In 2022 I called this lawsuit silly. Silly it remains. Judge Wesley Hsu doesn’t necessarily disagree with me, but the process seems determined to.
The latest order denies Sam Smith and Normani’s motion for summary judgment in Sound and Color, LLC v. Smith, the case claiming their 2019 hit “Dancing With A Stranger” copied a 2015 song of the same name by Jordan Vincent and the production duo SKX. It’s worth reading in full, especially in the context of what I covered in 2022, because it seems to turn on things I didn’t fixate upon: a particular facet of the hook’s melodies, and the samples inside the plaintiff’s own song.
First, a quick recap of the now years-long case. Sam Smith won summary judgment in 2023. The Ninth Circuit reversed in 2025, and I wasn’t thrilled. Then, in June of this year, the plaintiff withdrew its access argument. (No wide audience, no chain of custody stuff, no music-video similarities.) That leaves one route: the plaintiff has to prove the songs are so similar that independent creation is virtually impossible.
Harmony is gone too, which is sensible since my 2022 piece laid the two chord progressions side by side, and every chord was different. The plaintiff’s experts no longer contend that the chord progressions, the drum beat, or the overall feel are evidence of striking similarity. That argument is over. What remains is six syllables: “dancing with a stranger.”
Let’s begin with the fact that “dancing with a stranger” as a concept is unprotectable, and as an expression deserves very little protection. Ideas get no copyright at all, and four words describing a person dancing with someone they don’t know is an idea stated in about the only way it can be stated. Both sides agree the phrase appears in more than fifteen songs that predate Vincent’s, including in the title and hook of at least eight of them and as I pointed out four years ago, Cyndi Lauper recorded a song called “Dancing with a Stranger” in 1989. There’s a film by that name!
The melody that sustains
Actually, there is no “the melody.”
I’ll explain.
The plaintiff transcribed the respective hooks by scale degrees. (think Do-Re-Mi, but count 1-2-3 instead. Do to Ti is seven notes, right?)
And if even one reader comes back at me with minor solfege, so help me… it matters not. We’re using numbers anyway.
The claim is that Vincent’s melody is scale degrees (notes) 7-6-6-5-4-4-3. Smith’s is 7-5-5-5-4-4-3-3. And the phrases’ first syllables start on the same beats, with the final resolution landing half a beat later in Vincent’s.
Those numbers obviously just count down, right? So both are pieces of a descending scale, and the parties agree on that.
And yes, the phrases do start on the same beats if we frame things a certain way:
Vincent’s song is a two-line phrase. Smith’s is four. Vincent sings, “She said I’m gonna die, dancing with a stranger.” and it repeats this, usually four times. So “danc-” has a rhythmic placement on the “and of four” of the second, fourth, sixth, and eighth measures of its chorus. Smith sings, “Look what you made me do, I’m with somebody new, ooh baby baby, I’m dancing with a stranger,” a four-line stanza, and thus, its “danc-” lands on the “and of four” of its thirteenth measure, and if he repeats the whole thing, it’ll come back around on the twenty-sixth measure as well.
So, if you zoom in on just that phrase, they both come in on an “and of four.” But if you zoom out, they land in very different places within their respective choruses.
Another ax for me to grind: The defense asked the court to compare the pitches alone, then the rhythm alone, then the metric placement alone, then the pitch contour alone. Judge Hsu declined. “Disassemble any two works enough and neither would resemble anything.” He leaned on Ambrosetti v. Oregon Catholic Press and its line about not separating “each piece of glass within the kaleidoscope.” And that’s fine, except this logic is rarely applied the other way when it’s scaffolding for a bogus selection-and-arrangement claim.
I called the Ninth Circuit’s Ambrosetti revival “a resurrection of a non-case.” And now here it is out there resurrecting stuff itself!
I’ll be damned.
I’m gonna show my work
Remember I said, “There is no the melody?“
I transcribed every full statement of the hook in both songs, not just the one exemplar the court is working from, every “Dancing with a stranger” that appears in each work.
Vincent’s contains the phrase sixteen times, in four melodic variants!
Smith’s states it fourteen times, in eight variants, seven variants of which occur only once.
Plaintiff’s focus is on not the first but the second most common of the four variants. How common is the second? Five of sixteen.
Plaintiff’s Sam Smith transcription elects the most common of the eight, but it’s seven out of fourteen.
Vincent, “Dancing with a Stranger” (2015), four melodic paths, sixteen statements
| Path | Scale degrees | Times sung |
|---|---|---|
| J1 | 7 6 6 7 7 5 | 6 |
| J2 | 7 6 6 7 4 4~3 | 3 |
| J3 | 7 6 6 5 4 4~3 | 5 |
| J4 | 7 6 6 5 7 5 | 2 |
Smith and Normani, “Dancing With A Stranger” (2019), eight melodic paths, fourteen statements
| Path | Scale degrees | Times sung |
|---|---|---|
| S1 | 7~5 5 5 4 4~3 3 | 7 |
| S2 (duet, upper) | 7~5 5 5 4 4~5 5~4~3 | 1 |
| S3 (duet, lower) | 7~5 5 5 4 4~5 3~4~3 | 1 |
| S4 (duet, upper) | 7~5 5 5 4 4 4~5~4~3 | 1 |
| S5 (duet, lower) | 7~5 5 5 4 4 3~5~4~3 | 1 |
| S6 | 4 4 4 5 4~3 3 | 1 |
| S7 | 5~7 5 5 4 4 4~5~4~3~4~3 | 1 |
| S8 | 7~5 5 5 4 4~3 3~1~3~1~7, 1 | 1 |
That J3 version, the one shown in this decision? It’s not Vincent’s first, second or third melody. It appears as his 4th, 8th, 10th, 12th, and 16th choices.
Vincent’s version has four different melodies for expressing the six syllables in the unprotectable lyric and title, “Dancing with a stranger.” Smith has eight. Thirty statements sung between them. Not a single match between the two songs.
Vincent’s most popular variant appears six of sixteen times, but ends 7-7-5, far less comparable to Smith’s most used, and the one selected for comparison, which ends 4-4-3.
Four of Smith’s fourteen statements aren’t even a single melody, but duets with Normani. But no matches there either, even when I look at Normani’s notes as separate, plausible opportunities to find a match.
Every one of Vincent’s sixteen statements opens on 7 and steps down to 6. Twelve of Smith’s fourteen statements open on 7 and leap down to 5. Plaintiffs have an argument for this difference — that the 6s are passing notes, and as this decision accepts, “passing tones are defined as unessential.” Often, yes, that’s a valid point; some notes are more significant than others, and a passing tone passes, connecting two different notes and getting out of the way. Vincent’s 6 mostly doesn’t. It’s sung twice, on two different syllables, and this, in every one of the song’s sixteen statements. And hey, actually, now that we’re here, in most of the sixteen, the line goes 7-6-6 and then straight back to 7. So, most of the time it’s not a passing tone, but a “neighbor tone,” and these 6s aren’t all that quick about it.
It’s the melody. It’s part of the identity of each song’s phrase. These second pitch choices are the one thing about each song that, across 28 of the 30 expressions, never changes.
These songs share descending scales, one of the blockiest of building blocks, not the same, set to a phrase found in more than fifteen earlier songs.
Across all thirty statements, Vincent’s sixteen and Smith’s fourteen, not one matches the other. Not one. Zero for thirty.
One more callback to that 2022 article of mine. In 2022 I mentioned the plaintiff’s claim that if you slow Vincent’s track to Smith’s tempo, the keys match. I said I was very skeptical but couldn’t be bothered to check. The defense’s expert Paul Geluso did check and he says they don’t match.
Today, I could be bothered. Guess what? He’s quite right. If you slow Vincent’s down to Smith’s tempo, Vincent’s track moves to E minor, not Smith’s F minor.
One more thing — samples
Evidently, Vincent’s song contains two unlicensed samples: a full-measure vocal part from Masters at Work’s “The Ha Dance” and a two-measure portion of Lyn Collins’ “Think (About It).” Both loop through nearly all of the song, including under every instance of the hook. Section 103(a) of the Copyright Act says protection for a work that uses preexisting material unlawfully doesn’t extend to the part of the work where that material appears. The samples run under the whole hook, so the defense argued the hook has no protection at all.
That’s clever, but to be fair, musicologically I’d find that pretty irrelevant. Legally though, I’m fascinated. As I understand it, the court said the defense measured the wrong song. Under Newton v. Diamond, substantiality is measured by the significance of the copied portion in relation to the original work. Newton is famous. It’s a Beastie Boys case: a three-note flute figure looped throughout “Pass the Mic,” held to be de minimis, which is Latin for “who the hell cares?” The defense showed how much of Vincent’s song the samples occupy rather than how much of “The Ha Dance” or “Think (About It)” they took. Then Judge Hsu listened for himself. He found both samples “difficult to identify or even hear” at their low volume, and held that a jury could find them de minimis. I might be a little ticked off at the volume thing. I can move a fader and help make it easier to hear for you, Your Honor. But that wasn’t the point. The court said the question was wrong. (Note to would-be Musicologize plaintiffs: if your song contains uncleared samples, expect that to come up. Definitely divulge that.)
