Friday, August 11, 2017

IPSC Breakout Session V

Copyright Theory (my apologies; I had to leave early for a flight)

Redundancy and AntiRedundancy In Copyright – Oren Bracha & John M. Golden
Doctrines that overlap: fair use/functionality; fair use/improper appropriation; copyrightable subject matter originality; reproduction/derivative works rights (cautionary example).
Forms of redundancy: bidirectionally partial, standard Venn diagram; unidirectionally partial—safe harbor laid over a more general standard (§512 and secondary liability); complete (arguably “aid & abet,” “arbitrary & capricious,” eBay test prongs 1&2).  Fed. Cir. has resisted redundancy in things like patentable subject matter, SCt has pushed back.

Value in redundancy: error limitation and clarity at core; potential administrative advantages (dismissal/SJ, for example in Greatest American Hero case where the court says no improper appropriation w/o needing full fair use analysis/discovery); potential robustness against strategic behavior; evolutionary potential.  Copyrightable subject matter/originality: Copyright Office treatment of yoga sequences—usually not w/in CSM.  9th Cir.: functionality b/c of psychological and health benefits (claimed).  You won’t always have that kind of clear, easily available evidence of functionality in admission against interest.  Series of rules, each of which is an imperfect filter, might be the best.

Bad example: Harry Potter case—court decides Lexicon isn’t a reproduction but is a derivative work; the rights may not be producing good overlaps.

Buccafusco: how do these overlaps affect litigation choices and do they affect it asymmetrically? [I think w/TM the overlapping defenses can cause trouble, as Bill McGeveran has documented, b/c if you pick the wrong one you may be out of luck].

A: may cut both ways.

Rosenblatt: there are redundancies in protections, exceptions, liability doctrines (indirect infringement)—these might operate differently from the standpoint of litigants, creators, owners.

A: Maybe, and the point is more thought about design of overlaps and functions—backstops or partial redundancy or coequal partners.

Quantifying Copyright – James Grimmelmann
We encode almost anything digitally and count the bits—the number is a measure of complexity. Some number of bits can likely be squeezed out by compression, and those bits didn’t really matter in the first place.  You’re throwing out information, but if you do it right/high enough, you can throw out only information that humans didn’t care about.  Lyrics to Happy birthday have internal redundancy; can be compressed to subsequences that appear in the digital work.  Information theory: communications, processes, encoding, noise (redundancy)—Fromer & Scafidi have both done work in IP under the heading of redundancy. It’s about communications systems.  When do we want to be efficient and when redundant so we can recover the message even in presence of scrambling & noise?  Other branch of info theory: algorithmic.  It’s about computation, compression, individual works—a work is only as complex as its shortest encoding.

Tentative idea #1: Feist says there are hundreds of millions of ways to select 50,000 listings; about 1,500,000 bits to describe an arbitrary selection—but it’s a lot easier to describe Feist’s actual selection, and that is an indicator of lack of original creativity.

Tentative idea #3: quantify factor 3: a compressed image could be seen as 100%, but you could also look at how much human perception has been discarded going from big to small, which would be less.

Rule-based creativity: can’t extract more bits of expression than you put in; merger kicks in when there are only a ltd number of ways to express an idea. Scenes a faire are about predictability: in a hard-boiled detective novel, it adds almost no new info to learn that the hero drinks.

More ambition: you could try to directly quantify expression and do filtration and similarity tests.  Problems: full complexity is uncomputable exactly. We can only approximate. Also, that leaves off psychology and aesthetics and all the reasons people actually care about having expressive works.

Boundary arguments: there are some situations where there is no room for expression; you could use it as an input into expert testimony. Maybe something in exploring lossy compression and its relationship to psychology.

Linford: doesn’t music in Happy Birthday add a level of complexity, which also makes it difficult to compare text to music.

A: tells you immediately that musical works are simpler than sound recordings.

Christina Mulligan: how would you do the hard-boiled detective thing? You need a baseline for defining the genre.

A: Hard.  Mainly wants situations where there are actual numbers—file sizes for thumbnails.  A way of thinking about questions that involve choice and constraint.  Predictability/compressibility are throughlines that explain a lot of appeals to number of available options.

Rosenblatt: could really be misused where psychology etc. means that perceptions are completely different but analysis finds few differences.

A: yes, this does a faceplant in transformativeness. It clearly informs “extrinsic” analysis but has nothing to say about intrinsic analysis.  This is a formalist approach.

Buccafusco: compared to what? Everything else sucks; this approach only has to suck less.

Sheff: you might need different encoding selections to measure and to compare.

A: K complexity is coding independent.

Rewarding Derivative Works – Joseph P. Fishman

Derivative works’ share of all films’ combined box office keeps climbing. One explanation: it’s how they deal with risk.  But that’s not enough.  Derivative works rights subsidizes this investment, and not other works.  Franchises on average do earn more over production budgets than non-sequels. Merchandising is also huge: $118 billion entertainment/character related merchandise; $38 billion box office receipts. 

Jennifer Rothman: TMs are part of this; also, cross subsidization of smaller films both by studios and by directors/etc. other people who fund what they want to do w/the bigger films.

A: yes on TMs; maybe he needs a better term than adaptation. 

Q: Many novels were originally posted as serials; they weren’t short stories. Maybe we’re finding that longer visual stories are increasingly popular, and they come in segments.

Then I had to leave.

No comments: