Monday, June 20, 2005

The More Discerning Observer Test

Friday's blog dealt with the ordinary observer test; today's with the more discerning observer test. The more discerning observer test is employed where there is present in both parties' (but especially plaintiff's) a considerable amount of public domain material. The ordinary observer test is inappropriate in such a circumstance because unless the public domain elements are disregarded in the ultimate substantial similarity determination, that analysis will be skewered; it will result in the trier of fact finding similarities resulting from material that plaintiff does not have an ownership interest in. And from a policy perspective, failure to exclude substantial amounts of public domain material will also result in fewer derivative works based on that material.

There is a countervailing risk, though, and that is dissection of a work, such a work of visual art, where its originality (even when based on public domain elements) is based on the interrelationship between the constituent elements. When those elements are factored out and then separately compared or left out entirely (as pd), there is a risk that the trier of fact will overlook the creativity in the whole. (This concern with the whole is what led to the earlier "total concept and feel" test). Yet, where the unprotectible elements have a significant impact on the overall total concept and feel of the parties' work, it is wrong to ignore that impact: a thin copyright can, thereby, become impermissibly fat based on pd elements.

The Second Circuit tackled these problems in Folio Impressions, Inc. v. Byer Calif., 937 F.2d 759 (2d Cir. 1991) and directed the trier of fact to only consider the protectible elements, after factoring out the unprotectible elements. In Knitwaves, Inc. v. Lollytogs, Ltd., 71 F.3d 996 (2d Cir. 1995), however, the court backtracked some, describing Folio Impressions as involving "rather specialized facts," and expressing a concern that if one dissected works of visual art too far one, would be forced to say "there can be no originality in a painting because all colors of paint have been used somewhere in the past." This statement was hyperbole and as such not really helpful.

The solution may be that while the more discerning observer test factors out unprotectible elements, a total concept and feel analysis is still employed, but using the protectible elements in the parties' works.

The principles are not really that difficult to articulate (I accept others may have different ways of articulating them), but they are very hard to apply in practice. The Second Circuit's Tufenkian Import/Export Ventures, Inc. v. Einstein Moomjy, Inc., 338 F.3d 127 (2d Cir. 2003) opinion is a good example. (I played a very minor role on appeal for defendant). Plaintiff took substantial public domain elements from diverse sources and combined those elements in an unusual way, modified them, and added a few new elements. District Judge William Pauley found the work original although with a very thin copyright. (The court of appeals seems to have taken an even dimmer view on the scope of copyright issue than did Judge Pauley.) Defendant took the same public domain sources, modified them yet differently, and added new material too. Judge Pauley granted summary judgment of no infringement; the court of appeals reversed, granting summary judgment to plaintiff.

A correct analysis in that case entailed the following steps: (1) identification of the public domain elements in plaintiff's work; (2) identification of plaintiff's new elements; (3) identification of the public domain elements in defendant's work; (4) identification of those elements in defendant's work copied from plaintiff; (5) identification of the new elements added by defendant. In the end, the substantial similarity analysis, using the more discerning observer test, would have to exclude the pd material in both parties' works and would also have to exclude the new elements added by defendant. Infringement could only be found if the new elements added by plaintiff (including the original modifications of the pd material) constituted a material part of defendant's work.

In Tufenkian, I believe the total concept and feel of the parties' works (limited to only the protectible elements) was not materially similar. I further believe the court of appeals ignored (as Judge Pauley had not) the new elements added by defendant and that the circuit significantly downplayed the overwhelming impact of the pd elements. Both parties' works were exquisite high-end carpets, and both should have been left to compete in the marketplace, with plaintiff's thin copyright protecting him only against verbatim or non-verbatim copying, neither of which had occurred in that particular dispute.


Anonymous said...

Thanks for the great post and the great discussion that's sure to follow.

I think the "more discerning observer" is a lot easier to apply in a literary, rather than a visual context, although, AFAIK, it gets used a lot more in the VA context.

To that extent, I wonder if it is even possible to apply the more discerning observer test fairly in a VA case without making the standard "novelty."

As I understand it, in Tufenkian Defendant took the same, unrelated public domain elements and used them in a similar fashion. The yoking of different PD elements, which aren't natural to put together, to me, infuses the work with enough creativity to be protectible as a whole.

A quick example: I make a painting that's a combination of American Gothic and the Scream. Someone comes along and makes a very similar painting. In comparing our two works, should the Scream elements and the American Gothic elements be ignored, even though there was no natural way to combine them until I did it? (let's also assume that there is more than one way to do this combination).

Granting my painting "thicker" protection has taken nothing out of the public domain. Artists are still free to use either painting, and even to combine them. Assuming that the defendant did copy my work, does the argument that the works are "public domain" serve the underlying goals of the copyright act?

A final point-- my example above is an easy case because the PD sources are obvious. However, and this is from experience, using the More Discerning Observer test, the Defendant can/will produce as much of the "prior art" out there and say, "Well, if you look at the thousands of examples we've produced, you can see that Plaintiff's work consists only of public domain elements." This is the "it's out there somewhere" defense.

William Patry said...


Judge Pauley's district court opinion Tufenkian is a really good illustration, I think, of how the issue should be handled. He had the parties bring the full-sized rugs into the courtroom and he spent alot of time carefully examining them and getting the history of their development. He credited plaintiff's creativity in combining two pd elements in an unusual way, adding other elements, taking out other elements and melding it into a unified visual whole. Plaintiff deliberately departed from the symmetry of the central pd work to create a different feel. Judge Pauley gave that whole protection, but thin protection due to the still overwhelming predominance of pd material.

Defendant copied, I believe, plaintiff's choice of pd elements, but he adopted a very different feel, one closer to the original pd one (i.e., symmetrical), plus he added his own original elements. It is a very expensive rug. I have one in my work office. I wanted to get plaintiff's too (I could then teach a copyright class without leaving my office!), but I didn't like the way it looked. I disliked the asymmetry that was one of the principal features of what plaintiff changed from the pd work, and, importantly, defendant did not copy that change.

What Judge Pauley did was exactly right for me: he credited plaintiff's originality, excluded the pd stuff, credited defendant for his creativity but nevertheless compared the toal concept and feel of the protectible parts of plaintiff's work against defendant's. I don't think you can do more than that.

Anonymous said...

typo: you mean to say "verbatim or near-verbatim copying" [not non-verbatim]

anyway, a good post on an outstanding blog

Anonymous said...


I have read the Court of Appeals' opinion reversing the SDNY's finding in Tufenkian. The Court's citation to Knitwaves v. Lollytogs doesn't increase my confidence in the accuracy of its decision. The "coordination" of squirrels, leaves, and fall colors is not original, as we all remember from our elementary school classrooms.

The district court's opinion in Tufenkian included pictures of the accusing rug, the accused rug, and the public domain sources. Unfortunately the picture of the accused rug, the Bromley 514, is not very clear (at least in the copy I have). As a result it is hard to make an independent evaluation of either the SDNY's decision or the 2nd circuit's

But while I can't be completely sure (until I see clearer pictures of the rugs) that the Court of Appeals is wrong, I am (as I noted earlier) somewhat uneasy with its decision. Apart from my dislike of the citation to Knitwaves v. Lollytogs, I also get the impression that some judges on 2nd Circuit's Court of Appeals havn't completely cleared their minds of the superstitious notion that all copying is somehow inherently dirty.

I think I am less afraid than Tom of letting the standard of originality for copyright evolve, in some situations, into a de-facto standard of novelty. A trend in this direction has already been noted in musical copyright cases by at least one scholar. But however that may be, I think that if we have a doctrine of "unconscious infringement", it's only fair to allow the "it's out there somewhere" defense in some circumstances. If the defendant can unconsciously copy from the plaintiff, why can't the plaintiff unconsciously copy from the public domain ? If he did so, then what he copied from there should be able to be copied from him.

Anonymous said...

I am interested by the issue of language proficiency in court evaluations of similarity whether under the discerning observer or the original observer test; and by this I mean whether the court, in practice, can speak the language of the authors of the works in question.

In a music case in which the court knows music and can play an instrument proficiently the analysis and frequently the actual outcome will differ from a court that doesn't. I don't read music and I am always impressed by how Hand and Holmes and the other jurists of that day all knew how to read and play music as a matter of course - - no T.V.. They would never have considered an expert to determine whether in a European music idiom two chords played one after the other were original or common. That is not true today. I couldn't find a jury in America in which they all read and play music and I bet less than half of the Federal judges do proficiently.

This is also an issue in visual art cases. To the "un-schooled" two similar drawings or arrangements of geometric forms can be very similar when observed cold and quite extraordinary when observed with the language and context of fine art - - a language even fewer juries are likely to know.

Experts work and, of course, can be bought to cancel each other out but rarely on what I would call “language” issues. An element is either common or not and experts can usually take each other on convincingly when one or the other decides to go off the reservation, so to speak (in an American idiom that would seem original, perhaps, in Uzbekistan). But the only experts sitting in the court of appeals are the three judges and frequently their clerks who get their backs up (who know what that means in East African dialects) when someone suggests they might not know what they are talking about, or more importantly, that they might not know the language of the works at issue in the case. They usually know what they see when they see it - - know what they hear when they hear it - - and that’s a big problem in similarity cases. It goes a long way in explaining why they keep trying to come up with fancy-sounding analytical frameworks to cover for the fact that they are shooting from the gut.

William Patry said...

Thanks for the typo correction, Anonymous. I am a poor proof-reader and worry about it, but I have always done all my own research (never using law students or associates for anything), and my own writing and typing, so it is a proverbial difficulty. On the point about judges' proficiency, there is a belief that ignorance is somehow purer (the Second Circuit once reversed a district judge who professed to a more "Olympian perspective" than the common folk, the Am H'aretz), but I confess to siding with those who think that more knoweldge is better than less. There is an interesting CD, published by the Library fo Congress, called I think, "Field Recordings" that has two performances by Learned Hand singing sea chanties and explaining in a surprisingly strong accent how he came to learn them. among current judges, the most musically proficient is Judge Owen in the SDNY.

Anonymous said...

Judge Learned Hand singing "Iron Merrimac" along with songs performed by many others are available on CD for $17 on the Library of Congress website. The url for the selling page is too long and will not format to the columns on this blog. Just go to the LOC main home page and search for [field recordings "learned hand"]and the first result will take you where you want to go. Work of the United States government of a pd song with the recording unprotected even under state law and I'm still paying $17. Go figure. Thanks for the tip!

Anonymous said...

The "coordination" of squirrels, leaves, and fall colors is not original, as we all remember from our elementary school classrooms.

That's an interesting way of putting it, and I do agree, in part. But, there's an inherent difficult in expressing a visual work in words. I don't think a painting of jazz musicians is original per se, either, but then we have a Picasso that is certainly original.

And, I think this is a bigger problem than it seems because of the reductive power of works. It's very easy to reduce a complex (or more complex) work into a few easy words and make it sound simple and unoriginal.

Anonymous said...

I don't think a painting of jazz musicians is original per se... wrote Tom.

If it's a painting, then it is original per se unless it's an extreemly careful forgery attempting to replicate brushstrokes and slavishly copy every compositional element. It may be an infringing work because of its similarity to another work, or a thinly protected work if it borrows a great deal from a pd work but a painting will certainly contain original auhorship as well.

Anonymous said...

Re: Idea for another. . .

I 100% agree with you. I was building on Timothy's comment that says squirels, etc. are not "original."

Now is as good a time as any to throw out "idea" and "expression." The idea is not original, the expression may be. And, really what we're always talking about with infringement is the expression.

I doubt one could, in a court or otherwise, prove an idea was original. If so, I'd like you to be my lawyer, pro bono, for sure.

And that's the problem with the "novelty" standard for copyright. The idea is never novel and all forms of expression are continuously recycled.

I think in general, not enough credit is given to arrangement (not in a musical sense, for I know nothing about that) in the copyright law. There is lots of talk of "elements" in opinions, and I don't think that's a good paradigm for thinking about a creative work.


Anonymous said...

The more discerning ordinary observer test, if used in music copyright cases, would deem most music music created in the 20th centrury as un original. If such a test was used in Three Boys Music v. Bolton, Santrayall v. Burrell, wouldn't the outcome have been quite different as each of these cases found infringement but on the combination of otherwise unprotectable elements. If these were pulled out, the outcome would surely had been different.Why weren't they pullled out?