Tuesday, June 07, 2005

Does Copyright Registration Matter?

In the heady days of Berne adherence in 1988, the air was full of anti-formality sentiment. Maybe, the sentiment went, U.S. law was finally moving toward the formality-free approach taken on the Continent. Maybe, we would at long last be able to hold our heads high whilst walking along the Seine to attend yet another profound WIPO symposium. Maybe, we could at long last engage in all that fancy-schmanzy natural rights theorizing so beloved by Europeans, without worrying that behind our backs we were being laughed at for our colonial attachment to formalities.

There was no basis for such optimism. The 1990 VARA has a numbering and marking requirement, a true non sequitur for a moral rights provision. In 1992, the Copyright Office and book publishers fought efforts to repeal Section 412 (which conditions an award of attorneys fees and statutory damages on a pre-infringement registration) as if their very lives were at stake. The 1998 DMCA's copyright management information provisions have resurrected mandatory notice provisions, but with a reverse twist: it is users who have to worry, not copyright owners.

Formalities have always been an important part of U.S. copyright law and seem destined to remain so. The policy debates on them are well-known, and I don't want to take sides. Instead, since they exist, I want to point out one consequence: the importance of having the right kind of registration for "United States" works as a jurisdictional prerequisite to the bringing of an infringement action, a requirement expressed in Section 411(a).

Contrary to Nimmer's blatant misreading of the statute (he asserts one need only have sent in the check, the deposit copy, and the completed application), one must have either the issued registration or a rejection in hand. Failure to do so results in the court not having subject matter jurisdiction and thus having to dismiss the matter (without prejudice).

Claims of infringement in derivative works pose special problems. There are two scenarios, but each share this common fact: a work has gone through a number of iterations, but there is only one registration and plaintiff seeks to use that one registration to satisfy Section 411(a).

In scenario one, the registration is for the first, original version. No subsequent registration has been made for later versions. I call this a "forward-looking" registration. In scenario two, the registration is for the most recent version. No registrations have been obtained for earlier versions. I call this a "backward-looking" registration.

While backward-looking registrations satisfy Section 411(a), Murray Hill Publications v. ABC Communications, Inc., 264 F.3d 622 (6th Cir. 2001); Streetwise Maps, Inc. v. VanDam, Inc., 159 F.3d 739 (2d Cir. 1998), forward-looking ones don't, Johnson v. Gordon, 2005 U.S. App. LEXIS 9902 (1st Cir. May 31, 2005); Well-Made Toy Mfg. Corp. v. Goffa, 354 F.3d 112 (2d Cir. 2003).

The reasoning behind the distinction is fairly straightforward: with backward-looking registrations, the material in which copyright is claimed has been examined by the Copyright Office and is on public file. With forward-looking registrations, the Copyright Office has not examined the claim and the material is not on public file.


Anonymous said...

Prof. Patry:

Great blog . . . however, methinks you overstate things a bit.

Regarding your statement:

"Contrary to Nimmer's blatant misreading of the statute (he asserts one need only have sent in the check, the deposit copy, and the completed application), one must have either the issued registration or a rejection in hand. Failure to do so results in the court not having subject matter jurisdiction and thus having to dismiss the matter (without prejudice)."

In the Fifth Circuit this is demonstrably wrong. See, e.g., Apple Barrel Productions v. Beard, 730 F.2d 384 (5th Cir. 1984). Shep this case out and you'll find that Apple Barrel is followed religiously by district courts in the Fifth Circuit (indeed, CTA5 recently reaffirmed it), as well as elsewhere. See, e.g., International Kitchen Exhaust Cleaning Association v. Power Washers of North America, 53 USPQ2d 1505 (D. D.C. 2000).

Whether the Apple Barrel rule is intellectually "correct" is, of course, a point on which reasonable minds can differ, and other courts have done so. (I could similarly disagree with any number of truly goofy copyright decisions out of the First and Ninth Circuits, but my disagreement would hardly change such precedents in those circuits.) Nevertheless, your argument simply won't fly in a district court in the Fifth Circuit . . . I have defeated many, many motions to dismiss by invoking the Apple Barrel rule.

LKB in Houston

Anonymous said...

The registration requirements of 411 seem particularly weird in that they treat US copyright holders as third-class citizens, by imposing a requirement that doesn't have to be met by other potential litigants. Those other two classes include foreign copyright holders (or rather, holders of copyright in non-US works), who need not register to sue. The other class would be criminal prosecutors, who would appear to be able to bring criminal charges for copyright infringement (under 506) even without registration. (The statutory text speaks of "actions" for infringement, which isn't exactly clear, but would not seem to include prosecution, while the committee notes state that registration is a prerequisite to a "suit" for infringement, which would not include a criminal prosecution.)

Anonymous said...

Well, that's not really making them third-class citizens as much as it's indicative of the error we've made in not making registration a more universal requirement.

William Patry said...

I have an extensive discussion of Nimmer and Apple on Section 411(a) in my new book, explaining why Nimmer is so off-the-wall on this point. If anyone is interested, give me your email and I'll send it to you

Anonymous said...

I would be really interested in that discussion in your book. Pls email me at tjm60614@yahoo.com. I think you are right, but I prefer the incorrect result.

William Patry said...

the record, I am in favor of repeal of Sections 411(a) and 412. I do think, however, that the Copyright Office serves a valuable examination purpose and thus would like to find some carrots without the present sticks. But since 411(a) is on the books, I think it should be enforced as written and not tortured out of existence as Nimmer and Apple Barrel do. Baruch HaShem that I live in the Second Circuit.

Anonymous said...

Prof. Patry:

I would like to see your analysis of Apple Barrel / Nimmer. Please e-mail to bonham@oshaliang.com.

Please note that I don't dispute that Apple Barrel is a tortured interpretation of the statute (or that Nimmer -- particularly Nimmer the Younger -- can be prone to rather bizarre conclusions). My point is that correct or not, Apple Barrel is settled law in CTA5.

BTW, I share your belief that 411(a) and 412 should have been repealed or radically modified long ago . . . but any change that threatens the livelihood of an entire class of government employees isn't likely to occur.

LKB in Houston

Anonymous said...

I am really interested in your opinion that the Copyright Office "examination" procedure serves any purpose at all. As I see it, their main job is to reject things that aren't original enough to be granted protection, for one reason or another.

My feeling on this is "who is the copyright office to say my work doesn't have sufficient originality." Let a potential defendant make that argument, rather than imposing that hurdle at the start of the process.

Further, the rejection serves merely to prejudice a potential plaintiff, as they can bring suit anyway, but now have an adverse ruling against their work.

And, lots of things that probably should be rejected slip through, effectively making the examination a crap shoot.

Secondly, I don't think the Office's interpretations of the law are all that sound.

I am currently involved in a case regarding copyright on fully clothed teddy bears. Defendant cited up and down the Office's position on copyrightability for clothing, and rejections the Office had sent out for clothed teddy bears.

Fortunately, the judge was not persuaded by the Defendant's (and the Office's) reasoning.

Anonymous said...

Ideally we should withdraw from the Berne Union and apply simple formalities equally to all comers. Indeed, one of my solutions to the "orphan works" problem (besides the obvious one of reducing the duration of copyright) is to require the rightsholder to maintain the right to bring actions by renewing the registration periodically. The copyright itself will not lapse should the registration go stale. Only the right to bring actions or (alternatively) the right to bring actions for relief beyond the injunction. As long as the copyright lasts, the full right of suit would be restored as soon as the registration were updated.

As fas as abstract discussions in the Latin Quarter go, it is the "natural rights" theorists who should hang their heads in shame for refusing to recognize the scope of mankind's rights of free trade and free expression. The exclusive privileges of copyright are carved out of the margin of these inherent rights, theoretically with the consent of those who surrender them temporarily. Theoretically. In practice, they have expanded to where they have become unreasonable burdens on the public.

Anonymous said...

Timothy, I generally agree with you, but I think it would be better to follow tradition and end the term if the copyright holder fails to follow through with the periodic formalities. If his rights are so unimportant to him that he can't keep up with what should be a very low burden, then who are we to argue?

This roughly follows traditional copyright law, patent law, trademark registrations, etc. It's a good system.

I would also agree with Tom that the current CO isn't that great, but this to me only indicates that we need to reform the system and stop treating the CO as a rubber stamp. It is far better, after all, for plaintiffs to bear the burden demonstrating copyrightability, than to burden the world with improper claims of copyright, and the chilling effects that result.

Anonymous said...

I've got to go the other way on that issue, anon. Due to sheer volume, I don't think the Copyright Office can do a great examination. I'd much prefer a rubber stamp system, in this instance.

The same type of work, if not the same work, will be treated differently based on who at the Copyright Office looks at it.

(An aside: I think, in general, the USPTO does a good job of examination and reducing costs to society and trademark holders).

Anonymous said...

Professor, Patry, I looked at the Murray Hill v. ABC case , 60 USPQ 2d 1080, and I agree with the court that one can't bring an action for infringement of matter unique to a derivative work -- for infringement of the derivation -- unless the derivitave work be first registered. But I think they too blithely applied their rule to the case at bar. As far as I can make out from the report, the music of the underlying work "Jeanette", and the derivative work, "J. P.'s Theme", was the same. Only the words differed. The lack of registration only prevented plaintiff from suing over infringement of the words of "J.P.'s Theme", not from suing over the melody. An infringing use of "J.P.'s Theme" would necessarily have infringed the melody "Jeanette", in which a claim of copyright was registered with the Copyright Office. So I don't see how the court could avoid, as it claimed to be able to do, the question of whether the Foundation should have been a party to the suit and whether there was an implied license that made the use of "J.P.'s Theme" non-infringing.

It will be easier to evaluate the court's decision once the music at issue has been posted to Columbia's music copyright infringement case web site at


Anonymous said...

Your last paragraph in the original post (quoted below) correctly states the view of some cases, but does not raise a problem inherent in those decisions — a problem that may lead to confusion in many fact situations:

"The reasoning behind the distinction is fairly straightforward: with backward-looking registrations, the material in which copyright is claimed has been examined by the Copyright Office and is on public file. With forward-looking registrations, the Copyright Office has not examined the claim and the material is not on public file."

In "backward-looking" registrations the claim of authorship only relates to the new material added to the underlying work, so a registration should only provide jurisdiction for that new material. If no registration was made for the underlying work, the infringement suit should not address this material, notwithstanding what some courts have said.

While it is true that for practical purposes, the Office has received the entire derivative work, the registration only covers the new material added.

In "forward-looking" registrations, the contrary should be true -- the registration for the underlying work should cover any infringement of the underlying material, but cannot include infringement of the new, unregistered material.

William Patry said...


I'll ask my friends at the Copyright Office, but I am pretty sure that if the first registration for a work is in, say, version 5.5, the registration will cover everything in the deposit, including everything in versions 1 through 5.4. If I understand your comment, you seem to think the registration would cover only the new material from 5.4. In my opinion that would be true only if there had been a previous registration for 5.4

Anonymous said...

Prof. Patry--
Shouldn't a registration for v5.5 only cover those portions of previous versions which are present in v5.5? Anything that isn't present, presumably hasn't come before the CO by means of the v5.5 registration.

William Patry said...


I should get the answer according to the Copyright Office tomorrow (I don't have my Compendium II of Copyright Office Practices at home with me), but my view is based on my recollection that your approach is true only where there has been a prior registration for v1-4; the reason being that it is the registration that deterimes the scope of the registration and not mere status as a derivative work.

One possible error in my analysis involves screenplays where the movie has already been registered or publisherd (the latter being the key); here, as I recall, a registration for the screenplay will only cover material not in the movie. This has been criticized, though, since the screenplay is obviously not derivative of the movie; if anything its vice versa. But if I am right about this, the screenplay approach would apply also to say v5 of software where v1-4 were published, but not registered.

Hopefully, I'll be able to sort this out more tomorrow.

Anonymous said...

Copyright Registration at