A Fresh Look at Tests for Nonliteral Copyright Infringement

From AcaWiki
Jump to: navigation, search

Citation: Pamela Samuelson (2013) A Fresh Look at Tests for Nonliteral Copyright Infringement. Northwestern University Law Review (RSS)

Download: http://ssrn.com/abstract=2218838



"Nonliteral infringement may arise, for instance, when a second comer appropriates detailed plot sequences from another author's drama but uses different dialogue."

"There are too many tests and not enough guidance about which one to use in what kinds of cases.Occasionally,courts have applied several different tests, not being sure which is the right one." 5 described in paper.

A) Arnstein v. Porter: The Copying and Improper Appropriation Test

Two steps:

  1. copying happened (if no admission or witness, expert testimony about similarity substitutes)
  2. given (1), lay impression of inappropriate appropriation.

Author identifies three flaws:

  1. confusion about two different notions of similarity, which subsequently have been called "copying-in-fact" and "probative similarity"
  2. puts defendant at disadvantage because first step has already stipulated copying occurred, particularly unfair if first step decision relies on uncopyrightable elements
  3. should only be employed in case of independent creation defense

B. Krofft v. McDonalds:The Extrinsic/Intrinsic Test

Two step test "consciously adapted from Arnstein":

  1. extrinsic, includes expert analysis
  2. intrinsic, lay opinion

Difference with Arnstein is both steps focus on idea/expression divide, not creation. But still flawed:

  1. extrinsic/intrinsic confusing and inapt terminology
  2. "substantial similarity of ideas" inappropriate for copyright
  3. based on misunderstanding of Arnstein first step, which is only about copying, not ideas
  4. second step core inquiry came to be "total concept and feel", problematic, see next

C. Roth Greeting Cards: The Total Concept and Feel Test

Sometimes viewed as one-step test, sometimes viewed as second step of Krofft.


  1. does not focus on copying of specific expressive elements, which is what copyright case should focus on
  2. too easy for uncopyrightable elements to be swept into analysis
  3. unpredictable outcome, in particular by not limiting test to artistic or fanciful works
  4. difficult to square with first amendment
  5. illogical, eg in case of greeting card case -- few words thus uncopyrightable, different but similar art not copying, but "TCF" results greater than sum of parts

TCF rarely declined by courts, used in a very wide variety of cases including "fabric designs, masquerade costumes, slippers, duffel bags,rugs, jewelry, dolls, children's books, films and television programs, commercials, videogames, and buildings" but fallen out of favor for software.

D. Nichols v. Universal Pictures: Judge Hand's Abstractions Test Two steps akin to Arnstein:

  1. "dissective"
  2. "ingenuous"

Envisions a series of "patterns of abstraction", too far up which are unprotectable ideas.


  1. some very detailed and concrete elements are unprotectable
  2. encourages unpredictable outcomes

E. Computer Associates v. Altai: The Abstraction, Filtration, & Comparison Test

Described in The Strange Odyssey of Software Interfaces and Intellectual Property Law.

Expert/dissective analysis pertinent to all steps.

"Finding infringement is warranted If there is substantial similarity in program expressions and the these similarities resulted from copying.

Recognizes "that copyright protection does not extend to efficient program designs, standard programming techniques, and elements of programs dictated by external factors".


  1. "failed to call for exclusion of program 'procedure, process, system, [or] method of operation,' even though the copyright statute explicitly states that '[i]n no case does copyright protection extend' to such elements"


Author proposes 7 improvements:

  1. "triers of fact should do some analytic dissection and make some overall impression judgment in virtually every nonliteral similarity case, and the impression judgment should be done keeping in mind the existence of unprotectable elements in the works at issue."
  2. "courts should be more explicit than they have generally been about which elements in protected works are expressive, and which are unprotectable,as they go about making judgments as to nonliteral infringement"
  3. "dissection should be given more emphasis than impression when cases involve works that have a 'thin' scope of protections to ensure that unprotectable elements are not inadvertently protected in the impression step"
  4. "impression should be given more emphasis in cases involving artisticand fanciful works."
  5. "courts should be careful about the scope of permissible testimony to be offered by experts in non literal infringement cases, there should not be a blanket rule against admitting such testimony"
  6. "plaintiffs in non literal similarity cases should have to prove not only that the defendants' work is substantially similar to theirs, but also that the defendants appropriated a sufficient quantum of expression from the plaintiffs to have caused commercial harm"
  7. "only circumstances under which it is necessary to inquire whether the defendant had access to and copied something from the plaintiff's work is when the defendant has raised an independent creation defense. If the works at issue are not substantially similar in their expressive elements, there is no need to inquire about copying or independent creation. Independent creation defenses should thus be the last stage of an infringement analysis when they are raised"

Conclusion: "The easiest way to avoid the 'viscous quandary' that nonliteral infringement cases present would be to return to a rule that only exact or near-exact copies(i.e., servile imitations) are infringing" but this is unlikely and nonliteral infringement regime may encourage creation of more diverse works. Proposed improvements would make regime more coherent and consistent.