Ab), but directly to a natural ii (Am7), producing a very dissonant root movement of a tritone (Eb – A) in the process:

On its own, this progression is raw and awkward. It can work — in particular contexts — but are any of these musical contexts present in this passage? Let’s examine one of these contexts, and then find out whether it — or a variant of it — occurs in our piece.

Our first step is to consider an alternative approach: had the ii of bII – V of bII progression first resolved to bII (rather than ii), the passage could have easily progressed to the ii – V that follows, as part of a sequence:

Here, the complete ii – V – I pattern is repeated a minor 2nd lower, as part of a two-stage, descending sequence: the first ii-V-I relates to the bII (Ab), while the second ii-V-I relates to the actual tonic (G). As in any sequence, the listener’s primary attention is directed to the steps created by each transposition — in this case, to the bII and I that are the respective goals of those steps.

As we know, Just Friends removes the goal of that first step – the bII – and skips directly to the ii (Am7), thereby truncating the sequence:

. . . and giving us that same, awkward Eb7 – Am7 progression:

However, that same ii of bII – V of bII progression — even without its resolution to the bII — could be made to work, were it to act within the context of a smaller, “ii – V” sequence:

Here, the ii of bII – V of bII progression occurs within the context of a chromatically descending “ii – V” sequence, and all awkwardness seems completely mitigated. Why? Perhaps it’s because sequence always creates a kind of container for musical events; and the steps of that container – the actual sequential transpositions – act as the markers of a larger musical context. The listener’s attention is strongly directed to those sequential steps, so that other musical elements become secondary in importance. In our chromatically descending “ii – V” sequence, our attention is directed to the scale degrees upon which those steps begin: the mediant (B), lowered mediant (Bb), and supertonic (A). The awkward root movements (tritones) that join the last chord of one stage with the first chord of the next are less problematic because they are less noticeable — they’ve become secondary to the steps of the sequence. We hear “3 – b3 – 2”, rather than V of ii — ii of bII, etc.

In the context of this sequence, each V chord – being unresolved – acts more as an embellishment of its preceding “ii”, than as an actual dominant seeking resolution. Each V refers, backward, to the preceding minor 7 chord, rather than forward, to its non-existent resolution. We can hear this by playing the same progression, but without those V chords:

The passage seems to be pared down to its essentials; the steps of the sequence are clear, and the V chords aren’t even necessary. To my ears, this iii – biii – ii (V) is the underlying harmonic structure; the V chords are decorative, not functional.

We see, then, that the composer of Just Friends seemingly had at least several harmonic options: using sequence, he could have resolved the V of bII to a bII, and then progressed smoothly to the actual ii – V; or, he could have placed his ii of bII — V of bII progression in the context of a chroma-
tically descending “ii – V” sequence, to move from iii, through biii, to ii.

We know, however, that he chose a different route. Is there some other musical context, in this passage, in which his V of bII — ii can work smoothly?

Here are the first nine bars of Just Friends, with no attempt made to reconcile the chords Bbm7 – Eb7 with the key of G (the thick black line indicates that those chords do not apply to the key of G):

We see two separate ii – V occurences: Cm7 – F7, which belongs to the key of Bb major, and Bbm7 – Eb7, which belongs to the key of Ab major. These are spelled out a little more clearly here:

At face value, bars 3 & 4 can be said to belong to the key of Bb major, and bars 7 & 8 to the key of Ab major.

Is there any relationship between these two ii-V progressions? Do they form a sequence? Given that they are separated by two bars of non-sequential material (bars 5 & 6) – in an unrelated key – one would think not. However, it takes only a slight change to the song’s harmony to produce very different results:

If we change bar 5’s G maj7 to a Bb maj7, the two ii-V progressions instantly become part of a large, two-stage sequence. Each stage of the sequence begins with a major 7 chord, which then changes mode to become the ii7 of the next key. Here’s what it looks like in stage one:

Via sudden mode change, G major’s IV becomes g minor’s iv. That iv becomes a modulatory pivot chord, existing simultaneously as g minor’s iv, and Bb major’s ii. The progression continues in Bb major, to close the first stage on that key’s V.

Stage two begins on I of Bb major, and mirrors every harmonic aspect of stage one:

Via the same type of mode change that occurred in stage one, the Bb maj7 is transformed to a Bbm7, which then acts as the pivot chord in a modulation to they key of Ab major, in a ii – V progression in that key.

The sequence ends here, but the progression does not move on to an Abmaj chord — it proceeds to the same ii chord (Am7) that sounded so dissonant and awkward when we first examined it, earlier in this lesson. We still have that Eb7 resolving to the Am7, but — for some reason — it sounds fine in this situation. Why? Once again, the answer lies in the use of sequence, and its power to direct the listener’s attention toward its own longer-term structure:

Because of this passage’s clearly articulated sequential structure, its main structural points are the opening Cmaj7, the Bbmaj7 of bar 5, and the Am7 of bar 9. Cmaj7 is IV within the key of G major; Bbmaj7 is the III chord of the key of G minor, and Am7 is ii in the key of G major. If we refer to the III chord of G minor as “bIII” in the general key of G (ie, G major and G minor, combined), we see — as the illustration tries to show — that our progression is simply an elaboration of the progression IV – bIII – ii in the key of G. This clearly delineated structure absorbs our attention, so that we have little trouble accepting the final Eb7 – Am progression, even though it is – on its own – musically awkward. The Am is an expected and welcomed step on the scalar descent from subdominant to (eventual) tonic.

Here we have yet another sequential context in which the V of bII – ii progression can work well. Yet Klenner avoided it, too!

So, just what was the composer’s intention? Was he in control of his musical resources, pursuing a specific effect, or was he just winging it, and not that concerned about the clarity of his musical gestures?

Let’s give Klenner the benefit of the doubt, and temporarily assume that he knew what he wanted, and how to achieve it. Since we don’t find the usual harmonic processes at work in the piece, let’s be prepared to look for the unusual. Let’s return to Just Friends as it really is – i.e., with a Gmaj7 for bars 5 & 6, not the more logical Bbmaj7:

If we’re prepared to accept the unusual, perhaps we can entertain the idea that the harmonic structure of Just Friends may be working on two different levels simultaneously: a surface level, that stays pretty much in G major, and a superimposed level, that gives the ii-V of bIII and ii-V of bII of our last sequential ‘experiment’.   Like this:

From bar 1 to bar 6, the lower level comfortably moves along in G major and g minor. At bars 7 & 8, it’s somehow ‘pushed aside’ (by the upper level) —
and then innocently reappears at bar 9, as though nothing out of the ordinary had ever happened.

Meanwhile, the upper level has been developing its own musical logic: as we saw, the first four bars can be heard as:

But rather than resolve to a Bb “I”, bar five slides back into G major, via the bVII – i cadential formula borrowed from the parallel minor and the Picardy Third. It’s as if the entire musical reference to Bb major/bIII of g minor had suddenly evaporated — bars 5 & 6 just don’t exist for the upper level. But wait – here comes that level again, with its strong reference to Ab major/bII of G, in bars 9 & 10, and this time it’s strong enough to overpower the lower level’s tonality – temporarily, at least.

In other words, we can imagine two musical trains of thought, that strangely co-exist: the lower level’s straight-forward “G-ness”, and the upper level’s strong allusions to the kind of 8-bar, two-stage sequence that would exist, had the composer used a Bbmaj7 chord in bar 5, rather than his Gmaj7. The lower level has ‘gaps’, yet somehow manages to hold onto itself, while the upper level seems to weave in and out of the picture, sometimes present, sometimes not.

Is this, possibly, what Klenner had in mind? If so, what musical effect did this kind of “poly-harmony” serve? Was it used as a means of poetic expression, or to create a kind of musical gymnastics? Or, did the composer simply inadequately address the implications of his chord choices?

So far, I’ve attempted to describe the harmonic structure as it appears in the tune. If you’ll permit me a little ‘analytic license’, though – for there’s no way that I can prove my theory – I’ll share a far less complicated perspective: I have a hunch that the passage is based on a very straight-forward, traditional chord progression, that Klenner may have ‘updated’ — so that it would more fashionably suit the musical times in which he lived.

Here’s that progression:

I think you’ll recognize this as a venerable old chord progression upon which countless jazz standards were built. Before we examine the link between this progression and Just Friends, though, I’d like to quickly review one of its components: the ‘common-tone diminished 7 chord’ — the A#°7 — that occurs in bars 7 & 8.

In its role as vii°7, the diminished 7 chord has a dominant function. However, the diminished 7 chord built on the raised supertonic is not a vii°7 — and it does not have a dominant function. Here it is in the key of G major, where that chord is spelled A#C#EG:

Here, the A#C#EG chord (in 3rd inversion) acts as a neighbour chord to the I. We can hear that the note G is common to both chords — hence the term “common-tone” diminished 7 chord, by which it is sometimes identified. (Some theorists call it a “non-dominant” diminished 7 chord, because of its non-dominant role.)

I’ve identified this chord as being built on the raised 2nd degree (#ii). The thinking behind this comes from Walter Piston, who, in his Harmony, follows this musical logic:

Piston sees this chord as being the result of chromatic melodic movement: the 2nd degree (root of ii7) and 4th degree (third of ii7) are each chromatically raised, and then resolve into I6 — the raised 2nd degree ascends to the 3rd degree (where it is the third of I), and the raised 4th degree ascends to the 5th degree (where it is the fifth of I).

In fact, it’s in precisely that role that this chord appears in the old swing gesture:

Musicians evidently felt that this chord, that results from chromatic move-
ment, could just as easily descend as ascend:

. . . or:

Then, someone, somewhere, decided to use a substitute chord in place of the I6 — the iii7 chord:

. . . and it stuck. (Of course, A#°7 is also vii°7 of Bm.) There are thousands of tunes that employ that progression.

To sum up: the diminished 7 chord that’s built on the raised 2nd degree of the scale can act as:

Now, back to that old chord progression upon which Klenner might have built the first ten bars of Just Friends:

What would the composer have had to do to turn this into the beginning of Just Friends?

  1. Bbm7 instead of A#°7:   biii7 seems to have been a standard chord substitute for the common-tone diminished 7 chord, especially in this descending, stepwise progression:
  2. I’m not sure how or why the substitution works, but it’s ubiquitous.
  3. Giving Cm7 its “V”:  according to common practice, m7 chords are often treated as being “ii7” of another scale degree (ie, of the scale degree that is a major 2nd lower than the root of that m7 chord), and so can similarly be followed by the corresponding “V” (of that same scale degree). In this case, Cm7 is treated as a “ii”, and F7 is added as the corresponding “V”:
  4. However, adding an F7 also creates an awkward, tritone root movement between that chord and the Bm7 that follows:
  5. and that may have been a contributing factor in Klenner’s decision to use the F7 as bVII of G, in his move back to the key of G:
  6. Giving Bbm7 its “V”:  Finally, substituting Bbm7 (biii7) for A#°7 (bar 7) may have given Klenner the idea of treating that Bbm7 as a “ii”, and then adding the Eb7 as its corresponding “V”:

Although we have no way of knowing whether this traditional progression – and these steps – played any part in the creation of Just Friends, I find the speculation worthwhile. The question of how a work is created can lead to fresh perspectives, occasional revelations, and – quite often – to inspiration.

Whether Klenner actually thought like this or not, we can be sure that others have used – and still use -- these kinds of harmonic techniques to similarly ‘update’ the chord progressions of old standards.

As usual, Gustave, please don’t hesitate to email any questions or comments that may arise. We’ve covered a number of techniques, here, that easily point to other questions and musical situations, and I’ll be happy to address those with you.

All the best,

Michael Leibson is a composer, music consultant, and music educator, who specializes in jazz and classical harmony. To learn more about him, click here; for information about studying with him, click here; and for copyright information, click here.