The Nature and Scope of U.S. Copyright Law

(Presented at the NYC Bar Association on April 10, 2012; revised Jan-Mar 2013.)

Copyright may be the closest that American law gets to Chaos as scientists started using the term at the end of the last century: a system with so many variables constantly in play that, by definition, its development and at times even its basic character and shape are unpredictable and often not rational in the way that we expect a legal regime of its vintage to be. Certainly, there has never been any serious controversy as to whether copyright is necessary or intrinsically a "good thing." And what copyright is is simple enough - the word itself says it: the right to make copies of something inherently reproducible by virtue of one’s ownership interest in it. How any of those simple concepts actually work is where copyright immediately passes from the linear to the fractal. And the latter, in short, is what we mean by the “nature and scope” of copyright.

The power of the federal government to protect copyright (as well as patents) is established in Article I, Section 8, Clause 8 of the Constitution, which reads in its relevant entirety:

The Congress shall have power . . . [t]o 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[.]

This clause sparked no debate to speak of at the Constitutional Convention, and it gets barely a mention in The Federalist Papers. Britain’s Statute of Anne had already been protecting the rights of English authors to control the reproduction of their works since 1710. Accordingly, the enactment of the first American copyright law, the Act of 1790, preceded the ratification of the Bill of Rights by one year.

In the Act of 1790, the terms “author” and “writings” covered the subject matter of copyright more or less literally. “Authors” meant the literal writers of written and published “writings,” although the Act specifically excluded newspapers while including maps and charts. After 1803, “authors” also meant printmakers, etchers and engravers. After 1831, “authors” also meant authors of written musical compositions. The term “authors” now also includes, among others, authors of recorded musical compositions, architects, choreographers, painters, sculptors, filmmakers and writers of computer software programming.

Only in copyright law are all of these artists called “authors.” And only in copyright law are dances, films and sculptures called “writings.” The reason for these elastic definitions is part of the special genius of American constitutional law. All of American copyright law – absolutely all of it – is encompassed in Article I, Section 8, clause 8. Congress is afforded the power to secure to “authors” the “exclusive right” to their “writings,” for “limited times.” Those are the words and, to this day, everything copyright does has to fit within those words – literally.

This flexible definition of “writings” was embodied in statute before it was challenged in court, but challenged it has been – most notably in a Supreme Court decision from 1884 called Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1884). Napoleon Sarony was a well-known photographer of the time who had taken a well-known photograph of Oscar Wilde. The Burrow-Giles Company made a lithograph lifted directly from this photograph. There was no issue whatsoever as to whether the lithograph was a copy of the Sarony photograph, because it clearly was. Also, the Copyright Act of 1870 had extended copyright protection to photographs. So, what was the issue? The issue before the Court was whether the 1870 statute’s categorization of photographs as “writings” was constitutional under the copyright clause. Was photography by definition protectable – as a “writing” – under the copyright laws and within the copyright clause? The Court found this expansive application of the word “writing” to be wholly constitutional:

By writings in that clause is meant the literary productions of those authors, and congress very properly has declared these to include all forms of writing, printing, engravings, etchings, &c., by which the ideas in the mind of the author are given visible expression. …
…
We entertain no doubt that the constitution is broad enough to cover an act authorizing copyright of photographs, so far as they are representatives of original intellectual conceptions of the author.¹

The specific phrasing here is key. “Writings” is roughly defined as the ways in which “the ideas in the mind of the author are given visible expression.” And that formulation raises a couple of key questions. First, if an author’s ideas given visible expression are protectable under copyright, does that mean the author’s ideas themselves are protectable? The answer is no, and this may be the hardest fundamental concept to explain to someone dealing with copyright law for the first time: Copyright Does Not Protect Ideas. It can only protect an idea when it takes the form that the law recognizes as tangible property: the expression of that idea in a work of some kind. But precisely where the expression of an idea crosses the line into an expression protectable by copyright is often difficult to explain and often obscure.

In another formative case, Baker v. Selden, 101 U.S. 99 (1879), the author of a book on accounting that he had copyrighted, sued when another author used a similar accounting system in his own book, using similar graphics. In this case, the Court found that these bookkeeping graphics, notwithstanding that they were visible manifestations of the plaintiff’s original ideas about accounting, did not rise to the level of protectable expression. The book and graphics might convey the author’s ideas about accounting, but those ideas themselves were not protectable. As the Court stated:

The description of the art in a book, though entitled to the benefit of copyright, lays no foundation for an exclusive claim to the art itself. The object of the one is explanation; the object of the other is use. The former may be secured by copyright. The latter can only be secured, if it can be secured at all, by letters-patent.²

This formulation has long been codified in the operative statute. Section 102(a) of Title 17, regarding the “Subject Matter of Copyright,” states, in relevant part:

(a) Copyright protection subsists . . . in original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device. . . .

Further sub-section (b) states, à la Baker v. Selden:

(b) In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

You may note here that the current version of subsection (a) says “tangible medium of expression,” rather than “visible expression” as the Supreme Court phrased it in the Burrow-Giles case 130 odd years ago. This raises the second question implicit in that case’s holding: How could an 18th Century formulation of copyright law be made to extend to 21st Century formulations of expression?

One key development was actually a relatively backward-looking Supreme Court case, White-Smith Music Publishing Co. v Apollo Co., 209 U.S. 1 (1908), which involved player piano rolls. Player piano was a largely archaic technology even in 1908. Phonographs had been around for quite some time, although sound recordings proper were not protected under the copyright statutes until 1971. As mentioned previously, written musical composition had been protected since 1931. A musical score was clearly understood as a visual expression under the law, even though you obviously had to be able to read music for it to be intelligible. Player piano rolls were different, however. It was a long scroll with holes punched in it. You would mount it onto a mechanized pneumatic piano that blew streams of air through a row of 88 tiny holes. The punched paper would run from spool to spool with the sheet pressed against the holes. When a hole punched in the sheet crossed one of the pneumatic valves, the air would escape through the hole causing a corresponding piano key to strike. A sequence of punched holes in a line would make a chord, and so on. And it would sound like a disembodied pianist playing whatever song you bought the roll for.

The issue for the Court in the White-Smith case was whether music punched onto a player piano roll was sufficiently fixed expression such that it constituted an infringing copy of the original copyrighted musical composition. In this case, the Court said no. Even though you could easily hear a theoretically unauthorized “performance” of a song by slapping the roll for that song into the player piano and switching on the mechanism, the specific means of transmitting the work – the punched roll – was not “generally intelligible” in itself. No one could be expected to “read” the holes in the paper roll. Oliver Wendell Holmes dissented, expressing what would seem to us a more intuitive view:

[O]ne would expect that, if it was to be protected at all, that collocation would be protected according to what was its essence. One would expect the protection to be coextensive not only with the invention, which, though free to all, only one had the ability to achieve, but with the possibility of reproducing the result which gives to the invention its meaning and worth. A musical composition is a rational collocation of sounds apart from concepts, reduced to a tangible expression from which the collocation can be reproduced either with or without continuous human intervention. On principle anything that mechanically reproduces that collocation of sounds ought to be held a copy … .³

Ultimately, this was something that had to be corrected in the wording of the statute, as with the language in section 102(a) referencing “the aid of a machine or device.” The most obvious current applications of which are video (film is visually intelligible, while magnetic inscription on tape or disc is not) and, most importantly, computer software. The fact that software is today the province of copyright rather than patent – at least, for the present – would appear to turn Baker v. Selden on its head. The extent that it does not is attributable to case law requiring that the protectable content in computer programs be subject to a number of tests.

Most emblematic is the Second Circuit’s holding in Computer Assocs. Int’l Inc. v. Altai, 982 F.2d 693 (2d Cir. 1992). In order to determine what aspects of source code are protectable, the Court required a three-step process:

First, “abstraction” → You break down software into component parts.

Second, “filtration → You sift out non-protectable material or what is purely functional.

And, third, “comparison” → You compare the remainder with allegedly infringing program for substantial similarity of content.

Ostensibly, if you take out the functional aspects of the programming, you will leave something from which some kind of “expression” can be discerned.

So what about “functional” aspects? Can you copyright something that has a useful function? The answer is yes, to the extent that there are aspects of that something that are not functional and also protectable expression. In Mazer v. Stein, 347 U.S. 201 (1954), for example, the issue was whether the sculptural bases in table lamps were protectable. The answer was yes – and the pictures of the objects make it reasonably clear that one could take out the lamp portion of these pieces, leaving the sculptural elements intact. But what if the heads of the dancing figures themselves had lit up? That would be a closer call, but still easy to make – there’s nothing directly related to the function of a light in having it shaped like someone’s head. But under section 101, if a “pictorial, graphic, and sculptural work” has any intrinsic function, it has to meet the “separability” test whereby you can either physically separate the expressive features from the utilitarian aspects – as in the lamps in Mazer v. Stein – or conceptually separate them (as in lit-up heads).

Predictably, there are any number of other cases in which the distinction between what is copyrightable has and what is not has been similarly delineated – at least tentatively – but the handful I’ve discussed convey the flavor of the law’s development, and the ways in which it has resolved itself within the language of the Copyright Clause. But the real whammy in that clause is its built-in dialectic:

“The Congress shall have power . . . [t]o 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[.]”

If the objective is to “promote the progress of . . .useful arts,” the question is not protecting a right deemed to be preexisting and inalienable but actually creating one. The “exclusive right” references just how much protection is necessary to encourage the production and circulation of original works. If you write a novel and never show it to anybody, copyright is irrelevant to that work. If someone steals the manuscript and hides it from you, they have robbed you of your property, and that is a crime, but copyright is still irrelevant – unless they try to exploit the work by publishing it – making it public. Then they have robbed you and infringed your work. You have the exclusive right to make your own work public, or to authorize that being done. But “limited times” says that, by definition, there has to be a limit to that exclusive right – it’s not so exclusive, let alone inalienable, if it self-destructs. Where and when that happens – and why – is always in flux.

That, in short, is the nature and scope of copyright: a system as inherently chaotic – in its basic structure – as the law provides.

1 111 U.S. at 58.

2 101 U.S. at 105.

3 209 U.S. at 19-20.

The Second Circuit Defines “Interactive” Digital Radio in Arista Records v. Launch Media

(November 9, 2009)

The Second Circuit’s decision in Arista Records, LLC, et al. v. Launch Media Inc.,¹ came down in August 2009, eight years after the case below was commenced by most of the major American record companies existing at the time.² The case concerned a digital music service called LAUNCHcast, created and introduced in November 1999, by the defendant, Launch Media. LAUNCHcast no longer exists or operates as it did in 2001, the year the suit was brought. Yahoo! bought Launch Media that same year, and earlier in 2009, Yahoo! folded the LAUNCHcast service into a service that is administered by CBS Radio, comprising over 150 pre-programmed digital radio stations. Thus, LAUNCHcast no longer resembles the service at issue in the case.

However, before its 2009 change-over, LAUNCHcast was remarkably similar to Pandora.com, which is still very much in operation. Quite fortunately for Pandora, the latter was also never a party to the LAUNCHcast lawsuit, although the Second Circuit directly references Pandora in two footnotes.³ In fact, Pandora may be the one entity that now stands to benefit most directly from the substance of the LAUNCHcast decision. What the decision states, in brief, is that services like Pandora and LAUNCHcast (in its earlier freestanding form) are to be regarded as functionally equivalent to traditional broadcast radio stations with respect to the way their audiences may access them and, accordingly, such stations should only pay the same sort of statutory license fees for the use of music as traditional analogue radio broadcasters have long done, and need not pay individual licensing fees for the use of the sound recordings themselves.

The appeal of Pandora (and, ostensibly, LAUNCHcast in its earlier version) to its users is that it delivers a kind of streaming radio service that is programmed entirely through the user’s input without being entirely controlled by it. Pandora will take as much information as you can give it about the artists, composers or individual songs you are interested in hearing in order for the service to tailor its programming as close to your tastes as it can make it. It will play music that, by its computations, might be deemed similar or somehow related to the artists and songs you input. It will also play this music entirely for free, notwithstanding the occasional audio commercial and pop-up ad, and – even better – this may serve to introduce you to numerous other artists and songs with which you may not be familiar that have index-able aspects similar to those artists and songs you enter into the system. More to the point, Pandora is calculated to play music you are likely to want to buy, and the prominent links on the interface make it remarkably easy to go directly to Amazon or iTunes and download the precise recording you were just hearing, even as you’re hearing it. Unquestionably, this is a solid moneymaker for providers of paid downloads. In fact, the Second Circuit quotes Pandora CEO Joe Kennedy in one of the footnotes in the decision, claiming that “Pandora [is] ... among the top promotional partners of iTunes and Amazon.com” for music.⁴

The bottom line, however, is that Pandora does not play (nor did the original LAUNCHcast) exactly the song you want to hear when you want to hear it. Even if Pandora plays something you would want to own, you have no way of knowing ahead of time when it’s going to be played. As such, if you want your own copy to play at will, you can’t make one through a deliberate digital capture of a Pandora stream unless you essentially “leave the tape running” whenever you’re running any of your Pandora-programmed “stations.” Downloads end up being far more convenient for Pandora users than trying to get around them. So, what was the record industry’s problem with LAUNCHcast?

It’s questionable whether the industry at large really did have much of a problem with the service itself by the time Launch Media got its favorable jury verdict on April 27, 2007. By then, the industry’s far more obvious problem was its long-foreseen general implosion due to illegal downloads and file-sharing. In fact, out of all the plaintiffs in the original suit, only the BMG affiliated companies appealed the verdict.⁵ What they got for their money, thanks to the Second Circuit, was a fairly hard and clear definition of the term “interactive service” under section 114(j)(7), under the language inserted into the statute by the Digital Millennium Copyright Act of 1998 (or “DMCA”). To understand why that matters and why the record companies might have cared quite a bit more about it in 2001, when the case began, we need to go back to 1995.

Before 1995, section 106 did not accord holders of copyright in sound recordings any kind of right to restrict the performance of those works, in direct contrast with holders of copyright in musical compositions who had the full bundle of exclusive rights -- even including “public display.” Up to that point, an exclusive performance right for sound recordings didn’t make economic sense, because it would have meant collecting additional license fees from analogue broadcast radio stations, which already paid blanket license fees for compositions through ASCAP and BMI, and directly benefitted the record industry by giving the records airplay. By the mid-‘90s, however, although the technology facilitating rapid digital transmission and copying of audio files was not yet widespread, it was very much on the horizon. Accordingly, the record industry began lobbying very hard for an amendment to title 17 that would address this sort of technology without affecting the industry’s longstanding relationship with broadcast radio.

The result was the Digital Performance Right in Sound Recordings Act of 1995 (or “DPSR”), which first of all added subsection (6) to section 106, giving holders of copyrights in sound recordings, the exclusive right “…to perform the copyrighted work publicly by means of a digital audio transmission.” Now, by its terms, this language did not have any bearing on analogue radio broadcasts. But what about digital radio broadcasts? That was addressed under section 114 which stated that the only entities subject to compulsory licensing fees for digital audio transmissions were paid subscription services that were also “interactive.” Under the DPSR, an “interactive” service was defined as,

one that enables a member of the public to receive, on request, a transmission of a particular sound recording chosen by or on behalf of the recipient. The ability of individuals to request that particular sound recordings be performed for reception by the public at large does not make a service interactive.

The last sentence of this provision would appear by its terms to prohibit P2P file sharing without putting an onus on a radio station’s request line, for instance. Playing songs requested by listeners would not subject the broadcaster to a license fee to the record company, providing ostensibly that the requester would not know when exactly the station would play it, and wouldn’t be likely to record the broadcast until the song came on in lieu of buying a copy.

Under strict analysis, however, that language leaves open a number of other possibilities. Under the old Sony/“Betamax” decision, “time shifting” – or, in that case, recording a television broadcast in order to enjoy it later in roughly the same form – does not infringe copyright,⁶ and the same principle applies to recording a particular radio broadcast and listening to it at one’s leisure. Far more pressing for the record industry, then and now, is any kind of transmission that would enable the listener to make a copy of something the listener wants to own, obviating legitimate purchase of the music: where, for example, the listener knows the content of the transmission ahead of time and can record it or copy it with relative ease.

Even now, you can look up playlists for certain analogue radio music programs, stream the original broadcast digitally, and - if you so desire - make a digital recording of the portion of the broadcast that you’d like to keep, instead of buying it legally. In that sense, a listener’s ability to re-stream a prior broadcast in order to obtain a specific recording the listener sees on the playlist (or enjoyed on the original broadcast) is not directly proscribed by the language of the DPSR, nor would be any unforeseen technology that might take advantage of this loophole.

To foreclose the above situation, the industry continued to lobby Congress to amend the DPSR, and the result was the Digital Millennium Copyright Act of 1998 (or “DMCA”). Under the DMCA, the definition of interactive service under section 114(j) was significantly expanded, most relevantly, as follows:

(7) An “interactive service” is one that enables a member of the public to receive a transmission of a program specially created for the recipient, or on request, a transmission of a particular sound recording, whether or not as part of a program, which is selected by or on behalf of the recipient.

The phrase “transmission of a particular sound recording” “on request” covers downloads or file-sharing, much as the language in the DPSR does, but the key additions in the DMCA version are the surrounding phrases: “transmission of a program specially created for the recipient” and the transmission of a particular sound recording as “part of a program, which is selected by or on behalf of the recipient.”

This language would – and does – apply to streaming digital audio where the recipient knows the content ahead of time. Moreover, as the Third Circuit held in the 2003 Bonneville case, in support of an opinion to that effect by the Copyright Office issued in 2000, the DMCA’s exemptions for “non-subscription broadcast transmissions” do not necessarily apply to digital streaming of analogue AM/FM radio broadcasts, even if the same broadcaster is responsible for both transmissions.⁷

In the same year that the Copyright Office issued the opinion ratified by the Third Circuit, the Digital Media Association (or “DiMA”) petitioned the Copyright Office for an amendment to the administrative rules that would clarify what “interactive service” meant under the DMCA. The Copyright Office refused to do so, given the “rapidly changing business models emerging in today’s digital marketplace,” and ending with the following language:

What is not clear, however, is how much influence a consumer can have on the programming offered by a transmitting entity before that activity must be characterized as interactive. . . . Such a determination must be made on a case-by-case basis after the development of a full evidentiary record in accordance with the standards and precepts already set forth in the statute.⁸

The Copyright Office also, very unhelpfully, issued two versions of the following footnote:

RIAA and DiMA discussed the services offered by Launch Media, Inc., through its LAUNCHcast service, and MTV, through its Radio SonicNet service, to illustrate the type of offerings that are in dispute. . . . From these descriptions, there is considerable doubt whether either offering would qualify as an “interactive service.”⁹

As the Arista court later pointed out, there was another draft of this opinion where the Copyright Office also said the exact opposite. Predictably, the Court disregarded both.¹⁰

The Copyright Office issued its non-report in November 2000, after LAUNCHcast had been in operation for a solid year. The record companies filed their complaint in May 2001, claiming that LAUNCHcast was liable for copyright infringement from the time of its launch in November 1999 because it was interactive on its face under the DMCA language. Its users essentially created LAUNCHcast’s programming through their specific input, thus seeming to place the service well within the ambit of the statutory language, in that it “enables a member of the public to receive a transmission of a program specially created for the recipient.”

Given the Copyright Office’s refusal to formulate any kind of bright line rule on the “interactive service” definition – or even to give an opinion, for that matter – the progress of Arista Records v. Launch Media through the Southern District and the Second Circuit was appropriately eccentric. There was never a lower court decision on the issue of interactivity. Instead, the trial court jury found as an issue of fact that LAUNCHcast was not an interactive service under the DMCA language. On appeal, the record companies remaining in the case argued not only that this finding was in error, but also that Judge Owen had erroneously instructed the jury to reach a verdict on that issue at all, insofar as the question was properly one of law.¹¹ Curiously, the Court of Appeals agreed with the appellants that the issue was a matter of law, and then proceeded to affirm the lower court’s jury verdict in favor of Launch Media.¹²

Just as intriguing, much of the Second Circuit decision strongly resembles a finding of fact, rather than one of law. Despite the Court’s note at the outset that “[t]here is no material dispute between the parties with regard to how LAUNCHcast works,”¹³ the decision proceeds through a long and tedious summary of the data-sorting algorithms and other functions through which LAUNCHcast assessed its users’ input to generate appropriate songlists while simultaneously mediating and limiting its customers’ control over the music they streamed over the LAUNCHcast service. In truth, the technical details provide a hedge for the Court’s simple legal formulation: a service is “interactive” if the user’s control over the programming “approximates the predictability the music listener seeks when purchasing music.”¹⁴ As the Court states:

The language and development of the DPSR and DMCA make clear that Congress enacted both statutes to create a narrow copyright in the performance of digital audio transmissions to protect sound recording copyright holders – principally recording companies – from the diminution in record sales.¹⁵

Notwithstanding that this is as close to a bright line rule as a Court could come, the factual analysis of LAUNCHcast’s methodology also serves to restrict the rule to the facts – or, more precisely, to existing business models.

The deliberate narrowness of the Second Circuit’s decision in Arista mirrors the record companies’ likely motivations in bringing suit to begin with and – even in the face of LAUNCHcast and Pandora’s by then self-evident usefulness to the industry – in bothering to appeal the jury verdict. The DMCA provision was drafted broadly to keep it ahead of as-yet-unformulated business models that might escape the DPSR’s definition. The more restrictive the operative interpretation of the DMCA definition could be made to be, the less likely that a revenue stream based on a new model would escape direct licensing fees for digital audio transmissions of sound recordings, even if the definition consequently encompassed services that posed no threat to the industry. It is telling that the record companies deemed it necessary to alienate significant sources of revenue in the interests of securing an overbroad statutory scheme of very uncertain future usefulness. The Second Circuit’s Arista decision creditably curtails this self-defeating attempt by the record companies to make the definition unnecessarily expansive while preserving the use of the definition as against some yet unknown digital music service that directly threatens legitimate music purchase. For the record companies, however, it is now very late in the day for the kind of prophylaxis that characterized this action at its inception.

1 Arista Records, LLC, et al. v. Launch Media, Inc., Index No. 07-2576-cv, ___ F.3d ___ (2d Cir. Aug. 11, 2009) (hereinafter , “Arista”).

2 Arista Records, LLC, et al. v. Launch Media, Inc., Index No. 01-cv-4450 (RO) (S.D.N.Y.).

3 See Arista, slip op. at 33 n.19 & 41 n.24.

4 Arista, slip op. at 41 n.24.

5 The ten original plaintiffs at trial were: Arista Records, Inc.; Bad Boy Records; BMG Music d/b/a The RCA Records Label; Capitol Records, Inc.; Virgin Records America, Inc.; Sony Music Entertainment, Inc.; UMG Recordings, Inc.; Interscope Records; Motown Record Company, L.P.; and Zomba Recording Corporation. On appeal, six dropped out, leaving only Arista, Bad Boy, BMG and Zomba.

6 Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 442 (1984).

7 See Bonneville Int’l Corp. v. Peters, 347 F.3d 485, 500 (3d Cir. 2003) (“Section 114(d)(1)(A)’s nonsubscription broadcast transmission’s exemption implicates only over-the-air radio broadcast transmissions, and does not cover the internet streaming of AM/FM broadcast signals.”).

8 Public Performance of Sound Recordings: Definition of a Service, 65 FR 77330-01, 77332 (2000).

9 Id. n.1.

10 Arista, slip op. at 21-22 & n.6 (“Whatever the etiology of the Copyright Office’s inability to make up its mind, we find the Copyright Office efforts here of little help.”).

11 See Arista, slip op. at 4.

12 See id. at 8.

13 See id. at 6.

14 Id. at 34.

15 Id. at 33.