AI: Inspiration, Sampling, or Rip-Off?
In the 1970s, George Harrison’s songÂ My Sweet Lord wasÂ found to be a rip-off ofÂ a Chiffon’sÂ tune called He’s So Fine. In that case (Bright Tunes Music v. Harrisongs Music),Â the courtÂ decided thatÂ Harrison “unintentionally copied” the earlier song,Â andÂ ordered him toÂ relinquish royalties.
This type of supposed plagiarism has occurred many times in popular music over the years. The latest case involves the British band Coldplay and rock guitarÂ mainstay, Joe Satriani. Some say Coldplay’s most successful single to date,Â Viva La Vida, is a rip-off of Satriani’s earlier work,Â If I Could Fly.
The songs do seem to have the same melody and the same chord progression, but what do your critical thinking skills tell you in cases like this?
Is this plagiarism? Are the songs totally different? If they’re the same, is it unintentional? Is it a tribute? Is it coincidence? How do these types of similarities affect your consumption of music?
The Afternoon Inquisition (or AI) is a question posed to you, the Skepchick community. Look for it to appear daily at 3pm ET.
In my opinion, Nickleback should have to pay plagiarism fines, because crap had already been invented.
@Some Canadian Skeptic:
Ouch! An unsolicited attack.
I like that!
@Some Canadian Skeptic: Hilarious. COTW
I’m pretty sure Led Zeppelin ripped off all the tunes I was thinking about while I was in the womb so they owe me.
@Some Canadian Skeptic: And they like to steal their own crap. Over and over.
The artist formerly know as Cat Stevens claims Satriani’s melody sounds like one of his songs…
@Some Canadian Skeptic: Yep, this is how you remind me. You repeat the same crap over and over, and over…
With the simplcity of modern pop music, the repitious chords and the fact that most bands are formed as a way to get laid it is more surprising when something is original.
I think that music is a personal creation and if you want to sample or copy or imitate it too closely (you hear me, Vanilla Ice?!), you should pay. Tributes are usually pretty obvious, and claimed as such, with the necessary permissions sought and granted.
I also think that if a song is a “classic” and has been around forever (maybe 100 years is long enough?) then maybe it’s fair game.
Also, Lars Ulrich is still a douche.
I honestly can’t tell that there is a similarity between the two songs at all. There’s one point where maybe I heard something.. but it was still a stretch.
This guy does a break down of each song:
Outright plagiarism is not at all new to music. If you look at music from past centuries, many composers blatantly copied, added to, or recontextualized from earlier composers or even themselves. Medieval composers would take an extant melody (often liturgical chant) and use it as the structural bass line of a new work with added voices on top. Baroque composers often stole from one another, especially church musicians such as J.S. Bach, who were expected to continuously write ‘new’ pieces for the entire liturgical calendar. Same with Haydn, who had to come up with new stuff at the whim of the Esterhazy family.
If you reduce music to its basic elements you have 12 notes (or six intervals if you take inversions and assume equal temperament). Once that gets filtered down into the major/minor tonal system, you have 5-7 melodic notes and a handful of chords. The combinations at the local level are pretty limited. However, things like dynamics, timbre, harmonic rhythm, and formal structure offer more possibilities. Think of all the 12-bar blues songs that are virtually the same on paper, but are sonically quite distinguishable. Lyrics are also important, especially to pop music. A lot of pop music is more of a poetic art form where the music functions as a vehicle for delivery of the lyrics (not to disparage pop songwriting – its an art unto itself, just different than writing non-texted music).
I’m not surprised when I hear songs that sound similar to others. Given that the expectations are fairly well defined for a pop song and the thousands (maybe more, I don’t know the actual stats) written each year, it would be hard not to find coincidental likenesses every now and then. Personally, my listening tastes are quite adventurous, so I seek out a lot of the avant-garde composers – Stockhausen, Varese, Ligeti, etc, as well as a lot of experimental and electronic music.
FWIW, dinosaur comics has already touched on this one:
I agree VoxMachina how many different ways can you arrange those 12 notes in a unique way. I am surprised that it doesn’t happen more often.
I remember a few years ago, some one-hit wonder made a song-butterfly, and it “sampled” a RHCP chord. You could hear it in the background, just repeating. I think stuff like that is very lame.
Case in point: http://www.youtube.com/watch?v=JdxkVQy7QLM
Did they intentionally copy it? Hard to tell. It’s likely one of them heard Satriani’s song at some point and used the melody, either consciously or unconsciously.
A more blatant example would be “I Want a New Drug” versus “Ghostbusters”.
“Surfin’ USA” = “Sweet Little Sixteen”
“Sidewalk Surfin'” = “Catch A Wave”
“I Want Candy” = “Bo Diddley”
Well I know that Elton John totally copied REO Speedwagon.
Also I think there’s a lot of unintentional “inspiration” and coincidence, but for a certain degree of similarity it should count the same as outright plagiarism. I’m not sure how this can be objectively judged, but it shouldn’t have anything to do with perceived intent. That’s impossible to judge anyway.
Is it just me who wonders why you’d go with the unconsciously remembered version as-is?
@Bjornar: Oooo explain that one.
@Bjornar: Also, can’t seem to sign your guestbook for some reason.
<– not a bot, promise.
That the chord progressions and tempo are similar is not very incriminating, to me. Most songs probably have the same tempo and chords as at least one other song (particularly with blues, punk, country, and any descendants thereof, since they’re intentionally minimalistic).
The similar melodies are much more interesting. What if, given that chord progression, one tends to pick out a melody like the ones CP and JS did, for whatever reason? It may just be that those note movements are among the most compelling possibilities within the constraints of the chords.
It would be interesting (but probably impossible) to gather 100 people who’ve never heard Coldplay or Satriani, play them those chords over and over, have them sing whatever melody pops into mind, and see what they come up with.
Maybe you don’t realize it was a memory and not something you just invented on the spot.
I’m not about to listen to any Coldpay today, regardless of whether it’s in the name of critical thinking or science or whatever.
I will say this: I saw the unbelievably cheeseball video of the Ray Parker Jr. “Ghostbusters” song on VH-1 last weekend (nice cameo, Al Franken), which Huey Lewis accused of ripping off “I Want a New Drug”.
Then, on Tuesday, my shuffle spit out “I Want a New Duck” by Weird Al, which I hadn’t heard in forever, and remains AWESOME.
I always thought U2’s “Angel of Harlem” sounded a lot like Bob Dylan’s “Like a Rolling Stone”, but it’s been 10 years since I’ve listened to either.
@nollidge: I think you nailed it. It is a matter of musical grammar and syntax. The chord progression itself isn’t particularly unusual. Melodies tend to focus on pitches consonant to the underlying harmony, which limits the number of notes available as cadence and structural points. There is also a limited number of melodic patterns that smoothly connect the structural pitches. Also, both melodies are periodic, following an antecedent/consequent pattern, which further limits the grammatically ‘correct’ note options. Consider all the jazz standards and Tin Pan Alley tunes that use a ii-V-I progression, or early rock or blues songs that use IV-V-I. If you strip away the ornamental pitches, the remaining structural melodies tend to show a lot of similarities.
This makes it terribly difficult to legally prove plagiarism. Copyrighting a harmonic progression or melodic pattern would be the equivalent of copyrighting subject-verb-object. It is entirely possible that Coldplay did intentionally base their song on Satriani’s, but a musicological analysis won’t prove it either way, especially in a legal context. Besides, music is a cultural phenomenon, so new contributions are often based on existing traditions, whether consciously or not.
Eh. How many ways can you arrange 26 letters into unique patterns? And letters don’t get used simultaneously or for different durations.
At some point, the odds of coincidence become vanishingly small. I don’t know how many chords in that is, but I’d guess fairly quick. Then you get into the realm of “stolen” versus “inspired by,” which is tougher. Always be inspired by public domain stuff, people.
BTW, the Rolling Stones need to thank Beethoven’s “Ode to Joy” for making “Paint It Black” possible.
I think we should have the eminent mathematician Charlie Epps (from CBS’s “Numb3rs” show) run a Cambrian-Ordovician regression matrix analysis on the two tunes and that should settle that! ;)
I thought about hearing both and comparing the two, but that would require voluntarily listening to a coldplay song.
@VoxMachina: The idea of musicologists testifying as expert witnesses fills me with an odd glee.
@KristinMH: Are you at all familiar with Gerald Eskelin? He’s not strictly a musicologist, but he talks about testifying in similar cases in his book: http://www.amazon.com/Lies-My-Music-Teacher-Told/dp/1886209251/ref=ntt_at_ep_dpi_1
Pretty good read, IMHO.
@Steve: Except that Brian Wilson credited Chuck Berry as the main author of “Surfin’ USA”.
The most painful one for me is “Why Don’t You Get A Job?” by The Offspring, which blatantly rips off “Ob-la-di Ob-la-da” by The Beatles without credit.
Which is, I think, where the distinction comes into it: is the composer willing to acknowledge the source? Brian Wilson certainly did, and it is often referenced in the titles of works by baroque/neoclassical composers where the source was known; Coldplay to date haven’t AFAIK, although it may have been unintentionally ripped-off.
From my experience in bands, if it’s too close to an existing work, one of the band members will usually pick up on that. What you do about it after that is a different matter, though.
Sometimes when it quacks like a duck it is a playtpus.
I am sure someone else had said that first but never heard it before so is that stealing?
@phlebas: But how many ways can you arrange 26 letters into meaningful structures? Many fewer, and that is with a language where the individual parts (words) have inherent meaning. Musical meaning is dependent on syntax, grammar, and structure, the rules of which pare down the possibilities for harmonic and melodic motion. I find similarities between pop songs all the time – they all work on a common musical language. Blatant examples like the Coldplay and Satriani are a little less common, but if you take a broader perspective, all pop music shares similarities to some degree. Most pop music has phrase structures based on either 4 or 8 bar phrases, a harmonic rhythm of either 2 or 4 beats (sometimes longer, as in the blues), harmonic progressions following a predominant-dominant-tonic pattern (or variant thereof), and melodies mostly consonant with the harmony. Combine that with 3-4 diatonic chords and 5-7 diatonic notes available and variety in pop music seems more unlikely than similarity. All pop music is a combination of very simple (yet meaningful) musical patterns, so I’m not surprised when two songs happen to coincide. Just because two songs sound similar doesn’t necessarily indicate plagiarism. It could be the result of independent, yet parallel songcraft.
For a point of reference, it might be helpful to listen to some integral serialist composers (Pierre Boulez and Milton Babbitt are a couple good examples). Their music is based on a completely different (often stricter) grammar and syntax than most, yet I’m willing to bet that most new listeners will say they sound similar, even indistinguishable. All resulting from similarities in how the music is constructed.
Now, where to draw the line legally is a different question. Since music is a commodity (at least in the eyes of copyright law), it may make sense to draw an arbitrary yet well defined and agreed upon line. Perhaps this demonstrates how our legal concept of copyright isn’t based on the actual workings of our culture.
BTW Sam, thanks for the great AI. I thinks its great to flex the skepti-muscles on the arts as well as science and philosophy.
Jesus, Sam! Just one of those is a profoundly difficult question to answer. But all of them? Together?!? Yeech.
I am terribly poor at math, and so, much to my frustration I am unable to “prove” my argument. But I think the arguments about small numbers of note seqeuences and chords (and hence the possibility for distinct variations and distinctly differnt songs) are just way, way, way wrong.
Let’s look at it this way. A piano has 88 keys. A proper chord, standing alone, should contain three or more distinct notes. A second chord, played against the first chord need only contain two distinct notes, as it combines with the first chord to jointly form a third chord, or a harmonic dischord, or a melodic shift, of what-have-you.
So, add a couple of different melody lines, composed out of any combination of the 88 keys, so that’s potentially 88 times 88 time 2, I think, added to the two chords which are made up of anywhere from 3 to 7 or perhaps 8 distinct notes, times 88 times 88 times 2 times 2 (two chords for four hands — assuming two players) and assuming an arbitrary duration of 3 seconds before the next chord sequence appears….
Anyway, as I say I am no mathematician, but I think the reality, just like the letters of the alphabet, very easily and quickly surpasses millions. By the way, as far as I know the current full size Oxford has over one million distinct words.
And in all seriousness, I’d be very pleased if someone could do the proper math on this and clarify my mess of it.
Sure, but how many notes in a row does it take to make a melody, and hence, a copy? Or, additionally, how many short groups of similiar melodic structures goes beyond coincidence and becomes copy? And how closely must those groups be arranged? And how short or long must they be before they are considered a meaningful duplication? And how many of them?
Any experienced musician can hear far more simliarities in songs one from the next to the next, and so on, than your average lay person. But at what point does it go from being an example of forced limitation of the mathematics of music to being a copy?
Reference comment 17
Surfin’USA did not plagierize Sweet Little 16. If one looks at the published sheet music of Surfin’USA, it is written by Brian Wilson & Chuck Berry and published by the Chuck Berry’s Publisher at the time, Arc Music. Chuck Berry wrote Sweet Little 16
Yet another example of the wasted time, energy, and money generated by the nonsense of “owning” ideas. ALL art is derivative–if it weren’t we wouldn’t understand or enjoy it. Artists need to stop wasting their time worrying about keeping irrational control over their work and get on with creating business models that work with the reality of the nature of ideas–that they WILL be copied, changed, and take on a life of their own that their creators could not possibly control, or even imagine–and that this is a GOOD thing.
@SicPreFix: Go back to musical grammar, Helmholtz, and Rousseau. Those 88 keys repeat at the octave, so there really are only 12 notes, octave equivalency being integral to music theory since Rousseau. A chord with E, G, and B still sounds like an E minor chord no matter what octave the notes occur in. The major/minor tonal system has only 7 notes within a key, yielding three harmonic (chord) functions: predominant, dominant, and tonic (with some overlap between chords separated by thirds, which explains diatonic chord substitution – the difference between ii and IV is only one note). Harmonies beyond the triad (three notes) are heard as extensions of the base harmony, essentially negating large stacks of harmony (unless we venture into non-functional or non-tonal land, which pop music simply doesn’t do). Also, due to tonal function, some notes are simply more important, others being decorative turns or color tones. So rather than mathematically looking at how many combinations are possible, look at how many combinations fulfill the rules of the given grammar (and musical context – bluegrass music has different expectations than punk, even if they share some grammar).
Yes, there are over a million words in the English language, but music is an abstract language. For music to be similar to a verbal language, A-flat or C-sharp would have to mean something. Unless you’re synesthetic, they don’t mean anything. Even with chords and scales, the grade school music appreciation axiom of major=happy/minor=sad is woefully inadequate at actually describing the aesthetic result. There are plenty of uplifting pieces in minor and depressing pieces in major. Music only gets its meaning from the way it unfolds over time, which in the case of pop music follows a particular grammar. The exception to this is when the music is texted, but in this case the meaning is derived from the words and (usually) supported by the music.
There are also limits to human perception and memory. Even trained musicians have trouble distinguishing pitches at the extreme ends of the sonic spectrum. Most people can’t accurately recall melodies beyond about ten pitches, maybe a few more if it contains a distinctive melodic leap or rhythm, less if it is full of leaps, very long, or rhythmically consistent. Everyone remembers the theme from Beethoven’s fifth, but how many can recall the development note for note?
Some of the literature in psychoacoustics or musical perception would be a good source if anyone wants to take a look. Both are relatively new fields, but have made some interesting insights in recent years.
@Kaylia_Marie: It’s an old JREF forum in-joke based on this foolish post: Elton John totally copied REO Speedwagon (Oh my, that thread is over five years old.)
And my guestbook is apparently broken, and has probably been so for months without anyone noticing. Thanks for telling me, and for trying to write in it. I’ll see if I can’t fix it shortly. :)
I’ve stopped worrying about what is considered “good,” and just like what sounds good to me, and not care when others like something I don’t.
Besides, all pop music come down to four chords anyway.
I suppose the question the court also has to consider is “intent,” which is hard to prove. Is “similar” to another song the same as “plagiarism?” If so, where is the line and who gets to draw it?
VoxMachina and nollidge have a good point about the limits of music composition, too. In theory, I suppose there is close to an infinite number of ways to combine notes. But how many of those combinations are actually music? I’ve heard a few pop songs in my time that sounded very, very similar to some classical melodies, but were not labeled as such. Ripoff or coincidence? I don’t know.
@Steve: I’m not sure, but I think “Sidewalk Surfin” was a deliberate satire of “Catch a Wave.”
Yes, of course. I know and understand that. And as far as musical theory and the technical factuality of the twelve note scale, of course you’re right., But what about perception and how we hear? By which I mean, to the ear two notes one or two octaves apart seem to most listeners to be two different notes, not the same note at two different ranges.
Or is that where we begin to approach the topic of psycho-acoustics?
I guess what I’m trying to get at is the subjective perception of music versus the musical theory mechanics of it. A sort of perhaps airfy-fairy subjectivism versus absolutes. Or something like that.
I’m not diagreeing with you, I’m approching the topic from a different direction.
diagreeing = disagreeing
@SicPreFix: Yes, that is where we get into psychoacoustics. How we hear/perceive music is often very different from how it is constructed. There are some notable differences between the Satriani and the Coldplay, yet we still hear them as similar.
Notice in the analysis videos Sam linked to one of the chords in the Satriani is a suspended chord. Its a difference of one note, but we still hear the same function. Same with the substituted chords – there is a one note difference between them, yet we perceive them as fulfilling the same role. Also, there are some inverted patterns in the melody, but we hear them as similar because they cadence on the same structural pitch. The songs are also in different keys, which means the chords and notes are necessarily voiced differently. If those differences were so crucial, then we wouldn’t hear the JS and CP as strangely similar and potentially litigious. That is what I mean when I say that the musical grammar only allows limited options. There are notable differences between the two songs, but thanks to our perception, we hear the important stuff (the similarities) and leave out the (differentiating) details.
Tut, tut. Speak for yourself.
That`s an interesting point.
Ah, I see. Okay, now I get it. I had clearly misunderstood your earlier comments. But now they are making sense to me. Thanks. Really good analysis.
As someone who writes music for a living, I’d like to expand on some of the great analysis here, especially by VoxMachina.
We musicians listen to a LOT of music. Plus, we’re always listening critically and committing the music to memory to varying degrees. Everything we write is therefor going to be influenced by something do to our internal catalog of music (for lack of a better term). When a melody or idea pops into my head, I may initially think that it’s an original creation, but there’s no guarantee that I’m not just remembering someone else’s tune.
One of my professors in college said that “it’s impossible to compose a completely original tonal melody. Every combination of notes and rhythms have been done.” Which is why some composers have turned to atonalism, or even inventing their own tuning systems. When someone decides to write in a particular style (pop, jazz, romantic, etc), you get locked into a particular set of rules and any deviation runs the risk of breaking the style (although that may give interesting results). The choices become even more limited when you want to convey a specific emotion. If I wanted to write a “happy” piece for instance, it’s pretty much a given that I’m not going to use any diminished chords.
I’m usually willing to give songwriter’s the benefit of the doubt. Copyright infringement is such a tough thing to prove because if I’m not mistaken, the plaintiff does have to prove intent, and there’s just way too much material out there to unintentionally borrow from.
You must log in to post a comment.