July 14, 2021 Musicology No Comments

Still awaiting its day in court but Musicologize knows the arguments and has already reached judgment.

“Let’s Get It On” versus “Thinking Out Loud” is probably the most high-profile and interesting case out there. And since it follows a series of others that left the state of copyright law seemingly in disarray the question is whether this will bring balance or further chaos. Countless articles were written in the wake of “Blurred Lines,” “Dark Horse,” and “Stairway To Heaven” trying to get a handle on things. If “Blurred Lines” and “Dark Horse” struck fear in creatives as it was often reported, then the reversal of “Dark Horse” and the upholding of “Stairway” supposedly returned us to a steadier footing. We’ll see.

The north star in all this is that copyright law serves the important purpose of encouraging creativity through an admittedly delicate balance of protecting rights holders while enabling an appropriate derivation of prior works in new ones. Copyright is noble. It’s worth defending and preserving. And preserving it means getting cases like this right.

Thus this deep-dive, probably my last look before these cases (plural, yes, there’s more than one here; three actually) go to trial (hope, hope, hope at least one goes to trial). We’ll ignore the minor differences among the cases. We’ll economize and aggregate a bit. We will cover what the plaintiffs likely plan to argue and consider how reasonable it all seems, particularly as it pertains to the wellbeing of copyright itself. And if anyone wants to take issue, by all means, comment below.

First, the pretty short version.

I don’t have to bury the lede too much here. Sheeran should prevail. There’s nothing substantial enough from “Let’s Get It On” that appears also in “Thinking Out Loud” that’s original and protectable by copyright. That’s where this ends up. But the process matters.

Here first are the two songs. Not because you haven’t heard them before, but just as a refresher, an initial reference point.

Now, before any legal scholars shout at their screens, let’s address the fact that the recording of “Let’s” is probably not material to the case. Copyright law from a hundred years ago, applied in both the Blurred Lines and Stairway To Heaven trials, did not consider recordings as published works but instead regarded the sheet music deposit copy filed with the copyright office as the protected work. So in both of those cases, the plaintiffs had to find similarity not with the famous recordings we all know but with whatever notes were on the deposit copy. You can read plenty on Musicologize about how that went. It may go that way here as well. Filings in the cases so far are largely about whether the recordings should be excluded from the cases.

I’m of the mind that the law has been inadequate and that since copyright vests in the creator at creation, it’s artificial to then narrow the scope to a (usually) one-page sheet music distillation. It’s complicated admittedly. And my sympathy doesn’t matter. It’s just an aside. For our purposes let’s assume the plaintiffs get everything they want in that regard. They play the record in court. They’ve got it all. What points will they make, and why should they still expect to lose?

Mostly, plaintiffs will point to the similar but not identical chord progression, similar harmonic rhythm, similar bass parts, and similar drum parts, all of which appear or they’ll argue are implied by the sheet music deposit copy.

The defense will say the chord progression is similar but not identical, nor original, and was in common use prior to Lets Get It On. All that is true too of the bass and drums, and moreover neither appear in the deposit copy.

The defense is right in every way that matters.

Now the long, long musicologized version.

By far, the most compelling plaintiff argument is that some basic elements, collectively the “groove,” or “feel” or one might argue the propulsive engine of the song — are indeed similar. Critics of the “Blurred Lines” decisions, including the dissent from the appellate court, have said that case made it possible to own a “groove” or “style” — and this was a substantial shift in the established boundaries of copyright. Others, chiefly the plaintiffs, would argue this fails to recognize that broad and substantial similarity was proven to the jury’s satisfaction.

This case is less silly than Blurred Lines. The argument in the “Thinking vs Let’s” complaint avoids the word “groove,” deconstructs the groove, and examines and compares its individual components. It’s more impressive that way maybe. A pile of evidence or really anything else looks more substantial when it’s all spread out, right?

Foremost of those components is the respective chord progressions. Much will be made of this, so let’s look at that first. The main progressions in the two songs are these:

"Let's Get It On" chords:    Eb  Gm    Ab   Bb.
                             I   iii   IV   V

"Thinking Out Loud" chords:  Eb  Eb/G  Ab   Bb.
                             I   I/^3  IV   V

I might’ve skipped the roman numeral stuff. But if you’re wondering what this little bit of basic music theory is:

The diatonic chords (forgive the redundancy) in the key of Eb are, in order, numbered one through seven...
Chord Names:    Eb   Fm   Gm   Ab   Bb  Cm  D°
Roman Num:      I    ii   iii  IV   V   vi  vii°

That little carrot character thing I typed into the Thinking progression above means “degree.” When I put a slash in the chord name, I am specifying the bass note. In this case, it’s letter name “G” which is the third degree of the Eb major scale. (Eb major scale is Eb F G Ab Bb C D Eb) Get it? If not there’s my very quick theory lesson over here.

Obviously, three of the four chords in both songs are identical; only the second ones are different. How much difference does that one chord make to the overall similarity of the four-chord progression?

I expect the plaintiffs will argue it makes almost no difference, and not because it’s only one chord out of four, but because the two different chords sound hardly different at all; I’d argue they’re mostly equivalent; they’re interchangeable.

Q: How can chords that are not identical be similar, equivalent, and interchangeable?

A: By having some notes in common, and by having the same “harmonic function,” which is a music theorist’s way of saying, “they both make the listener feel kinda the same way at hearing them and impart the same set of expectations for what chords and notes might be coming next.”

Here’s a quick look at the notes that make up the two chords.

The notes in an Gm chord are...  G   Bb   D.
The notes in a Eb/G chord are... G   Bb   Eb.

So the plaintiff will be right that the four chords are very nearly the same. The chords are not identical though, indicative perhaps that they weren’t copied, but musically, they’re arguably interchangeable.

Further, they will argue, sometimes in music a bass note of a chord has so much harmonic function CLOUT that the other notes are of lesser relevance. In other words, the “G’ in the bass in both songs is far more important than the other notes accompanying it. And the different notes Eb and D by virtue of being immediately adjacent notes only a half-step apart sound very alike? And the plaintiff argues that if G is in the bass and is reinforced with a G in the melody, you’ll possibly imagine you’re hearing “iii” even if you’re not because of the overtone series. When you play G, one of the more prominent overtones is D. He offers that in “The Weight” by The Band, when Levon Helm sings the first line and lands on the word “Nazareth,” the listener might imagine they hear “iii” in the accompaniment even if “I/^3″ is what’s actually played. Then he cites the Commodores’ “Easy” as well. He makes the correct observation that ^3 is both in the bass and vocal of the words “Nazareth” and the second syllable of “Easy.” Wanna hear those?

Let’s evaluate. I would agree that the bass note is sometimes the most persuasive, but it depends upon the functional strength of the accompanying notes, case by case. I’ll give a case in a moment where the accompanying notes are super persuasive and not easily overwhelmed.

As for the half-step apart nearness of Eb and D making them interchangeable? Odd, I’d say; misleading at best. Adjacency creates, as often as not, the very least equivalence available in music. Indeed half-step relationships are often the best example of contrast and directional necessity. Consider for example, “Shave and a haircut, two bits.”


The notes in question are “Two” and “Bits.” We showed our two songs in the key of Eb, where “Two” and “Bits” would be “D” and “Eb” respectively. They’re right next to each other on a piano, but could they sound much less interchangeable to you?! Me either. One is the solid landing spot and the other is shifting sand.

Moving onto “The Weight” and “Easy,” perhaps I’m missing what’s persuasive about this. Yes, G appears in both the bass and in the melody, and yes, the overtone series of that G would include D, but…

  1. the pitches of the overtone series are largely imperceptible.
  2. The Weight and Easy actually employ iii; nobody needs to imagine hearing the overtones; Im missing something here?
  3. Giving some rope though, would they sound okay with Eb/G instead of Gm? The Weight would sound weird to me, but I have music ears, and those ears know it’s supposed to be iii. But I would expect “Easy” to sound weird to anyone. You know those soothing “ah ah ah ah’s” that fall between “That’s why I’m easy: and “I’m easy like Sunday morning?” The iii chord falls on the second “ah;” along with a “D” in the melody. If you play Eb/G instead of Gm, it’s gonna make a world of diff.

But when he points out that in the verses of Credence Clearwater Revival’s “Who’ll Stop The Rain” Mr. Fogarty alternately employs both Gm and Eb/G over the third lines of consecutive stanzas, the point gets more traction with me.

The third line of verse one “Clouds of mystery pourin” stays on an Eb chord — a repetition of the Eb played over the first line “As long as I remember,” — but the third line of the second verse “Caught up in the fable” is accompanied by Gm — a contrast to the first line, “Went down to Virginia” (or something like that.) So the point seems to be that both chords provide the same overall arch to the verses they serve. And I agree.

Perhaps the defense will argue that since it’s only four chords across just four measures of music (though repeated throughout the song) musicological norms would demand that the similarity be identical or very nearly so to be deemed significantly similar. Sometimes we hear of copyright being “thin.”

It gets interesting immediately.

When a “I” chord changes by bass note only to become a “I/^3″ you haven’t added any new tones. You’ve only changed which of the three notes is heard lowest. This has the effect, functionally, of making the chord less stable, less comfortable, meaning the listener expects the chord to change soon rather than persist. But we’ve changed the quality of the I chord only very minimally.

When a I chord moves to a iii chord, the added note D is a new color, with greater meaning and greater instability. And you’ve abandoned the note for which the key is named Eb, again greater instability. You’ve changed the chord more than the previous example, but still minimally.

So this is relative. Both chords are still fairly stable and both are mostly we call “Tonic,” the most stable harmonic function. “Tonic” is where most songs begin and end. Tonic is like harmonic home base. I should clarify, the so-called “tonic” chord is the I chord. But the iii chord and vi chords, each with two notes in common with the I chord, are considered by most musicologists to be tonic in function.

Talk about tedious. But that’s the discussion we’re looking at.

What if the chord written in the deposit copy is wrong?

This would be fun.

Again on this same topic of how similar these second chords are, plaintiff expert report makes an interesting observation. When “Let’s” chord moves from Eb to Gm, an Eb note occurs in the melody much of the time whereas D notes occur much more rarely. The expert concludes that this Eb note implies the second chord in “Let’s” is or at least could be Eb/G after all.

That’s not crazy, but the argument has weaknesses.

The counter arguments are that, first, where the sheet music contains Eb notes in time with Gm chords symbol, this more readily conveys that you are expressly NOT to play Eb/G. Is that not the default meaning — to very deliberately play “iii” with its tonal color ^7 “D?” Is it not the greater logical leap to infer otherwise — that the “D” while expressly stated is unimportant to the composition?

I’d point out also that the “C” and “Gb” notes you find sung over “iii” chords are the more prominent out of place melody notes in that circumstance. But they don’t support the story that I/^3 is the “true” harmony.

If we’re determined to blame the sheet music scribe, taking into account the C and the Eb, ignoring the bluesy Gb’s, I might’ve labeled the chord as Ebmaj13/G. Then again I could’ve called it a “Gm6 add 11” or something less supportive of the plaintiff’s point. Frankly, the sheet music has a more plain-spoken style than any of that.

On balance, my Ebmaj13/G opinion supports the plaintiff’s argument somewhat, but I think it cuts both ways.

That’s PLENTY about whether “I/^3” is equivalent to “iii.”

We need to look at harmonic rhythm.

And then, get back to the elephant, that in my opinion wherever we land on that relative equivalence, the bigger question has to be “SO WHAT? It’s only a four-chord accompaniment!”

But before we do that, let’s see if there’s anything else that matters. The gold standard of copyright in music after all is MELODY. Are the melodies similar?

These are the kinds of arguments anticipated for the melodies being significantly similar, again paraphrased by me for brevity, and hopefully clarity.

The first point I’ve seen is that the melodies enter on the same note “G,” and rhythmically at the same time as that contentious second chord.

This is true though one might brush it aside. Trying to establish similarity from the placement of one identical note, when the very next note doesn’t match is a tough row to hoe. But I would argue that rests are important as well. And the one and a half beats of rest that precede those “G”s are a part of the plaintiff’s observation.

One expert report observed that both melodies begin on ^3, ascend stepwise through ^4, and to ^5 before falling stepwise. This is truer of “Let’s” than of “Thinking,” but no. No, they don’t. I’m really not in the gotcha business and I could be missing something. There might be an explanation. Could be an honest mistake. Or maybe there’s a way to interpret this such that it’s sensible. But I don’t see it.

It continued that both melodies descend stepwise from Bb down to Eb. And here again, I must be missing something.

There’s an observation that “Let’s” uses bluesy Gb’s (that’s ^b3) and that “Thinking” similarly finds it’s way to ^3, both as dissonant sevenths over the IV chords. 

This is more contrast than similarity. First, while ^3 and ^b3 might appear similar, they’re two different notes, just as “two” and “bits” were. And moreover, remember just a few points ago when they argued that a melody note should inform how we think of its accompanying chord? Here we’re pointing to”G” and “Gb,” which by the same token would result in two different chords in the progression, “Abmaj7 and Ab7. A new dissimilarity to contend with.

"Let's Get It On" chords:    Eb  Gm    Ab7      Bb.
                             I   iii   IV       V

"Thinking Out Loud" chords:  Eb  Eb/G  Abmaj7   Bb.
                             I   I/^3  IV       V

The plaintiff report notes that both verse melodies end on the note Eb against Bb in the accompaniment.

This is arguably true, and arguably interesting.

There’s a lengthy argument that again the conspicuous ^b3’s in “Let’s” are equivalent to the less bluesy ^3’s in “Thinking.” And that “Thinking” is a just a less bluesy cover of “Let’s.”

I always reject that sort of logic. Different melodies are not made the same through the premise that you might’ve created one melody through consideration of the other. Different is different however it comes to be.

8. “Both songs use the same pitch relationship in the same voice for the same ornamental function relating the same scale degree to the same non-chord tone.”

He means “Let’s” employs ^6 as an ornament of ^5. This is true as far as it goes. In the latter half of the chorus, the second time Gaye sings “Let’s Get It On,” “On” lands on C and resolves to Bb one beat later. The report is likening this to an ornament in Thinking. But in “Let’s” it’s mostly a “suspension.” Gaye lands on the non-chord tone “C” on the strong beat and releases to the chord tone Bb. It’s a basic musical device.

Sheeran’s ornament is rhythmically different. It’s an anacrusis — a couple notes at the end of the phrase that clear the way for the first strong beat of the next phrase, a “pickup.” Also, IT’S NOT IN THE MELODY. It’s a little guitar figure! 

This is a mostly inconsequential observation.

9.  “The Gaye – Townsend composition’s upper-voice emphasis on ^5 and ^3 shown throughout the deposit copy is unusual when supported by the iii chord, which is more characteristic with ^7 in the upper voice.” 

Trying to imagine where this is going, but there are three chord tones in a iii triad, ^3 ^5 and ^7 and the notion that two of them are unusual is bizarre on its face.

The point though is that neither of these songs use ^7 very much over three. This harps on the argument that iii and I/iii are the same thing. So the chord progressions are the same. 

But there’s a logical fallacy here.

Sheeran doesn’t sing ^7 and he doesn’t play iii. He’s consistent. You can try to argue that Gaye’s tune might’ve been written that way as well, except that the ^7 and its iii are plain as day in the recording and on the sheet music.

It’s hard to make the point that Sheeran is not doing things Gaye OVERTLY DID in service to your copying argument.

And Different Drum plays the same four chords for two verses, a loop, NEVER hits ^7. Red Rubber Ball? Never. Georgy Girl? Nah. Jean? Nope.

Seems to me you hit ^7 if it’s what you want, and don’t if it isn’t.

Red Rubber Ball, Georgy Girl, Puff The Magic Dragon, Different Drum, Jean, My Cup Runneth Over, some stupid movie theme called “Brother Of The Wind” that just popped into my head from when I was like 5 — this progression is simple and basic; arguing otherwise is obtuse.

I – iii – IV – V is literally the first chord progression I can remember being excited about as a kid.

10. In relation to other parts of songs, the I – iii – IV – V progression class may appear as part of any number of functions within the formal structure.

The report lists a dozen or so uses of I iii IV V in other songs, again anticipating and pre-arguing, presumably to show that Let’s And Thinking are different from those, yes, but similar to each other.

10. Most four-chord loops (repetitions of the four-chord pattern) do not contain the iii chord;”

Even I can’t maintain indulging them. This is statistically insignificant, and a few song you just mentioned use it repeatedly.

12. It is very unusual to repeat the I – iii – IV – V progression as a loop without contrast, and beyond rare to do so throughout the entirety of both verses and choruses;

Okay, yeah, the Commodores’ “Easy” might be the only such example known to me other than our Gaye – Townsend and Sheeran exhibits. But originality through interminable simple repetition? Ew. Especially forty years ago.

It used to be unusual for songs to be this static in general. Nowadays it’s common. So expect to find this and every other common progression comprising the entirety of pop songs.

13. One challenge Sheeran has is finding other occurrences of the chord progression with the syncopation. The database has only a couple. Georgy Girl is the main one.

This argument needs to be put down though.

Is syncopation a common musical device? Yes.

It’s so pedantic to even explore this. A thousand tunes should be written over that progression and that groove without one significantly copying another. That’s how copyright is supposed to work.

What we really have is the Tonic Chord for a four-count, modified slightly for another four-count, followed by a SubDominant Chord and then a Dominant Chord. The widely regarded “most common chord progression” I vi IV V is built in an equivalent manner. I can substitute here and there, show you ten variations in five minutes all having that very familiar, not at all novel chord function pattern. Rhymically, I might them played straight, swung, syncopated or combinations thereof and you’d regard them as common — the stuff of western popular music. The stuff most of us whether we realize it or not expect and prefer to hear as the basis for most everything on our pop Spotify channel.

The defense is going to say, “none of that is protectable, so tell us what IS protectable in “Let’s” and is common to both songs.” And there will be nothing but light smoke, if that. The plaintiffs need to defeat the question, for they cannot satisfy it. The melodies and lyrics are very different. And they’re the only parts of either of these songs that are original and protected by copyright.

And all this by the way is us looking at the two recordings — which is almost certainly NOT going to happen in court. Instead, they’re more likely going to be comparing Thinking to the sheet music deposit copy of Let’s. There are no drum parts indicated there. There is far less musical information on the page. The plaintiffs will need to argue that there are notes implied via musical shorthand. They will try to get the recording heard in court. Laypersons can be persuaded that similarity, any similarity, is theft. It’s not true. But that’s court. I don’t see this happening.

But I could’ve cut to this legality and skipped most of the musical analysis. I didn’t because unlike in the Stairway case, I don’t see it as terribly significant. The deposit copy of Let’s is a mostly sufficient representation of Let’s. The expected exclusion of the audio recording does NOT leave the plaintiffs extremely shorthanded. I’m not so sure they aren’t better off. That’d be a longer discussion.

Ultimately the two songs are alike only in ways that are simple, basic, common, and not worthy of protection by copyright. So, frankly, this is just silly; so silly that Sheeran, who I’d say has been known to give in to spurious assaults in the past, does not appear interested in doing so here. Nor should he.

Written by Brian McBrearty