Scrivener's Error: Warped Weft

You're Staring At My Story's Package

originally posted February 2010

This essay has little to do with salacious imagery... except insofar as copyright law in the US almost always involves the actual, human creator(s) getting kneed in the groin. Recent litigation and attempts to amend the Copyright Act — and, by "recent," I mean since the US joined the Berne Convention — have pretty uniformly reflected a meme that is not supported by the IP Clause,1 Berne Convention2 or other treaty obligations,3 or the Copyright Act4 itself.

Lemma: The dominant factor in determining the outcome of any dispute over creative IP rights is the package in which those rights are presented to end-users, regardless of distinctions found (or not found) within the legal framework that creates and protects those rights.

This is, perhaps, most apparent in the way that written works have been treated in litigation in the last fifteen years.

First, there was the ineptly handled Tasini matter,5 which concerned only copyrighted/copyrightable textual works that appeared in periodicals. As Tasini was not a class action, it ran into a two-element buzzsaw of bad civil procedure decisions in the Second Circuit, before the Supreme Court "clarified" matters. Muchnick6 is an attempt to expand the Tasini cause of action to include a class settlement; unfortunately, Muchnick is at least in part governed by the Second Circuit's ill-reasoned decision in Morris,7 one of the poster children for agency capture and self-interest by bureaucrats overwhelming the underlying legal authority.

At the opposite extreme, we've got the GBS litigation,8 which concerns only works published in nonperiodical text-on-paper form. Ironically, many individual works that appeared in periodicals at one time also appeared in books at another, which leads to some interesting fact patterns and maneuvering over remedies.

One example of the publishing package problem appeared in Ellison,9 which concerned unauthorized electronic display of Mr Ellison's copyrighted works as they had appeared in bound collections. Most of the individual stories concerned had been previously published — and individually registered, I might add — in periodicals. One package-related question that this raises, but that (due to the settlement reached after the Ninth Circuit decision) was not actually litigated: Is the collection Deathbird Stories — one of the infringed collections (and in that edition) — one unit (the book), 19 units (the individually registered stories), or 20 units (the stories plus the new material added to the book)? Then, too, there's the currently unanswered question of how Deathbird Stories is properly treated under the proposed GBS settlement, and indeed under the lawsuit itself. Consider, for example, whether Mr Ellison would be required to prove the provenance of an infringing copy of the story "The Deathbird" that appears somewhere on Google's servers (not just via index pointers). On the one hand, if it could be proven that it was from the book version — and that it was the version scanned by Google itself under the Google Library Project — it might be covered under the GBS settlement, and therefore liability had been already released (assuming Mr Ellison was both a proper and in fact member of the settlement class, which is another issue entirely). Conversely, one might question whether the provenance of the infringing copy was a different book edition of Deathbird Stories that was not included in the Google scanning, or to any of many other anthologies containing the story, or indeed to original publication in The Magazine of Fantasy & Science Fiction (which might — at least in theory — fall under Muchnick!). None of those possibilities, though, concerns the fact of infringement or the nature of the work itself — they concern only the delivery package, and subsequently the infringement package.

Another example, although perhaps a bit less obvious, is the Harry Potter Lexicon matter.10 In the Lexicon matter, nothing was filed until the proprietor of the Lexicon attempted to convert it from a diffuse set of webpages to a printed volume. Warner then interceded and brought Joanne Rowling along as a public face, considering (rightly) that people would be even more likely to direct their ire at a greedy film producer than at a greedy author. As I remarked at the time, though, the giveaway of what is really going on can be found in the copyright notices printed in the more-recent US editions of the novels:

Text copyright 2007 by J.K. Rowling
Illustrations by Mary GrandPre copyright 2007 by Warner Bros.
HARRY POTTER & all related characters and elements are TM of and copyright WBEI
Harry Potter Publishing Rights copyright J.K. Rowling
All rights reserved.

This makes excrutiatingly clear that we're really concerned, in the Lexicon lawsuit, with the form of the original works compared to the form of the alleged infringement. In short, all anyone really cares about is the package... and that is a consideration that is not relevant to the Copyright Act, which — except for works made for hire — does not distinguish among packages for textual works. The Copyright Act does distinguish among packages for certain types of nontextual works, such as phonorecordings; but we'll burn that bridge when we get to it.

Turning away from the Copyright Act for a moment, it becomes substantially clearer what we mean by "package" by looking at the commercial reality of publishing. As an example, Tad Williams's doorstop novel11 To Green Angel Tower (New York: DAW, 1993) illustrates the package problem rather clearly. A quick look at the entries at Powell's discloses:

  • A casebound edition, used, for the bargain price of $7.95 (original price on publication was either $25 from the publisher or $12.95 from the Science Fiction Book Club; this edition is no longer in print) — 1104 pages
  • A trade paper edition, used (2005), for $12.95 (original price on publication was $16.95, now $18.95) — 1104 pages
  • A mass-market paperback edition (1994), divided into two volumes; volume I, used for $4.50 (original price on publication was $5.99, now $8.98), and volume II, used for $4.95 (original price on publication was $5.99, now $8.98) — total of 816 + 816 = 1632 pages
  • An Adobe electronic edition, divided into two volumes; volume I for $7.98, and volume II for $7.98 — total of 816 + 816 = 1632 pages, or 49 + 32 = 81mb

And that's just the U.S. editions... and neglects the real problem: That To Green Angel Tower is not a stand-alone novel, or even the third book in a trilogy that can be read on its own, but the third publishing increment of a single work — Memory, Sorrow, and Thorn. It's also quite enlightening to look at the choice for the basis of an electronic edition, which chops up the novel into the mass market paperback form and pagination, rather than the trade paperback form and pagination. There might be a good production reason for dividing the mass market edition into two subvolumes12 — but that reasoning simply does not apply to e-books.

This really is about packaging. I chose Williams' novel because there is an unusual degree of fidelity — if you want to call it that — among the editions that I've personally examined over the years. Even though the mass-market paperback had to be reflowed for production, it maintains many of the same typographical and production errors that appear in the casebound edition (such as doubled commas), and contains no "new and improved author afterword" or anything like that. Although I've not put my grubby little fins on the electronic edition or the trade paperback edition for long enough to look, I strongly suspect the same — if only due to page-count identity, which is awfully hard to maintain for a book that long if one is doing actual changes and corrections between editions.

But the copyrightable expression is identical. As a Gedankenexperiment, let's put the copyrightable expression of To Green Angel Tower into Schrödinger's box.13 When the photon gets emitted, the copyrightable expression gets divided into either one volume, or two; on paper, or as electronic files; maybe even an audiobook edition or Braille edition. Now, without opening the box:

  • How many dead trees are involved?
  • How much does it cost Joe Consumer to purchase a new copy at list price?
  • Most difficult of all, if the book had been scanned and placed online — whether by Google or by some kid in his parents' garage — which edition has been infringed, and what does that mean for measuring damages if statutory damages are not available?14

Now imagine that perhaps — just perhaps — this was a reissue of a work that had been published, then went out of print, had the rights revert to the author, and then got reprinted by a different publisher just in time for the author's next novel to win the National Book Award.15

One other variable that goes into the packaging problem comes from the timing/pricing/returns problem endemic in the Western and Japanese publishing industry today.16 Although there are huge variations, a trade book-length widget typically comes out on date x for approximately price y in a casebound edition, followed on approximately x+360 by a mass market paperback edition priced at approximately 0.3y; there may be a trade paper edition, too, priced at approximately 0.7y in that mass-market paperback time slot, pushing the mass market paperback edition back about 300 days. On average. With huge variations. Further complicating efforts to determine real pricing, casebound and trade paperback editions are (ordinarily) fully returnable — that is, what you see in the store is not owned by the store, but is on consignment at the store, and is still owned by the publisher. For the purposes of this discussion, though, keep in mind that it is rare — not unheard of, by any means, but rare — for the copyrightable expression to be altered between any of those editions in any but the most minor ways. Thus, we don't pay for the words; we pay for the package in which we obtain those words.

To say the least, John Locke never imagined this potential problem when he worked toward what became the Statute of Anne.17 Similarly, the drafters of the Berne Convention, the US Copyright Act, and the various copyright-related WIPO treaties didn't consider it either. Next time, we'll start to see the real price being paid for that failure of imagination.

  1. U.S. Const. Art. I, § 8, cl. 8 ("[The Congress shall have power to] promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries").
  2. Berne Convention for the Protection of Literary and Artistic Works.
  3. See, e.g., the WIPO Copyright Treaty.
  4. 17 U.S.C. § 1 et seq.
  5. New York Times, Inc. v. Tasini, 533 U.S. 483 (2001).
  6. Reed Elsevier, Inc., et al., v. Muchnick, No. 08–103 (argued (U.S.) 07 Oct 2009; decision pending at this writing).
  7. Morris v. Business Concepts, Inc., 259 F.3d 65 (2d Cir. 2001), abrogated and replaced by 283 F.3d 502 (2d Cir. 2002).

    The agency capture lurking behind this decision was reflected in the Register of Copyrights' support for the defendant's position that each individual copyrightable element of a periodical must be separately registered; under that reasoning, the publisher's registration of the entire issue of a periodical does not cover the individual works appearing therein, unless the copyright of those works belonged to the publisher at the time of registration (as was required under the 1909 Act). Leaving aside the inconsistency between this analysis and the Berne Convention's derogation of formalities — and leaving aside the illogic of it all — the Register's position reflects an attempt to protect the Copyright Office's income stream from registrations. Cf. generally Conrad Peter Schmidt, Changing Bureaucratic Behavior: Acquisition Reform in the United States Army (2000) (esp. ch. 3) (PDF); John Hendry, Beyond Self-Interest: Agency Theory and the Board in a Satisficing World, 16 Brit. J. Mgt. S55 (Spr. 2005); Dan Kahan, Democracy Schmemocracy, 20 Cardozo L. Rev. 795 (1999).

  8. The Author's [sic] Guild, et. al. v. Google, Inc., No. 2005–8136 (filed 20 Sep 2005), class settlement proposed. n.b. If you're reading this essay on my blawg, I probably don't have to tell you that I've harshly criticized both the underlying lawsuit and the proposed settlement from the beginning.
  9. See, e.g., Ellison v. Robertson, et al., 189 F. Supp. 2d 1051 (C.D. Cal. 2002), rev'd in part and remanded sub nom. Ellison v. America Online, Inc., 357 F.3d 1072 (9th Cir. 2004). n.b. I was lead counsel for plaintiff Harlan Ellison in the matter; the caption in the Google scan of the West reporter for the appellate decision is wrong, as the Ninth Circuit itself had redesignated the matter Ellison v. AOL for all purposes.
  10. Warner Bros. Entertainment, Inc. v. RDR Books, Inc., 575 F. Supp. 2d 513 (S.D.N.Y. 2008). This was less a copyright suit than a trademark suit that borrows copyright remedies... which reinforces the point of this already extended essay.
  11. This is most emphatically not a judgment of the book's merit from either a literary or a fanboy perspective. It's actually quite a good book, and integrates its thematic material much more deeply into the plot, the characters, and the prose than almost anything else from that imprint in the early 1990s. It is, instead, a description of its sheer heft.
  12. Or maybe not; the layout and production quality of the mass market paperback editions somehow manages to combine wasteful and execreble. That's not a smackdown of all mass market paperbacks, by any means; William Gaddis's longer-by-wordcount unappreciated masterpiece The Recognitions clocks in under 1000 pages and yet is (physically, anyway!) more readable. It's not even a smackdown of all DAW, or all Penguin USA, mass market paperbacks.
  13. The cat is, in fact, dead already, because we've already done this experiment — so long ago that the cat died, and it doesn't really matter whether or not it was in the box when it did so. All of this points out a critical error in Schrödinger's frame: It does not control for time elapsed between the event and observation of the event's effects.
  14. In this particular instance, they would be, as the casebound edition was registered... albeit not timely (see No. TX0003857430). Like I said — Gedankenexperiment. Of course, this also brings up the question of whether Mr Williams might have a claim against DAW... and any professors out there perusing this have my gleeful permission to include precisely such a fact pattern on their exams. It brings up further questions regarding other registration problems, too...
  15. This is not a hypothetical; see the work listed in note 12 supra for one example.
  16. I simply don't have enough data to make either a similar or a dissimilar inference for China and other emerging print-publishing markets, or even for established print-publishing markets like Latin America. There's even more chaos than one usually finds in publishing!
  17. Happy 300th birthday in about six weeks! See 8 Anne, ch. 19 (10 Apr 1710).

Law and reality in publishing (seldom the same thing!) from the author's side of the slush pile, with occasional forays into military affairs, censorship and the First Amendment, legal theory, and anything else that strikes me as interesting.

Ritual disclaimer: This blog contains legal commentary, but it is only general commentary. It does not constitute legal advice for your situation. It does not create an attorney-client relationship or any other expectation of confidentiality, nor is it an offer of representation.

All material © 2003–11 except where otherwise indicated. All rights reserved. N.B. This blawg does not use the Creative Commons License, although I'm usually pretty good-natured about permissions for attributed reuse.


Warped Weft

Essays in Warped Weft are consolidated from Scrivener's Error with only formatting corrections (and the occasional typographical correction).