27 December 2013

A Study in Scarlet Letters

This isn't really a link sausage platter today; it's instead about the difference between one and many, as reflected by the interplay between Alan Turing and Sherlock Holmes.

Alan Turing was finally granted a long-overdue royal pardon last week. That article in the Grauniad both underplays and overplays Turing. It's very easy to quibble with some of the more grandiose statements; for example, John von Neumann was more the "father" of the modern computer than Turing was (if anything, Turing was the pimp for von Neumann and whoever the mother was... and there's a reason that I'm invoking that particular image that will become apparent below). But that is beside the point: He was treated inhumanely and substantively incorrectly by a system that was imposing a certain upper-middle-class view of protestant values upon everyone through actions that would have been treated as war crimes a couple of years earlier at Nuremburg. There have been proposals to pardon (or otherwise rehabilitate) Turing percolating for at least thirty years — indeed, one of those came surprisingly close to success in Thatcherite England while I was living over there, although it never really made it into public consciousness. There's a simple reason for this: There is, and was, widespread recognition that what the particular bigots in charge of security did in the early 1950s was both wrong and ineffective (such as completely missing Kim Philby and his cohorts), even if nobody will make the obvious connection between a certain class solidarity and... the "t" word. One of them finally came to fruition, and at least now new books that will be written to introduce children to Turing and his achievements will mention that the Crown tried to correct the government's mistake.

And then, in the same newspaper, another asshole claims that pardoning Turing is "wrong" because it wasn't a blanket pardon. With due respect (which is to say none whatsoever), Mr Fogg, you're a bloodyminded fool. By your reasoning, Brown v. Board of Education (both decisions together) was wrong because it did not completely destroy all aspects of segregation right then — not just the schools, but housing and employment discrimination and ballot access and everything else. The perfect is the enemy of the good, particularly for anything regarding recognition of past inequities, mistakes, blunders, and idiocy. And flipping this around demonstrates why you're so wrong: By the reasoning in your screed, recognizing Rosa Parks as an iconic civil rights figure was "wrong" just because not everyone who ever sat down on a bus against the driver's orders is not simultaneously recognized, regardless of whether they were as (or more) "objectively deserving" of such recognition and iconic status. One does not delay compassion for one in the name of waiting (probably forever) for consensus compassion for all. Then, too, there are probably more than a few truly nasty actors among those Mr Fogg would have pardoned equally with Mr Turing; statistically, there's more than one rapist in there (they were, after all, men). Perhaps a blanket pardon is indeed the best acknowledgement of the UK's mistakes — I'm certainly not against it! — but that does not mean maintaining the official shame heaped on Mr Turing's reputation merely because 74,999 others are not being simultaneously "rehabilitated" is in any way justified. That we can continue to criticize "us" for not doing more is no reason to do nothing. I'm afraid that we'll need Sherlock Holmes to unravel everything... and even then, we'll still be making judgment calls.

Fortunately, we can get Sherlock Holmes to unravel everything (and not just because Sherlock returns in a few days). The Doyle estate has persistently taken the unreasonable position that the clearly-in-the-public-domain aspects of Holmes — published at times that would place them in the public domain under either US or UK law — are kept out of the public domain because other Holmes works are not yet in the public domain (although they will be within a decade). Even Disney wasn't this obnoxious; for all that I disagree with the Mickey Mouse Sonny Bono Copyright Term Extension Act of 1998, it at least has a clear legal basis. Indeed, even the "perpetual copyright" of Peter Pan in the UK is founded on specific legislative acts and not upon heirs who want the public to stay off of grandpa's beach... or at least charge the public a fee to use it, which seems far more likely.

The Doyle Estate's position was just rhetorically castrated in (of all places) Chicago.1 Chief Judge Rubén Castillo held a few days ago that the early Holmes works and their copyrightable elements are clearly in the public domain (PDF), and that the later works do not provide any reverse-derivative-work protection to the earlier ones. Although it was not an unvarnished victory (for example, the specific relief requested was denied because the request was too broad), the following passage illustrates rather well both what is at stake and its resolution:

Conan Doyle['s estate] argues that the effect of such a holding will be to dismantle Sir Arthur Conan Doyle's characters into a public domain version and a copyrighted version. This is, however, precisely what prior courts have done. Silverman [v. CBS, Inc., 870 F.2d 40 (2d Cir. 1989)] and Pannonia Farms[, Inc. v. USA Cable, No. 03civ7841 (S.D.N.Y. 08 Jun 2004), aff'd on other grounds, 426 F.3d 650 (2d Cir. 2005)] instruct that characters and story elements first articulated in public domain works are free for public use, while the further delineation of the characters and story elements in protected works retain their protected status. Silverman, 870 F.2d at 50; Pannonia Farms, 2004 WL 1276842 at *9. Conan Doyle argues that the precedent exemplified in Silverman should pertain only to two-dimensional, "flat" characters and not to complex, three-dimensional characters such as Sherlock Holmes and Dr. Watson. Conan Doyle fails to offer a bright line rule or workable legal standard for determining when characters are sufficiently developed to warrant copyright protection through an entire series, nor does it provide any case law that supports its position. Conan Doyle's proposed distinction runs counter to prevailing case law. See Siegel v. Warner Bros. Entm't Inc., 690 F. Supp. 2d 1048, 1058–59 (C.D. Cal. 2009) ("[T]he copyrightable aspects of a character… are protected only to the extent the work in which that particular aspect of the character was first delineated remains protected."); see also Gaiman v. McFarlane, 360 F.3d 644, 660 (7th Cir. 2004) (holding that once a comic book character was drawn, named, and given speech, it was sufficiently distinctive to be copyrightable). The effect of adopting Conan Doyle's position would be to extend impermissibly the copyright of certain character elements of Holmes and Watson beyond their statutory period, contrary to the goals of the Copyright Act. See id. at 661 (citing Lee v. A.R.T. Co., 125 F.3d 580, 581–83 (7th Cir. 1997)); see also Stewart v. Abend, 495 U.S. 207, 228 (1990) ("The copyright term is limited so that the public will not be permanently deprived of the fruits of an artist's labors."). Accordingly, the Pre–1923 Story Elements are free for public use.

Slip op. at 13–14 (record citations omitted, hyperlink and full citations to case law supplied).2

The common thread between the Turing pardon and the Holmes declaration of partial public domain status is that neither one deals with the entire issue... but each represents a step forward. That it is an insufficient step in the eyes of some activists3 does not make it an inappropriate step, particularly given that there were petitions for limited relief in front of the decisionmakers in both instances — a petition as to Turing only, and a grantable petition as to pre-1923-in-the-US works as to Holmes only. Criticizing a decisionmaker's choice to act within that decisionmaker's power is not a productive, constructive, or indeed valid criticism.


  1. disclosure: I have a great deal of personal respect for Judge Castillo. Although I never had an intellectual property case before him, I did have several other complex-litigation matters before him during his first years on the bench. Even in cutting-edge areas of civil procedure, his decisions were always both clear and nuanced, and provided more than adequate guidance to the parties to know what steps would be next — even, or perhaps especially, when ruling from the bench.
  2. I find it both revealing and disturbing that nobody appears to have cited the leading Supreme Court case on overextending copyright terms. See Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003). On one hand, it's at least understandable because Dastar concerned not a natural expiration of an undoubted expressive copyright, but a failure to renew a compilation copyright (that was rather weak in the first place). On the other hand, the core of Dastar is directly on point:

    The right to copy, and to copy without attribution, once a copyright has expired, like "the right to make [an article whose patent has expired]—including the right to make it in precisely the shape it carried when patented—passes to the public." Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 230 (1964); see also Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 121–22 (1938). "In general, unless an intellectual property right such as a patent or copyright protects an item, it will be subject to copying." TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 29 (2001). The rights of a patentee or copyright holder are part of a "carefully crafted bargain," Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 150–51 (1989), under which, once the patent or copyright monopoly has expired, the public may use the invention or work at will and without attribution.

    Dastar, 539 U.S. at 33—34. It takes Congressional action to bring an expired work back from the public domain... although this tangent is getting rather too extended.

  3. Regardless of the reason for the "chemical castration," I believe it should be completely pardoned (and, indeed, that reparations should be paid to any survivors... but not to their descendants) because it was objectively an inappropriate and inhumane means under any reasonable standard of human conduct. Put another way, if it was condemned either directly or by analogy at Nuremberg, it's not an appropriate way for the State to act.