According to Billboard, Luke Combs’ “Fast Car” was the eighth biggest song of 2023. Needless to say, the success of the song has likely made a lot of people happy. It’s likely made Luke Combs happy. It’s likely made his label happy. It’s also likely made the millions of listeners happy. But the song’s success has also made another person very happy: Tracy Chapman.
Fixing copyright is certainly on the SB reform agenda. Thank you for writing this piece.
My family had to deal with a bunch of music copyright law when handling my aunt and uncle's estate and it was not pleasant. My aunt was a performer while my uncle was a producer, writer, and performer.
Really interesting, especially compared to something I just learned yesterday on Blocked & Reported*... they gave an update on the “Bad Art Friend” lawsuit, when apparently the judge has ruled that Sonya Larson did not plagiarize Dawn Dorland because the several hundred words she copied verbatim from Dawn’s kidney donor letter were used with a different “character of use” in Larsons short story. Apparently because the words were used with clearly different intent of expression than the original letter, it doesn’t count as plagiarism? And because Dawn suffered no financial loss, arguably gained from it, she also loses standing to claim damages.
The vast majority of music sampling would meet both a “character of use” test as well as not cause financial damages to the original artist. Seems interesting that the written word would be treated so differently than a composition or recording? Even if you argue the written word is more like a composition than a recording, at least the composition has compulsory licensing.
Sorry to be That Guy, but based on my read of the order it's not that it "doesn't count as plagiarism," but that it's that it's fair use (as to which both the "Character of the Use" and "Effect on the Market Value of the Work" are factors in making the determination), for which the gloss is more like "this implicates the copyright holder's rights, but it's okay and there's no remedy warranted."
In layman's terms instead of copyright-legalese this is closer to "it's plagiarism but that's okay" than "this isn't plagiarism."
Fair use is one of the fuzzier parts of copyright law and typically an affirmative defense, which makes it hard to know in advance if an appropriator is coloring inside the lines because the scope of per se rights is subject to this four-factor balancing test, although some uses are significantly more likely to be protected than others.
It does seem to me that there’s some logic for the original artist to be able to deny sampling (while being unable to deny a cover).
A cover is their whole song, faithfully reproduced as the author notes, effectively in the same context as the original artist recorded it.
A sample is a small snippet, taken out of context, and by definition not faithfully reproduced.
What if an artist doesn’t want their music associated with a rap recording (because of, say, objections to violent references in the lyrics)? That feels quite reasonable.
While I generally agree with the author and thought he made a compelling case, I would have liked to see him do more to address some potential counter arguments.
A sample is by definition faithfully reproduced, because it is the recording itself (though it can have additional modifications). A cover has much more flexibility - just look at Johnny Cash’s cover of “Hurt”
And as in the case of Cash's "Hurt" often the cover is far better than the original and vastly increases its popularity. One wonders if Leonard Cohen should have paid Jeff Buckley for his cover of "Hallellujah" rather than the other way around. (Same for "A Little Help from My Friends" and countless others.)
"What if an artist doesn’t want their music associated with a rap recording (because of, say, objections to violent references in the lyrics)? That feels quite reasonable."
My instincts pull the opposite direction. If a license is compulsory, then you have no responsibility or capability to police how people use samples. If it is voluntary, then you open yourself up to scrutiny every time you approve use of a sample, which incentivizes denying a license even when you don't object.
This is a good point - but fundamentally the logic of copyright is to give the creator control. Whether there is scrutiny of what the creator does with that control is second order. And since the author is making a case for why there should be an exception to creator control in this instance, he could have addressed some of the objections to giving up control.
Yes, but the reason that's the logic is that creator control is supposed to incentivize creation in general. This post is largely arguing that the current logic, when applied to sampling, does the opposite.
I appreciate the concern you outlined, but it doesn't make any sense to me. An artist can't prevent Gary Glitter from recording a faithful cover of their song, but they should be able to stop a three-second snare sample from being used in a rap song's hook because they interpret the lyrics to be violent? Seems backwards, if anything. To be clear, I agree with you and the author, and both licenses should be compulsory.
Yeah, it makes me think of "Who I Smoke" by Jacksonville rapper Julio Foolio. He samples Vanessa Carlton's "A Thousand Miles" on a song where he explicitly brags about the multiple people he or his crew actually murdered in real life. If I was Vanessa Carlton, I would not be happy about this, and would want a way to stop it!
Now granted it did happen so perhaps she was fine with it, or maybe she doesn't think its worth the hassle to fight the legal battle to stop, but still, I think I'd want that option to stop this from happening.
Except Weird Al seeks permission for his works, both presumably because he doesn't want to spend money defending court cases even if he'd win them and because many of his songs probably wouldn't actually pass the SCOTUS standard for being parodies. (Yankovic arguably lucked out with "Amish Paradise," which infamously involved confusion over whether Coolio had authorized it or not, because it's pretty squarely within the parody exception.)
The parody exception is not "you can use something copyrighted without permisison in a parody". It is "you can use something copyrighted without permission if you are parodying that copyrighted work". If the target of the parody is not the work itself that you are using, then it's not a fair use.
Most Weird Al songs are parodying something, but they are often using a popular song to parody something else.
To the extent that I have any opinion, it seems more logical that both written music and lyrics and the actual song produced should be under sole control of the creator. Why have a centrally-fixed price for licensing for the former at all?
But in turn, copyright terms need to be curtailed. 50 years, inheritable only by a surviving spouse, child, or grandchild with no extension possible.
But honestly, I don't really see it as important and don't know if I'm smoking something.
One reason for compulsory licensing of covers is that it also applies to public performance. Without it, it would be a copyright violation to sing (or otherwise perform) any copyrighted song in a public place. That would make buskers and karaoke bars illegal, as they would need song-by-song licensing. Similarly, there are plenty of musicians who will do a live cover of a different song every night on their tour. That's only legal because they will use compulsory licensing for cover songs.
Releasing a recorded version ("phonorecord" to use the language of the copyright acts, which is where ℗ comes from) is different and I could see an argument ab initio for restricting that to operating with permission. But it's absolutely against the traditions of the music industry for the last century, and changing that would be a major change to the law.
Though I doubt most cover bands, let alone buskers, are actually paying licensing fees in the first place unless they get good enough to tour in their own right.
I don't think the former paragraph becomes impossible here. If my half-assed understanding is right, there are cataloguing/intermediary firms who build lists of music for the sorts of purposes you've described and lease them to end-users, then ensure licensing fees go where they should. If that's correct, then doesn't an end to mandatory licensing just require those firms to get permission from/negotiate prices with the record labels and artists with whom they already communicate and to whom they already send money?
Most music venues carry licenses with "performing rights organisations" (PROs) and just take a note of the setlist for any given night and send it with a fee to the PRO, which then distributes the money to the songwriters. BMI and ASCAP are the main US ones. For bands playing songs they wrote themselves, professional venues will generally require them to register with a PRO and then send the licensing fee to the PRO and the songwriter gets the money later (rather than just paying the songwriter as part of the band's fee, which is legal if they haven't signed with a PRO, but is bad practice and will make their next audit a nightmare, especially if the songwriter later signs with a PRO).
This very much applies to small cover bands playing in music venues, ie the sorts of small places that small bands play. Playing in somewhere that doesn't carry a PRO licence because they aren't really a music venue, like a bar or a mall, is a different story; the PROs vary from time to time on how strictly they enforce on places like that. But even they will often carry a PRO licence for a jukebox in a bar, or for piped-in music in a mall, and it's easy to then notify the PRO of an occasional live performance and pay the relevant fee (which probably isn't much for a one-off in a small venue). Karaoke bars very much do carry PRO licences and keep records of which songs are performed and pay fees. I'm sure buskers don't, though I think some places that license buskers (e.g. London Underground) do include music licensing in the busking licence.
But compulsory licensing means that it's simple: there's a standard schedule of fees and you don't have to check lists of what can and can't be licensed. If you want a good example of how this becomes a nightmare, try having college TV broadcast a concert from a college venue. Because there are moving images synchronised with the music (ie the band performing), you are required to have a synchronisation licence. Even though it's only appearing on a few hundred screens in dorms that have the college's TV channel added to the cable TV.
You negotiate with the band and agree a fee with them for the concert and include the rights for the TV broadcast in that fee. Then you get their setlist. If they write all their own songs then you just say to them "no covers tonight, guys", and you're OK. But if they don't write all their own songs, you have to clear the setlist. So you contact the PROs and they will say "these songs, we have a standard fee; these songs, the publisher (ie record company) has a standard fee; these songs, you negotiate with the publisher; these songs, you negotiate with the songwriter". Then you pay the standard fees and you contact the publishers and songwriters. If you can't get hold of them, or you can't get an agreement in time, or you can't afford what they want you to pay, you now have a choice: shut down the broadcast when that song is on, or tell the band you can't clear that song and they have to adjust the setlist. For a small college venue, you're not going to get the band to change their setlist, so you just have to have three minutes of dead air in the middle of the concert. Then you have to tell the band that they can't deviate from the setlist because you had to clear it for copyright and if they play anything else (unless they wrote it themselves) then it's illegal (or you shut down the broadcast for that song).
This is what every concert would be like if you got rid of mandatory licensing. It would kill mid-sized venues. Not from the costs of the licensing, but from the cost of the bureaucracy. It would make it illegal to spontaneously play a song. I can't think how many bands will spontaneously add a song to their act when something happens: for instance, when Shane MacGowan died recently, loads of people played "Fairytale of New York" in his honour. Having to clear the rights to that with Jem Finer and Shane MacGowan's estate (which was almost certainly not a well-run machine the day after he died) would be a nightmare.
In a normal (ie not TV broadcast) concert, if a song gets played that isn't on a PRO list, you have made a "reasonable effort" to find the songwriter (by checking with the PROs), so you can pay the mandatory license fee to the Copyright Office and then the songwriter can collect that if they find out. But if you need a synchronisation licence (ie to film the concert for broadcast), and they're not on a PRO list, then you need to contact them or their publisher. If you can't find who their publisher is, and you can't contact them, then there's no "reasonable effort" defence - you just can't legally play the song.
What you'd end up with is lists of songs you are and aren't allowed to play, but the biggest venues and bands would be able to play the songs that everyone else isn't allowed to, because they could afford the lawyers needed to negotiate custom licences for them only, while your average 100-ticket venue would just have to ban anything that isn't on a PRO list.
This is a great comment. It was outside the scope of what I was trying to cover, but one of th difficulties is that the licensing is so vastly different for music in all types of situations that it can quickly become the administrative nightmare you describe.
I'd add that the idea that musicians, artists etc should have some kind of moral claim to control who uses their work is deeply offensive to scientists and mathematicians.
Our theorems and theories are every bit as much the result of hard work and creativity (and no they aren't just facts about the world -- the same laws can be phrased many different ways but some are more elegant and useful). However, no one ever suggests we should have moral rights to determine who gets to use them.
Yes, limited copyright for artistic expression is a worthwhile incentive for creation and we don't think it's a good tradeoff in the case of scientific discovery. That's fair enough. But it's unreasonable to suggest that artists should somehow have moral control over their creations beyond what's the societally optimal incentive while not giving that same control to mathematicians and scientists (I obviously think no one should get it).
Totally agree with this. There is no other form of intellectual property where anybody would even fathom you having those rights past your death ... other than trademarks which are a whole different thing and very different imo
That’s true. I wasn’t careful enough with my phrasing. Music is generally life of the author + 70 years. Patents are 20 years to my understanding. So the minimum length for music is 70 years. Assuming someone lives an average life span, you are looking at protection for close to 150 years which I think is hard to justify. If it were a short fixed term like a patent, the persisting after death does make sense.
I was actually almost done writing this as a response to Ethics Gradient when your response popped up, so I'll put it here instead:
Right of publicity (a.k.a. "name, image, likeness rights") also survives death for some period of time as a matter of law in most, if not all, US states, sometimes longer than copyright. (Indiana allows it for 100 years post-death; Tennessee technically only allows it for 10 years post-death, but it can be indefinitely renewed by the estate/heirs.)
Trade secrets last as long as the criteria for qualifying as such are met and thus theoretically can last forever, not being dependent on the life of the original creator at all.
Moral rights under US law are the only type of intellectual property that specifically terminate on the creator's death. (And I suppose also "reputation," if you classify that as a form of intellectual property, since defamation claims can't be based on post-death statements about the decedent.)
I admire the Bold Decision made by Chris Dall Riva to delve into musical copyright law without invoking Taylor Swift and her "Taylor's Version" covers of her own past albums.
Her old label controlled (and sold) their rights to her original recordings. But now Ms. Swift controls the rights to her cover versions. If a movie, television or advertising exec wants to stay in the good graces of the Swifties, they'll license Taylor's Version for their own purposes and she needn't share royalties with whomever owns the masters of the original versions.
Do compulsary license payments go to the songwriter or the person who owns the recording? Like, is Taylor paying compulsary license fees to the owner of the recordings?
This reminds me of a weird feature of Arabic pop music: Love songs are usually--if not almost invariably--written in the second person masculine,* even when the singer is a man and even when the lyrics make as much or even more sense as being addressed to a woman. (And no, these aren't same-sex love songs, this is the Middle East we're talking about.) To me, who grew up with this stuff, it's perfectly normal, but it throws off a lot of language learners to recognize that Amr Diab's impassioned "habibis" in "Nour El Ain" are treated as being addressed to a woman only by the universal unspoken agreement to default to the masculine in songwriting. I'm not sure if this convention is to make it easier to market songs (since Arabic singers usually don't write their own songs, having a one-gender convention makes it easy for lyricists to write for both male and female singers) or arises from rhyme/prosody considerations, but it's there.
*Note: Yes, Arabic distinguishes gender in the second person. As do most Semitic languages.
Yeah, Nour El Ain was super-popular when it came out and even trickled in marginally onto Western radio marginally. If you've ever heard an Arabic pop song, it's probably that one.
Is there a different number of syllables in the masculine or feminine second person? Are there traditional rhyme or alliteration schemes that might be changed by shifting the gender?
It can affect the number of syllables depending on the word and the dialect/register. It also can affect the rhyme; for instance, in Egyptian and Levantine Arabic, it changes the quality of the vowel preceding the final -k that marks the second-person possessive (i.e. "your X") and in most varieties of Arabic it changes the final vowel of second-person past-tense verbs.
To be honest whenever I see articles about copyright centered on sampling (and there are a lot, for obvious reasons), there’s always a certain Onion-headline quality to it: “Art form based on wholesale appropriation inexplicably finds itself having difficulties with copyright.”
Less snarkily, because I think at a minimum the arguments about transaction costs are good ones, I wonder if the author has considered the effect this would have on, e.g., sample-based instrumentation licenses or VLSI’s, which AFAIK have a robust private market that would presumably be somewhat harmed by compulsory licensing.
Also 30 seconds seems like a long time - that’s going to be 1/5 to 1/6 of a typical song length, easily enough to get the entire chorus or hook in there. I’d suggest at most 10.
“Art form based on wholesale appropriation inexplicably finds itself having difficulties with copyright.”
It’s a funny quip, but this applies to a significant portion of rock music in the 70s. I’m a huge Zeppelin guy but a decent portion of the catalogue is just an appropriation of Howlin Wolf and others.
I don’t see any ethical or moral daylight between that and sampling.
I've been down so many Zeppelin youtube black holes. Love them. My only hot take with the "appropriation" bit is ... ok, but just listen to Page and Bonham. I don't think it's hyperbolic to say Bonham's grooves were unlike anything that had ever been played before.
Great point on the 12-bar blues. Thought this Polyphonic video raises interesting authorship questions on how the blues is built on folk tradition and folk tradition is more collective. There's a 1941 interview with Muddy Waters (6 minute mark) where he says he learned his song from Son House which was similar to Robert Johnston's Walkin' Blue but the song was out way before Robert recorded it.
+1 re the Onion headline issue, as I have a similar reaction. I'm not going to claim sampling is an inferior artform, but, as lawyer, all I can say is that I would tell a client they should get authorization from the necessary rights holders before sampling a work unless they have a really clear fair use defense (which will rarely be true for commercially released music) or otherwise have a solid argument for why copyright shouldn't even apply, and you probably should still get authorization because it will almost always be cheaper than defending a lawsuit.
Man, paper planes still absolutely slaps. Also great post! I worried that at the start it was going to be about people complaining about 'appropriation' or something but it was actually really interesting.
I think the cultural discussion around “appropriation” and the post-Napster copyright discussion could learn a lot from each other, and complicate the simplistic views that many people take of one or the other.
This argument doesn’t work for me. Treating a composition and a small portion of a recording of that composition as equivalent is not sensible. You even kind of acknowledge that by using a trivial example (the hi hat part), rather than a relevant one, like say the famous guitar lick. In the current scheme, those different sounds would be wildly different costs to license, for obvious reasons, but your proposal treats them as equivalent. That makes this a non-starter.
I don't think the piece argues that the costs should be the same, but that samplers and covering artists should be have the same rights to pay a fee for a non-negotiable license. That's completely compatible with samples costing more
I see the value in a set formula for non-negotiable licenses, but I think disinterested makes a good point with the famous guitar lick example. If we use a set time based formula it will be used to relatively cheaply license short but iconic parts of songs. I don't think it makes the whole idea a "non-starter" but it is worth considering.
To argue with myself, you could use the same argument to do away with compulsory licensing for songs as well, since the perceived value of a cover of "stairway to heaven" is much greater than a cover of some lesser known Led Zeppelin song. I don't really have a good reason to say why this is OK but compulsory licenses for samples is not.
Why does that make it a non-starter? What's so bad about a world where that happens? I don't think it's going to substantially reduce the incentive to create new music and if it means we get to use great bits of music more broadly that's great.
I’m not talking about covers in my comment. I’m saying different samples from the same song can be wildly different in terms of their usage and recognizability. If you sampled the hi-hat, as the article says, that is a trivial sample that no one would recognize. If you sampled the guitar, people would hear it as “the guitar from fast car” and arguably you should owe more royalties because it’s a key component of the song.
My point is that a scheme that treats those samples as equivalent is not one recording artists would support.
This is a great piece, as someone in the indie electronic music industry who also DJs. I never thought I'd see a Francis Grasso shout out in Slow Boring, what a pleasant surprise!
The idea that authors should retain some kind of control over how their works are used (eg moral rights) is one of the worst ideas of all time and I don't understand why many otherwise left leaning individuals support it. Many European countries are happy to apply a high tax rate to earnings but yet allow an artist to deny other people the benefit of reusing their creation.
All of human society is built on using other people's ideas and innovations. Why should this be different because you filled a copyright. People who come up with policy ideas don't get to decide who can use them (eg if you suggested the Obama/RomneyCare idea you can't say one party can't use it). Scientists don't get to deny other people the use of their laws.
The world is made better off when we can all build on the ideas that go before us. And it's not like it follows from any philosophical notion of private property (eg on a Lockean perspective asserting copyright is actually violating other people's property rights to make what they want).
Yes, copyright is important to encourage innovation but that's the only justification.
I largely agree, but the very literal protection from someone copying your work word for word and selling it shouldn't be allowed for some amount of time... Maybe 25-50 years tops.
Agreed. I think ideally intellectual property protections are very strong for some limited amount of time. For a long time it was 28 years with an option to extend another 28, which feels like an ample amount of time.
I agree that it's desierable that some kind of compensation be given for that use to incentivize creation but I don't think that (in an ideal world) that should give the author the ability to, say, refuse to sell their work in a particular area or format even within that time period.
For instance, even if you believe that your work is designed to be read if you refuse to allow an audiobook or if you try and refuse to let your story be included in an anthology I think it's totally appropriate that there be a compulsory licensing scheme where the government says, ok you get this much in royalties per copy of the work but they get to create an audiobook/anthology even if you don't approve.
To be clear, this is what is ideal, I realize in actuality there will be all sorts of concerns and complications which might make this a bad idea. But it shouldn't be ruled out in principle.
Very interesting and unexpected piece. I certainly didn't think I was going to learn so much about sampling when I woke up this morning. I'm especially grateful for learning where the beginning of "Paper Planes" -- a piece I love -- came from.
And it got me to thinking. How did earlier periods handle this kind of thing? And I thought specifically about that masterpiece, Brahms' "Variations on a Theme by Haydn" and how he got the rights from the long-dead Haydn, only to learn this from Wikipedia:
"Brahms composed the work on a theme entitled "Chorale St. Antoni", which Brahms found in a wind ensemble composition. When Brahms discovered it, the wind ensemble piece carried an attribution to the composer Joseph Haydn. Brahms titled his own composition accordingly, crediting Haydn for the theme. However, music publishers in the early nineteenth century often attached the names of famous composers to works by unknown or lesser-known composers, to make the pieces more saleable. Subsequent research has concluded that the wind piece Brahms used as a source does not fit Haydn's style. The wind ensemble piece remains without clear attribution."
Loved the read, thanks! The bit about samples containing samples is craaaazy. In a couple decades this scheme could have literal hundreds or thousands of people getting paid out for a huge web of samples of samples of samples. Not intractable but logistically challenging.
It's worth saying that, while Sinéad O'Connor could make the recording of Nothing Compares 2 U using a compulsory licence, she did need a synchronization licence from Prince for that iconic music video.
That is wild, that the video requires extra licenses not needed for the recording, even though the licenses are only about things that are already used in the recording!
The video is legally considered to a be a short film that uses the music. In order to use music in a film, you have to license the film (same for TV and ads), and those require voluntary licences, one for the phonorecord (the sound recording - that's what the P in a circle symbol stands for on your record labels) and one for the composition (lyrics and music, which may have separate copyright holders).
A lot of TV made before the rise of the boxset used only licensed the music for broadcast (ie not for reproduction onto media) or only licensed it for a period of time, which has meant that a lot of TV made before 1995 or so was unable to be released on home media or streaming, or was released with the music changed (sometimes with a cover of the original song if the composition rights could be cleared but not the phonorecord rights, sometimes with an entirely different song). Film had home media as a major part of the revenue from far earlier on (VHS rentals, etc) and also has always had far larger budgets, much more likely to pay for music to be composed, and much more able to afford to buy rights to music it used.
Music videos are exactly in the same situation, but, obviously, the phonorecord synchronization is never an issue, and they do usually have the ability to get rights from the composer if they have the budget for a music video - it's record companies negotiating with each other at that point, usually.
Note that amateurs uploading covers to YouTube do technically require a synchronization license from the composer (but YouTube has negotiated blanket terms with the performance rights organisations so this is not normally an issue in practices). Putting the cover anywhere other than YouTube is legally more dubious, but in practice doesn't get pursued.
Thanks for having me! I really enjoyed writing this piece.
Fixing copyright is certainly on the SB reform agenda. Thank you for writing this piece.
My family had to deal with a bunch of music copyright law when handling my aunt and uncle's estate and it was not pleasant. My aunt was a performer while my uncle was a producer, writer, and performer.
This rules, and I’d love to see more guest posts like this about niche topics!
Noted!
Would have never been exposed to this ...well written and interesting!
Really interesting, especially compared to something I just learned yesterday on Blocked & Reported*... they gave an update on the “Bad Art Friend” lawsuit, when apparently the judge has ruled that Sonya Larson did not plagiarize Dawn Dorland because the several hundred words she copied verbatim from Dawn’s kidney donor letter were used with a different “character of use” in Larsons short story. Apparently because the words were used with clearly different intent of expression than the original letter, it doesn’t count as plagiarism? And because Dawn suffered no financial loss, arguably gained from it, she also loses standing to claim damages.
The vast majority of music sampling would meet both a “character of use” test as well as not cause financial damages to the original artist. Seems interesting that the written word would be treated so differently than a composition or recording? Even if you argue the written word is more like a composition than a recording, at least the composition has compulsory licensing.
*Timestamp 34:00-36:15 or so: https://open.substack.com/pub/blockedandreported/p/premium-how-the-allies-took-berlin (if you’re not a primo and want a free trial sub, drop your email, I think I have 3 more to share.) Or you can find the ruling here: https://www.bloomberglaw.com/public/desktop/document/LarsonvPerryetalDocketNo119cv10203DMassJan302019CourtDocket/2?doc_id=X4C1JG8DIN58IVP76AC7631DIMB
Sorry to be That Guy, but based on my read of the order it's not that it "doesn't count as plagiarism," but that it's that it's fair use (as to which both the "Character of the Use" and "Effect on the Market Value of the Work" are factors in making the determination), for which the gloss is more like "this implicates the copyright holder's rights, but it's okay and there's no remedy warranted."
In layman's terms instead of copyright-legalese this is closer to "it's plagiarism but that's okay" than "this isn't plagiarism."
Fair use is one of the fuzzier parts of copyright law and typically an affirmative defense, which makes it hard to know in advance if an appropriator is coloring inside the lines because the scope of per se rights is subject to this four-factor balancing test, although some uses are significantly more likely to be protected than others.
Thank you for being That Guy, this helps!
It does seem to me that there’s some logic for the original artist to be able to deny sampling (while being unable to deny a cover).
A cover is their whole song, faithfully reproduced as the author notes, effectively in the same context as the original artist recorded it.
A sample is a small snippet, taken out of context, and by definition not faithfully reproduced.
What if an artist doesn’t want their music associated with a rap recording (because of, say, objections to violent references in the lyrics)? That feels quite reasonable.
While I generally agree with the author and thought he made a compelling case, I would have liked to see him do more to address some potential counter arguments.
A sample is by definition faithfully reproduced, because it is the recording itself (though it can have additional modifications). A cover has much more flexibility - just look at Johnny Cash’s cover of “Hurt”
https://m.youtube.com/watch?v=8AHCfZTRGiI
or Alanis Morisette’s cover of “My Humps”.
https://m.youtube.com/watch?v=VJg4rwDkkBA
And as in the case of Cash's "Hurt" often the cover is far better than the original and vastly increases its popularity. One wonders if Leonard Cohen should have paid Jeff Buckley for his cover of "Hallellujah" rather than the other way around. (Same for "A Little Help from My Friends" and countless others.)
Cash’s version of “Hurt” is actually my go-to for testing headphones and earbuds.
Lots of samples involve interpolations, which aren’t really faithful reproduction.
"What if an artist doesn’t want their music associated with a rap recording (because of, say, objections to violent references in the lyrics)? That feels quite reasonable."
My instincts pull the opposite direction. If a license is compulsory, then you have no responsibility or capability to police how people use samples. If it is voluntary, then you open yourself up to scrutiny every time you approve use of a sample, which incentivizes denying a license even when you don't object.
This is a good point - but fundamentally the logic of copyright is to give the creator control. Whether there is scrutiny of what the creator does with that control is second order. And since the author is making a case for why there should be an exception to creator control in this instance, he could have addressed some of the objections to giving up control.
Yes, but the reason that's the logic is that creator control is supposed to incentivize creation in general. This post is largely arguing that the current logic, when applied to sampling, does the opposite.
I appreciate the concern you outlined, but it doesn't make any sense to me. An artist can't prevent Gary Glitter from recording a faithful cover of their song, but they should be able to stop a three-second snare sample from being used in a rap song's hook because they interpret the lyrics to be violent? Seems backwards, if anything. To be clear, I agree with you and the author, and both licenses should be compulsory.
Yeah, it makes me think of "Who I Smoke" by Jacksonville rapper Julio Foolio. He samples Vanessa Carlton's "A Thousand Miles" on a song where he explicitly brags about the multiple people he or his crew actually murdered in real life. If I was Vanessa Carlton, I would not be happy about this, and would want a way to stop it!
Now granted it did happen so perhaps she was fine with it, or maybe she doesn't think its worth the hassle to fight the legal battle to stop, but still, I think I'd want that option to stop this from happening.
I'd likely want that option too, if I were an artist. But "what do artists want?" is a different question from "what is best for society?"
I think the first and most obvious counterargument to that would be that parody is protected.
The "Weird Al" exception.
Except Weird Al seeks permission for his works, both presumably because he doesn't want to spend money defending court cases even if he'd win them and because many of his songs probably wouldn't actually pass the SCOTUS standard for being parodies. (Yankovic arguably lucked out with "Amish Paradise," which infamously involved confusion over whether Coolio had authorized it or not, because it's pretty squarely within the parody exception.)
Yeah, a lot of people don't get this:
The parody exception is not "you can use something copyrighted without permisison in a parody". It is "you can use something copyrighted without permission if you are parodying that copyrighted work". If the target of the parody is not the work itself that you are using, then it's not a fair use.
Most Weird Al songs are parodying something, but they are often using a popular song to parody something else.
To the extent that I have any opinion, it seems more logical that both written music and lyrics and the actual song produced should be under sole control of the creator. Why have a centrally-fixed price for licensing for the former at all?
But in turn, copyright terms need to be curtailed. 50 years, inheritable only by a surviving spouse, child, or grandchild with no extension possible.
But honestly, I don't really see it as important and don't know if I'm smoking something.
One reason for compulsory licensing of covers is that it also applies to public performance. Without it, it would be a copyright violation to sing (or otherwise perform) any copyrighted song in a public place. That would make buskers and karaoke bars illegal, as they would need song-by-song licensing. Similarly, there are plenty of musicians who will do a live cover of a different song every night on their tour. That's only legal because they will use compulsory licensing for cover songs.
Releasing a recorded version ("phonorecord" to use the language of the copyright acts, which is where ℗ comes from) is different and I could see an argument ab initio for restricting that to operating with permission. But it's absolutely against the traditions of the music industry for the last century, and changing that would be a major change to the law.
Fair point.
Though I doubt most cover bands, let alone buskers, are actually paying licensing fees in the first place unless they get good enough to tour in their own right.
I don't think the former paragraph becomes impossible here. If my half-assed understanding is right, there are cataloguing/intermediary firms who build lists of music for the sorts of purposes you've described and lease them to end-users, then ensure licensing fees go where they should. If that's correct, then doesn't an end to mandatory licensing just require those firms to get permission from/negotiate prices with the record labels and artists with whom they already communicate and to whom they already send money?
Most music venues carry licenses with "performing rights organisations" (PROs) and just take a note of the setlist for any given night and send it with a fee to the PRO, which then distributes the money to the songwriters. BMI and ASCAP are the main US ones. For bands playing songs they wrote themselves, professional venues will generally require them to register with a PRO and then send the licensing fee to the PRO and the songwriter gets the money later (rather than just paying the songwriter as part of the band's fee, which is legal if they haven't signed with a PRO, but is bad practice and will make their next audit a nightmare, especially if the songwriter later signs with a PRO).
This very much applies to small cover bands playing in music venues, ie the sorts of small places that small bands play. Playing in somewhere that doesn't carry a PRO licence because they aren't really a music venue, like a bar or a mall, is a different story; the PROs vary from time to time on how strictly they enforce on places like that. But even they will often carry a PRO licence for a jukebox in a bar, or for piped-in music in a mall, and it's easy to then notify the PRO of an occasional live performance and pay the relevant fee (which probably isn't much for a one-off in a small venue). Karaoke bars very much do carry PRO licences and keep records of which songs are performed and pay fees. I'm sure buskers don't, though I think some places that license buskers (e.g. London Underground) do include music licensing in the busking licence.
But compulsory licensing means that it's simple: there's a standard schedule of fees and you don't have to check lists of what can and can't be licensed. If you want a good example of how this becomes a nightmare, try having college TV broadcast a concert from a college venue. Because there are moving images synchronised with the music (ie the band performing), you are required to have a synchronisation licence. Even though it's only appearing on a few hundred screens in dorms that have the college's TV channel added to the cable TV.
You negotiate with the band and agree a fee with them for the concert and include the rights for the TV broadcast in that fee. Then you get their setlist. If they write all their own songs then you just say to them "no covers tonight, guys", and you're OK. But if they don't write all their own songs, you have to clear the setlist. So you contact the PROs and they will say "these songs, we have a standard fee; these songs, the publisher (ie record company) has a standard fee; these songs, you negotiate with the publisher; these songs, you negotiate with the songwriter". Then you pay the standard fees and you contact the publishers and songwriters. If you can't get hold of them, or you can't get an agreement in time, or you can't afford what they want you to pay, you now have a choice: shut down the broadcast when that song is on, or tell the band you can't clear that song and they have to adjust the setlist. For a small college venue, you're not going to get the band to change their setlist, so you just have to have three minutes of dead air in the middle of the concert. Then you have to tell the band that they can't deviate from the setlist because you had to clear it for copyright and if they play anything else (unless they wrote it themselves) then it's illegal (or you shut down the broadcast for that song).
This is what every concert would be like if you got rid of mandatory licensing. It would kill mid-sized venues. Not from the costs of the licensing, but from the cost of the bureaucracy. It would make it illegal to spontaneously play a song. I can't think how many bands will spontaneously add a song to their act when something happens: for instance, when Shane MacGowan died recently, loads of people played "Fairytale of New York" in his honour. Having to clear the rights to that with Jem Finer and Shane MacGowan's estate (which was almost certainly not a well-run machine the day after he died) would be a nightmare.
In a normal (ie not TV broadcast) concert, if a song gets played that isn't on a PRO list, you have made a "reasonable effort" to find the songwriter (by checking with the PROs), so you can pay the mandatory license fee to the Copyright Office and then the songwriter can collect that if they find out. But if you need a synchronisation licence (ie to film the concert for broadcast), and they're not on a PRO list, then you need to contact them or their publisher. If you can't find who their publisher is, and you can't contact them, then there's no "reasonable effort" defence - you just can't legally play the song.
What you'd end up with is lists of songs you are and aren't allowed to play, but the biggest venues and bands would be able to play the songs that everyone else isn't allowed to, because they could afford the lawyers needed to negotiate custom licences for them only, while your average 100-ticket venue would just have to ban anything that isn't on a PRO list.
This is a great comment. It was outside the scope of what I was trying to cover, but one of th difficulties is that the licensing is so vastly different for music in all types of situations that it can quickly become the administrative nightmare you describe.
Like Eminem sampling (without permission) French pianist Jacques Loussier’s song Pulsion on the track ‘Kill You’
https://en.wikipedia.org/wiki/Kill_You_(Eminem_song)
Except you can use for free if you make it a parody.
See 2 live crew
I'd add that the idea that musicians, artists etc should have some kind of moral claim to control who uses their work is deeply offensive to scientists and mathematicians.
Our theorems and theories are every bit as much the result of hard work and creativity (and no they aren't just facts about the world -- the same laws can be phrased many different ways but some are more elegant and useful). However, no one ever suggests we should have moral rights to determine who gets to use them.
Yes, limited copyright for artistic expression is a worthwhile incentive for creation and we don't think it's a good tradeoff in the case of scientific discovery. That's fair enough. But it's unreasonable to suggest that artists should somehow have moral control over their creations beyond what's the societally optimal incentive while not giving that same control to mathematicians and scientists (I obviously think no one should get it).
Totally agree with this. There is no other form of intellectual property where anybody would even fathom you having those rights past your death ... other than trademarks which are a whole different thing and very different imo
Patent terms are also a fixed number of years regardless of whether the inventors or original claimants are living or dead.
That’s true. I wasn’t careful enough with my phrasing. Music is generally life of the author + 70 years. Patents are 20 years to my understanding. So the minimum length for music is 70 years. Assuming someone lives an average life span, you are looking at protection for close to 150 years which I think is hard to justify. If it were a short fixed term like a patent, the persisting after death does make sense.
I was actually almost done writing this as a response to Ethics Gradient when your response popped up, so I'll put it here instead:
Right of publicity (a.k.a. "name, image, likeness rights") also survives death for some period of time as a matter of law in most, if not all, US states, sometimes longer than copyright. (Indiana allows it for 100 years post-death; Tennessee technically only allows it for 10 years post-death, but it can be indefinitely renewed by the estate/heirs.)
Trade secrets last as long as the criteria for qualifying as such are met and thus theoretically can last forever, not being dependent on the life of the original creator at all.
Moral rights under US law are the only type of intellectual property that specifically terminate on the creator's death. (And I suppose also "reputation," if you classify that as a form of intellectual property, since defamation claims can't be based on post-death statements about the decedent.)
I admire the Bold Decision made by Chris Dall Riva to delve into musical copyright law without invoking Taylor Swift and her "Taylor's Version" covers of her own past albums.
Her old label controlled (and sold) their rights to her original recordings. But now Ms. Swift controls the rights to her cover versions. If a movie, television or advertising exec wants to stay in the good graces of the Swifties, they'll license Taylor's Version for their own purposes and she needn't share royalties with whomever owns the masters of the original versions.
https://wjlta.com/2023/03/07/copyright-law-taylors-version/
Do compulsary license payments go to the songwriter or the person who owns the recording? Like, is Taylor paying compulsary license fees to the owner of the recordings?
I just love that Combs didn't change the line that goes "Now I work in the market as a checkout girl."
It’s always interesting to see where people do and don’t edit gendered terms in a gender reversed cover!
Here’s a fun listicle of some covers (both by queer and straight artists) that kept the pronouns:
https://www.buzzfeed.com/skarlan/girls-want-to-steal-jessies-girl-too
This reminds me of a weird feature of Arabic pop music: Love songs are usually--if not almost invariably--written in the second person masculine,* even when the singer is a man and even when the lyrics make as much or even more sense as being addressed to a woman. (And no, these aren't same-sex love songs, this is the Middle East we're talking about.) To me, who grew up with this stuff, it's perfectly normal, but it throws off a lot of language learners to recognize that Amr Diab's impassioned "habibis" in "Nour El Ain" are treated as being addressed to a woman only by the universal unspoken agreement to default to the masculine in songwriting. I'm not sure if this convention is to make it easier to market songs (since Arabic singers usually don't write their own songs, having a one-gender convention makes it easy for lyricists to write for both male and female singers) or arises from rhyme/prosody considerations, but it's there.
*Note: Yes, Arabic distinguishes gender in the second person. As do most Semitic languages.
And here's the song for anyone who isn't familiar with it: https://www.youtube.com/watch?v=KLJA-srM_yM
(Now that I'm listening to it, I think it does sound familiar, though I definitely didn't know the song by name or description.)
And here's an article I found about gendering "habibi": https://www.thenationalnews.com/lifestyle/comment/2022/06/06/what-does-habibi-mean-when-and-when-not-to-use-the-term-in-the-uae/#:~:text=Habibi%20(male)%20and%20habibti%20(,t%20drop%20it%20too%20casually.
Yeah, Nour El Ain was super-popular when it came out and even trickled in marginally onto Western radio marginally. If you've ever heard an Arabic pop song, it's probably that one.
That's really interesting!
Is there a different number of syllables in the masculine or feminine second person? Are there traditional rhyme or alliteration schemes that might be changed by shifting the gender?
It can affect the number of syllables depending on the word and the dialect/register. It also can affect the rhyme; for instance, in Egyptian and Levantine Arabic, it changes the quality of the vowel preceding the final -k that marks the second-person possessive (i.e. "your X") and in most varieties of Arabic it changes the final vowel of second-person past-tense verbs.
To be honest whenever I see articles about copyright centered on sampling (and there are a lot, for obvious reasons), there’s always a certain Onion-headline quality to it: “Art form based on wholesale appropriation inexplicably finds itself having difficulties with copyright.”
Less snarkily, because I think at a minimum the arguments about transaction costs are good ones, I wonder if the author has considered the effect this would have on, e.g., sample-based instrumentation licenses or VLSI’s, which AFAIK have a robust private market that would presumably be somewhat harmed by compulsory licensing.
Also 30 seconds seems like a long time - that’s going to be 1/5 to 1/6 of a typical song length, easily enough to get the entire chorus or hook in there. I’d suggest at most 10.
“Art form based on wholesale appropriation inexplicably finds itself having difficulties with copyright.”
It’s a funny quip, but this applies to a significant portion of rock music in the 70s. I’m a huge Zeppelin guy but a decent portion of the catalogue is just an appropriation of Howlin Wolf and others.
I don’t see any ethical or moral daylight between that and sampling.
I've been down so many Zeppelin youtube black holes. Love them. My only hot take with the "appropriation" bit is ... ok, but just listen to Page and Bonham. I don't think it's hyperbolic to say Bonham's grooves were unlike anything that had ever been played before.
Great point on the 12-bar blues. Thought this Polyphonic video raises interesting authorship questions on how the blues is built on folk tradition and folk tradition is more collective. There's a 1941 interview with Muddy Waters (6 minute mark) where he says he learned his song from Son House which was similar to Robert Johnston's Walkin' Blue but the song was out way before Robert recorded it.
https://www.youtube.com/watch?v=KGmMjAGYyYA
+1 re the Onion headline issue, as I have a similar reaction. I'm not going to claim sampling is an inferior artform, but, as lawyer, all I can say is that I would tell a client they should get authorization from the necessary rights holders before sampling a work unless they have a really clear fair use defense (which will rarely be true for commercially released music) or otherwise have a solid argument for why copyright shouldn't even apply, and you probably should still get authorization because it will almost always be cheaper than defending a lawsuit.
Man, paper planes still absolutely slaps. Also great post! I worried that at the start it was going to be about people complaining about 'appropriation' or something but it was actually really interesting.
Lol I was actually lamenting while reading this that copyright law hadn’t spared us from “Paper Planes.” De gustibus non disputandem est.
I think the cultural discussion around “appropriation” and the post-Napster copyright discussion could learn a lot from each other, and complicate the simplistic views that many people take of one or the other.
This argument doesn’t work for me. Treating a composition and a small portion of a recording of that composition as equivalent is not sensible. You even kind of acknowledge that by using a trivial example (the hi hat part), rather than a relevant one, like say the famous guitar lick. In the current scheme, those different sounds would be wildly different costs to license, for obvious reasons, but your proposal treats them as equivalent. That makes this a non-starter.
I don't think the piece argues that the costs should be the same, but that samplers and covering artists should be have the same rights to pay a fee for a non-negotiable license. That's completely compatible with samples costing more
I see the value in a set formula for non-negotiable licenses, but I think disinterested makes a good point with the famous guitar lick example. If we use a set time based formula it will be used to relatively cheaply license short but iconic parts of songs. I don't think it makes the whole idea a "non-starter" but it is worth considering.
To argue with myself, you could use the same argument to do away with compulsory licensing for songs as well, since the perceived value of a cover of "stairway to heaven" is much greater than a cover of some lesser known Led Zeppelin song. I don't really have a good reason to say why this is OK but compulsory licenses for samples is not.
Why does that make it a non-starter? What's so bad about a world where that happens? I don't think it's going to substantially reduce the incentive to create new music and if it means we get to use great bits of music more broadly that's great.
The piece advocates the rate be set based on sample length alone.
That does not in any way conflict with what I said. It could be the case that even a one-second sample could be priced higher than the fee for a cover
I’m not talking about covers in my comment. I’m saying different samples from the same song can be wildly different in terms of their usage and recognizability. If you sampled the hi-hat, as the article says, that is a trivial sample that no one would recognize. If you sampled the guitar, people would hear it as “the guitar from fast car” and arguably you should owe more royalties because it’s a key component of the song.
My point is that a scheme that treats those samples as equivalent is not one recording artists would support.
This is a great piece, as someone in the indie electronic music industry who also DJs. I never thought I'd see a Francis Grasso shout out in Slow Boring, what a pleasant surprise!
The idea that authors should retain some kind of control over how their works are used (eg moral rights) is one of the worst ideas of all time and I don't understand why many otherwise left leaning individuals support it. Many European countries are happy to apply a high tax rate to earnings but yet allow an artist to deny other people the benefit of reusing their creation.
All of human society is built on using other people's ideas and innovations. Why should this be different because you filled a copyright. People who come up with policy ideas don't get to decide who can use them (eg if you suggested the Obama/RomneyCare idea you can't say one party can't use it). Scientists don't get to deny other people the use of their laws.
The world is made better off when we can all build on the ideas that go before us. And it's not like it follows from any philosophical notion of private property (eg on a Lockean perspective asserting copyright is actually violating other people's property rights to make what they want).
Yes, copyright is important to encourage innovation but that's the only justification.
I largely agree, but the very literal protection from someone copying your work word for word and selling it shouldn't be allowed for some amount of time... Maybe 25-50 years tops.
Agreed. I think ideally intellectual property protections are very strong for some limited amount of time. For a long time it was 28 years with an option to extend another 28, which feels like an ample amount of time.
I agree that it's desierable that some kind of compensation be given for that use to incentivize creation but I don't think that (in an ideal world) that should give the author the ability to, say, refuse to sell their work in a particular area or format even within that time period.
For instance, even if you believe that your work is designed to be read if you refuse to allow an audiobook or if you try and refuse to let your story be included in an anthology I think it's totally appropriate that there be a compulsory licensing scheme where the government says, ok you get this much in royalties per copy of the work but they get to create an audiobook/anthology even if you don't approve.
To be clear, this is what is ideal, I realize in actuality there will be all sorts of concerns and complications which might make this a bad idea. But it shouldn't be ruled out in principle.
I'm all for this idea
Just want to chime in to say that, to the extent I believe I grasp his argument, I support Chris Dalla Riva's position on this issue.
Very interesting and unexpected piece. I certainly didn't think I was going to learn so much about sampling when I woke up this morning. I'm especially grateful for learning where the beginning of "Paper Planes" -- a piece I love -- came from.
And it got me to thinking. How did earlier periods handle this kind of thing? And I thought specifically about that masterpiece, Brahms' "Variations on a Theme by Haydn" and how he got the rights from the long-dead Haydn, only to learn this from Wikipedia:
"Brahms composed the work on a theme entitled "Chorale St. Antoni", which Brahms found in a wind ensemble composition. When Brahms discovered it, the wind ensemble piece carried an attribution to the composer Joseph Haydn. Brahms titled his own composition accordingly, crediting Haydn for the theme. However, music publishers in the early nineteenth century often attached the names of famous composers to works by unknown or lesser-known composers, to make the pieces more saleable. Subsequent research has concluded that the wind piece Brahms used as a source does not fit Haydn's style. The wind ensemble piece remains without clear attribution."
https://en.wikipedia.org/wiki/Variations_on_a_Theme_by_Haydn
My guess is that rules in Europe might be different than in America, plus copyright didn't last as long in the nineteenth century.
Loved the read, thanks! The bit about samples containing samples is craaaazy. In a couple decades this scheme could have literal hundreds or thousands of people getting paid out for a huge web of samples of samples of samples. Not intractable but logistically challenging.
It's worth saying that, while Sinéad O'Connor could make the recording of Nothing Compares 2 U using a compulsory licence, she did need a synchronization licence from Prince for that iconic music video.
That is wild, that the video requires extra licenses not needed for the recording, even though the licenses are only about things that are already used in the recording!
The video is legally considered to a be a short film that uses the music. In order to use music in a film, you have to license the film (same for TV and ads), and those require voluntary licences, one for the phonorecord (the sound recording - that's what the P in a circle symbol stands for on your record labels) and one for the composition (lyrics and music, which may have separate copyright holders).
A lot of TV made before the rise of the boxset used only licensed the music for broadcast (ie not for reproduction onto media) or only licensed it for a period of time, which has meant that a lot of TV made before 1995 or so was unable to be released on home media or streaming, or was released with the music changed (sometimes with a cover of the original song if the composition rights could be cleared but not the phonorecord rights, sometimes with an entirely different song). Film had home media as a major part of the revenue from far earlier on (VHS rentals, etc) and also has always had far larger budgets, much more likely to pay for music to be composed, and much more able to afford to buy rights to music it used.
Music videos are exactly in the same situation, but, obviously, the phonorecord synchronization is never an issue, and they do usually have the ability to get rights from the composer if they have the budget for a music video - it's record companies negotiating with each other at that point, usually.
Note that amateurs uploading covers to YouTube do technically require a synchronization license from the composer (but YouTube has negotiated blanket terms with the performance rights organisations so this is not normally an issue in practices). Putting the cover anywhere other than YouTube is legally more dubious, but in practice doesn't get pursued.