We want our laws to be clear so that we can know what is commanded of us. In the case of copyright, we want to know who owns what, for how long, how far the copyright owner's rights reach, which unauthorized uses are excused and which aren't. Given the hefty civil and criminal penalties for infringement, clarity is not only desirable but seems necessary to comport with Due Process. At the same time, we want our laws to be fair, and fairness frequently means being flexible. Flexibility is, though, an enemy of clarity since we do not know in advance if a court will hold that the literal language of the law will bend to the particular facts at hand.
Yesterday's discussion of striking similarity raised this tension between clarity and fairness from a somewhat different angle, but I think the general contours are equivalent. Let's say the "law" (in this case judge-made), working at an acceptable degree of clarity says that if a plaintiff can't prove access plus copying, plaintiff loses even if the two works are substantially similar. That approach provides certainty in application: if not before suit is filed (hopefully), at least after discovery, one will likely know whether a case can go to the trier of fact. Litigation risks can be assessed and perhaps a reasonable settlement effectuated.
But to some, the existence of striking similarity calls, by itself, for a different approach. That approach would suspend the usual rules in favor of incredulity over defendant's denial of copying given the striking similarities. This incredulity may be the result of what Smoky called yesterday the lie factor, or what Judge Posner would call common sense: common sense tells us, at least inferentially, that such striking similarities must be the result of copying even though there is no evidence of it and notwithstanding plaintiff's burden of persuasion on that issue.
If infringement cases where striking similarity is alleged are permitted to go to a jury (based, lest we forget, solely on the existence of expert testimony) we are in treacherous territory as the BeeGees case (Selle v. Gibb) demonstrated. In Selle, I believe it was Barry BeeGee who mistook his own composition for plaintiff's in a deposition, evidence certainly of similarity, but not of copying. And in Gaste v. Kaiserman, there is no doubt in my ear that the introduction to the two songs were strikingly similar. But it is also clear to me that Gaste never should have gone to a jury due to a lack of copying and that that verdict to the contrary was, therefore, a travesty of justice.
In Selle, the district judge did the right thing by granting a jnov; that did not happen in Gaste. It may be some will think both decisions are "correct" or that such disagreements as to the result are inevitable. But it is also clear that absent the availability of striking similarity neither case would have gone to the jury. Those whose favor clarity (and respectfully, I think that group is larger than insurance companies) think that is a good idea. Those who want to maintain flexibility (or as Judge Posner would say, common sense) favor maintaining it. Both approaches have their costs.
Starting Monday I will be on vacation through July 4th, returning to the blog on Tuesday, July 5th. When Grokster comes out, hopefully Monday June 27, I will be guest blogging (with others) on Tom Goldstein's Scotusblog site. Happy 4th and thanks for reading.
After the smoke cleared from my long-winded tale, I meant to suggest only that where there is striking similarity, and in the context of summary judgment, the defendant should be required to establish authorship in some fashion greater than a reliance on the prima facie presumptions from his or her registration. Access is a substantive requirement mostly because of the independent creation rule and that premise is suspect except as it regards common elements - - elements which are likely not to produce *striking* similarities - - just similarities. It is plausible to think two authors of approximating works would draw from common elements without access or copying. It is implausible (or certainly less plausible) to presume that they could produce strikingly similar works if the analysis of *striking* requires a filtering out of unprotectable elements. What is unfair or imbalanced in shifting the burden of proof on authorship in these cases? What barrier or difficulties am I missing? You can keep the burden on access and copying on plaintiff. The defendant should not have to prove a negative, just the circumstances of his or her own creation.
ReplyDeleteAnother famous unconscious-infringement case, which the Professor mentioned only in passing, was Fisher v. Dillingham:
ReplyDeletehttp://ccnmtl.columbia.edu/projects/law/library/cases/case_fisherdill.html
which I think was wrongly decided.
The Fisher v. Dillingham case illustrates a problem with the presumption that access+substantial similarity implies copying. Some works are "contagious" in the sense that they are learned and transmitted easily and can circulate independently of any indication of their origins. This happened, famously, to the Christmas hymn "Silent Night" which was for a time attributed to Michael Haydn. It took a government investigation to track down its true origins. Loss-of-origin also occured in the case of the song "Home on the Range". The result was litigation in the 1930s: The plaintiffs were a couple who had been perhaps the first to capture the song from oral tradition. Southern Music Publishing Company v. Bibo-Lang 26 USPQ 321 (S.D.N.Y, 1935), 26 USPQ 324 (S.D.N.Y., 1935). When "access" is almost impossible to disprove, and "substantial similarity" can be interpreted quite broadly, and some works can circulate through undocumented channels, the presumption of copying can only be disproved by producing a daily log of all one's movements and of everything one has heard or read. Maybe it would be fairer to abandon the fiction that independent creation is an infallible defense to a charge of infringement, and to say that in some genres, copyrights will sometimes operate like patents, shutting out both copiers and independent followers alike.
I never tire of stating it: Whatever is wrong with copyright is only made worse by an absurdly long copyright duration. A shorter term of copyright is not only desirable for its own sake, it would also make some of the system's other shortcomings more bearable. If we had a moderate term of copyright second-comer who found that he had independently created a "substantially similar" work and that he had no way to disprove access or copying, would be more likely than under current conditions to have a realistic option of waiting until the first copyright expired.