Music suffers from a unique copyright problem. The limited number of possible note combinations that make up melodies means it’s easy for new songs to duplicate existing songs unintentionally, leaving unwitting composers open to plagiarism lawsuits. A pair of musician/coders, one of whom is also a lawyer, think that they’ve found a solution. The idea is to brute-force compile every possible one-octave 12-note melody onto a hard drive, fixing it in tangible form and thus copyrighting it, then releasing it in the public domain. At first glance, this seems like a potentially clever idea.

But is it, really?

The Problem

Music has a copyright problem all its own. In other forms of art, there are billions upon billions of possible color combinations, or billions of ways to string words and characters together. It’s nearly impossible for one painting or written story to duplicate another exactly without intentional plagiarism being involved.

But there are a much more finite number of musical notes, and a much more limited number of ways to string them together. And even out of all those ways, far fewer still are aesthetically pleasing, and many of the songs in particular genres of music often end up sounding very similar because they use similar chord progressions. This means that “collisions” are much more likely—cases where composers independently stumbled upon the same series of notes for their own compositions.

As if that weren’t bad enough, human memory can play tricks. A songwriter might hear a song and enjoy it, but consciously forget about it—and years or decades later, write what they think is an original tune but turns out to be a virtual duplication of one from back in the day. Most famously, this happened to ex-Beatle George Harrison, whose 1970 song “My Sweet Lord” was found to have been “subconsciously plagiarized” from the Chiffons’ “He’s So Fine” from seven years earlier. These musical copyright issues, and that court case, were addressed in Spider Robinson’s science fiction story “Melancholy Elephants,” which argued against perpetual copyright.

It seems that, as years go by, we see more and more cases of one artist suing because another artist’s song was too close to theirs. In some cases, the plagiarism is fairly cut-and-dried—the producers of Ghostbusters had approached Huey Lewis to do a theme for their movie, and even used his song “I Want a New Drug” as a temp track for scoring the film, so it wasn’t exactly surprising when Ray Parker Jr.’s theme came out sounding very similar to that song. That case was settled out of court.

But then there are cases like Katy Perry’s “Dark Horse” sounding similar enough to Flame’s “Joyful Noise” that a jury awarded Flame $2.8 millionAre the two songs really all that similar? Not according to some.

The two songs don’t share the same melody, chord progression or drum groove, [popular YouTuber and musician Adam Neely] explains. But they do share a similar ostinato, which Merriam-Webster describes as “a musical figure repeated persistently at the same pitch throughout a composition.” Neely argues the ostinato isn’t a significant part of the song.

Even though Katy had never heard “Joyful Noise,” the case turned on whether the other songwriters could have heard and borrowed from it (consciously or subconsciously). And since it was popular enough to have over 3 million hits on YouTube, the jury was willing to believe that could have been the case.

This is another area where music copyright faces its own unique challenges. In other media, the hurdle to clear for exposure is much greater. When N.K. Stouffer sued J.K. Rowling for plagiarizing the term “muggle” from her obscure children’s book, it was obviously extremely unlikely that Rowling could have been exposed to that book. But  the ubiquity of music on streaming sites, radio, and YouTube, combined with the doctrine of “subconscious plagiarism,” means that juries are a lot easier to convince that someone could have been exposed to a song and just not consciously remember it.

The cases go on and on. There was the “Blurred Lines” copyright lawsuit, in which Marvin Gaye’s family successfully argued that the Robin Thicke/Pharrell Williams song infringed on Gaye’s “Got to Give It Up.” 

Many disagreed with the verdict since “Blurred Lines” and “Got to Give It Up” were only similar in feel and genre. According to the original suit, “the Gaye defendants are claiming ownership of an entire genre, as opposed to a specific work.”

In 1988, John Fogarty was actually accused of plagiarizing himself. When he was younger, he had made the mistake of selling the copyrights to his own compositions, and the owner of the record label that now held the copyrights (with whom he had a history of mutual enmity) felt that his new song “The Old Man Down the Road” was just the old Creedence Clearwater Revival track “Run Through the Jungle” with new words. Fortunately, Fogarty was able to convince a jury that it was only to be expected that two songs composed by the same writer would have a similar sound. He was even awarded legal fees, though he had to go all the way to the Supreme Court to get them (setting a precedent that would help the rare songwriter defendent who did win a plagiarism suit).

And in 2015, British singer Sam Smith’s Grammy-nominated “Stay With Me” turned out to be remarkably similar to Tom Petty’s “I won’t Back Down.” Even though Smith insisted he’d never heard Petty’s song (that had been recorded before he was even born), and Petty seemed to understand that the similarity was entirely accidental, Petty still ended up collecting half the royalties and co-writing credit on that song in an out-of-court settlement because that was less damaging to Smith than the expense of a lawsuit.

The effect of all these lawsuits is to make songwriters very nervous, because they never know when some “new” melody they came up with is going to turn out to be litigiously similar to something that already exists. How can even they be sure that they’re not subconsciously remembering and duplicating some fragment of something they heard years before?

A Solution…Or Is It?

This is the problem that lawyer Damien Riehl, also a musician and coder, set out to address with fellow musician and coder Noah Rubin. As Riehl explains in a TED talk, and a subsequent video with YouTuber Adam Neely, they wanted to disrupt musical copyright by generating all possible melodies that could be made with twelve notes from a single octave, and saving them as MIDI files. They used the same kind of “brute-force” technique used to guess passwords, creating every possible variation simply by starting with aaaaaaaaaaaa, incrementing to aaaaaaaaaaab, and so on. They ended up with 68.7 billion melodies.

Under US copyright law, any composition is automatically protected by copyright once it’s fixed in a tangible form, so the act of saving these MIDI files to disk theoretically copyrights them. Riehl and Rubin then released all these files, as well as the code used to make them, into the public domain (or, at least, under a Creative Commons Zero license, which is effectively the same thing).

But will this actually make a difference in future music plagiarism lawsuits? Lawyers and legal experts don’t think so. In another YouTube video, lawyer Leonard French explains that, while this might be a clever thought experiment, there are a number of hurdles it fails to clear.

Melodies generated in this way don’t meet the requirement of originality (though the algorithm that outputted them might); and crucially, copyright law allows two artists “to create the same thing, as long as they didn’t have access (…) copyright infringement means copying, or access and substantial similarity.”

And proving prior access to 68.7 billion melodies might get more than a little tricky.

Music created by algorithm might not actually be any more copyrightable than, say, a photo taken by a macaque. There’s no actual human creativity involved, any more than there is in compiling a list of names in a phone book. But more importantly, it doesn’t matter whether the melody exists somewhere as a MIDI file nobody ever actually listens to. That doesn’t make any difference to a courtroom argument over whether someone who wrote Song A had previously been exposed to a similar-sounding Song B. And there’s also the fact that, in compiling every possible 12-note melody, they’ve also infringed basically every song that’s ever been recorded already themselves.

(Also, judges tend to have a low tolerence for attempts at being clever–so someone who tried to introduce this into an actual court case would probably hurt their own cause more than they’d help.)

That being said, the experiment has gotten plenty of media coverage over the last few days, and gotten people discussing just how absurd music copyright has become. Under the doctrine of subconscious plagiarism, it’s now incumbent on anyone accused of plagiarism to prove a negative—that they never heard the music they’re accused of infringing. How can you even do that? Hence, the deck is stacked in favor of anyone accusing someone who came later of ripping off their song. Bringing more publicity to that issue certainly can’t hurt. Indeed, even Riehl doesn’t really want to see it adjudicated in court—he told The Atlantic that he hoped Congress could address the issue instead.

(But good luck with that—Congressional copyright legislation of any kind seems to be effectively a moot point right now.)

As for trying to undercut the mechanism of music plagiarism lawsuits, Riehl and Rubin’s bold project probably won’t cut it. But if it gets people talking and brings more awareness to the issue, at least that’s something.

Featured photo by Ylanite Koppens on

If you found this post worth reading and want to kick in a buck or two to the author, click here.