Faire Use

What with all the plagiarism this month, it seemed like a good time to revisit one of the most plagiarized composers of all time—at least in his own mind, that is. The perennially litigous Ira Arnstein, born in 1879, would only be remembered today, if at all, for a few not-unpleasant Tin Pan Alley and religious numbers from the 1920s, except for the fact that he was convinced that other, more successful songwriters were constantly ripping him off. Arnstein brought no less than five lawsuits between 1936 and 1946, against the likes of Harry Warren, Joe Burke, Cole Porter, and others; and in the process, he indirectly paved the way for a refined legal definition of music plagiarism, one that, for better or for worse, persists today.

Arnstein was born in Kiev, and emigrated to the United States when he was eleven. A sometime pianist, music teacher, and composer, by the 1930s, he had become something of a full-time plaintiff. His first effort, Arnstein v. Edward B. Marks Music Corp. (82 F. 2d 275 [2d Cir. 1936]), set the pattern: Arnstein would take a song of his own and another, more popular song, and, by highlighting selected pitches, altering rhythms, changing octaves, and elevating accompanying notes to melodic status, conjure similarities between them. Jack Lawrence had co-written one of the songs named in Arnstein's suit, "Play, Fiddle, Play." Lawrence recalls:

Arnstein's lawyer had a piano and fiddle player in court plus huge music charts, an intriguing presentation. The melody line of a song consists of single notes in the clef treble. Arnstein's chart highlighted notes in both the clef and bass and when the fiddler played only the highlighted notes... lo and behold! — it sounded exactly like our song! Our attorneys spent hours trying to explain this to the judge, but he would only accept what he was hearing.

In fact, as you can compare for yourself via the Columbia Law Library's Music Plagiarism Project, the songs barely rate even a charitable resemblance. Arnstein didn't help his cause by admitting that he had threatened the defendants. The New York Herald Tribune reported Arnstein's testimony: "'I was desperate,' Arnstein said quaveringly. 'I heard my song being played everywhere, and I was starving. I was out of my mind and might have committed murder.'" (Lawrence remembered that Arnstein "paraded in front of the ASCAP offices wearing a sandwich sign that read: 'My songs have been plagiarized by the following writers: Irving Berlin, George Gershwin, Cole Porter, Jerome Kern, Rodgers and Hart.'") Theatrics aside, Arnstein's suit was eventually dismissed by the Second Circuit. (Judge Learned Hand took the opportunity to excoriate popular music in general, suggesting that pop songwriters were too witless to plagiarize: "[The defendant's] gifts were very limited, and to attribute to him the ingenuity and penetration so to truncate and modify, and thus really to create a melody out of other elements, is harder than to suppose that the extremely simple theme should have occurred to him out of his own mind.")

Undeterred, Arnstein next brought suit against ASCAP itself. Arnstein v. ASCAP (29 F. Supp. 388 [S.D.N.Y. 1939]) was much in the vein of his previous action, right down to Arnstein's familiar allegations of an enormous conspiracy to pirate his compositions. From his complaint:

That they have in cooperation with the other defendants and their attorneys conceived the plan of branding the complainant as a lunatic and have worked in harmony with the officers of ASCAP and MPPA (Music Publishers Protective Association) to oust the complainant from the W.P.A. and have caused him to starve.
...
That his room was constantly ransacked and many manuscripts and letters stolen. When he complained to the Police Department no action was taken but two gorillas beat plaintiff up and plaintiff produced a Doctor's certificate at the trial to prove that he received medical attention for several weeks....
...
That the conspiracy extended even to the Court Room during the trial. Witnesses and Musicians were accosted by defendants attorneys and induced to disappear. That twenty-five (25) musicians from the Union who signed affidavits to the similarity of the Music, were given jobs in the Russian Ballet as an inducement for not testifying at this trial.

Arnstein managed to come up with two expert witnesses willing to testify that his song "Whisper to Me" bore certain resemblances to Joe Burke's "My Wishing Song," although the force of said testimony was blunted when the experts admitted under cross-examination that both songs bore certain resemblances to a previously existing number called "Are You Lonesome Tonight?" Arnstein similarly tried to sue BMI a few years later (Arnstein v. Broadcast Music, 137 F.2d 410 [2d Cir. 1943]), brazenly citing "Whisper to Me" again, this time comparing it to a tune called "It All Comes Back to Me Now." His next try, Arnstein v. Twentieth Century Fox Film (52 F. Supp. 114 [S.D.N.Y. 1943]), was his most far-fetched yet, asserting that Harry Warren's "I've Got a Gal in Kalamazoo" had infringed Arnstein's Wagnerian parody chorus "Kalamazoo" pretty much on the grounds that they both mention the same city.

Finally, though, Arnstein caught a break. For his final trip through the judicial system, Arnstein took aim at none other than Cole Porter. "Begin the Beguine" had been stolen from Arnstein's setting of "The Lord Is My Shepherd." "Don't Fence Me In" had lifted from Arnstein's "A Modern Messiah." "My Heart Belongs to Daddy" had its illicit origins in an Arnstein instrumental called "A Mother's Prayer." Arnstein had become accustomed to grants of summary judgement for the defendants—the presiding judge saying, basically, that there wasn't enough evidence to even bother with a jury trial—and he had always lost on appeal. But Arnstein v. Porter (154 F.2d 464 [2d Cir. 1946]) was, surprisingly, remanded for trial by the district court, and Arnstein got his chance in front of a jury. Of course, he lost—but it took some doing. Charles Schwartz relates in his biography of Porter:

In the two-week long jury trial that followed, Monty Woolley, Deems Taylor, and Sigmund Spaeth all appeared in Cole's defense to support his contention that he had never taken material from Arnstein. Cole also testified that he neither knew Arnstein nor was familiar with his work. When the case finally went to the jury it was dismissed as being without merit after a deliberation of nearly two hours. But though Cole won the case, it was perhaps as a result of the experience gained from this trial that, when asked if he ever went out without a carnation in his boutonniere, Cole answered, "Only when I'm being sued, because a carnation in the buttonhole never helps your case before a jury."

What changed? The Second District had decided to raise the bar for summary judgements, making it that much harder to simply dismiss nuisance lawsuits like Arnstein's. Now, the plaintiff was only required to demonstrate two things: a) that the defendant might have indeed copied from the plaintiff, and b) that the copying constituted "improper appropriation." Although subsequent case law has once again lowered the impediments to summary judgements, the Arnstein test remains the basis for music plagiarism cases. And both parts of the test are, well, complicated.

In the absence of a direct confession, to establish the fact of copying, the plaintiff needs to show that the defendant had access to the pirated material, and that the material was actually pirated—that the pieces in question are, in fact, suspiciously similar. (Note that this makes all the evidence for this part of the test circumstantial.) The court recognized the tricky nature of the access question; if that couldn't be proved, the plaintiff could still satisfy the test if the similarities were "so striking as to preclude the possibility that the plaintiff and the defendant independently arrived at the same result." How striking would that be, exactly? Good question. And while the test says that the less evidence of access, the more obvious the similarities need to be, what no court has ever really cleared up is whether the relationship works in reverse—that is, whether more compelling evidence of access allows for less of a resemblance between the original and the alleged copy.

Demonstrating that the copying amounted to "improper appropriation" is an even stickier wicket. The plaintiff needs to show that what was stolen is what made the original piece distinct and memorable to the ear of "the ordinary lay hearer." That might seem reasonable—anyone can hear if two songs are the same, right? Maybe so, but most people aren't aware that a lot of what they're hearing may not actually be copyrightable. In legalese, such elements are called scènes à faire (literally, "scenes that must be made"), meaning they're so common to a given situation that they've lost any claim to distinction. To allege that one romantic comedy stole from another because they both end with weddings would fail on the grounds that such a denouement is a scène à faire for romantic comedies. So just because two songs both start on do, mi, or sol, or both end on a V-I cadence, or both use, say, a twelve-bar blues progression shouldn't be enough to establish plagiarism, but that's assuming the judge or jury are musically literate enough to know the basic building blocks of tonal music.

And here's where the system fails, because expert testimony on the question of improper appropriation is, at best, severely limited. Under Arnstein, a musical expert can only testify as to the hypothetical effect on that "ordinary lay hearer," completely ignoring any opinion as to whether perceived similarities are scènes à faire or not. (As it is, having an expert witness try to put him or herself into the mind of a non-expert may, in fact, violate Federal rules of evidence regarding expert testimony.) Under a later, non-music plagiarism decision, Sid and Marty Krofft v. McDonald's (562 F.2nd 1157 [9th Cir. 1977])—the first case to concern "total concept and feel" plagiarism—expert witnesses are simply excluded from the improper appropriation test. As lawyer (and trained composer) Jeffrey Cadwell points out in his article "Expert Testimony, Scènes à Faire, and Tonal Music: A (Not So) New Test for Infringement," this means that musicians are shut out of the very part of the process where they're most needed: establishing whether or not similarities between two works are proof of theivery or just part of a generic vocabulary.

Cadwell points out that the obvious solution—requiring proof of access and letting musical experts testify as to whether similarities between songs are trivial or not—was actually proposed prior to Arnstein, in a 1932 book on music copyright by Alfred Shafter. (Shafter's book, from what I've seen, contains a number of howlers on musical substance, but he correctly foresaw the difficulty courts would have in dealing with it.) And since mass media and digital distribution have made access more and more easy to assume, others have proposed taking the guesswork out of the access question via compulsory music use licenses: basically cheap, no-permission-needed shout-outs to copyright holders that a particular piece of music is being covered, sampled, or otherwise borrowed (see, for example, this article by J. Michael Keyes). Are any of these going to happen soon? Probably not—the wheels of justice do grind exceedingly fine, and judges historically don't like to admit that there's a subject matter that might be beyond their ken.

So what eventually happened to Arnstein? I don't know; after the Porter trial, he seems to have vanished—there were no further lawsuits, and I can't find any record of him after 1946. Wherever he is, though, he'd probably be pleased to know that he's still causing other composers legal troubles—if only indirectly.