Fifty-Five Years of Software Copyright
Greetings from INTA 2019 in Boston! It’s been a few months since a post, entirely my own fault of course, but today is a good day to bring readers something new – the Copyright Office announced 55 years ago today (May 19, 1964) that they would begin accepting registrations of software copyrights. This was of course an enormously consequential decision, and while it has been written about before, as far as I know a proper history of the era has yet to be written (Pam Samuelson’s article The Uneasy Case for Software Copyrights Revisited has a lot of useful material on the subject though). As such, I thought I’d provide a few aspects of the early story of software copyrights that some may not be aware of.
Added 4/9/2021 – William Patry notes that there were actually 3 earlier registrations – a FORTRAN program entitled “Gaze-2, A One-Dimensional, Multigroup, Neutron Diffusion Theory Code for the IBM-7090,” written by the General Atomic Division of General Dynamics, registered on February 1, 1963 (registration number A607663). In early 1964, General Dynamics registered two other computer programs (A677198 and A678350).
The best source for the prehistory of this decision was the article Assistant Register of Copyrights George Cary (who briefly served as Register 1971-1973) published in what is now the Journal of the Copyright Society of the USA, entitled Copyright Registration and Computer Programs . He records that the first known attempt to register a computer program for copyright had been in 1961, when North American Aviation filed a registration for a program written for the Autonetics Recomp II on magnetic tape. That program was what began the investigation of software and copyright by the Copyright Office, but the program was complex and Cary indicates that North American was completing “certain required procedures” before it could be registered. It was however registered in June of 1964 following the institution of rules for registering computer programs.
In April of 1964, while the earlier application was pending John Banzhaf, then a law student at Columbia, filed two short programs with the Copyright Office – one on magnetic tape, and one which had been published in a Columbia student newspaper. The latter especially, as a contribution to a publication which would otherwise be copyrightable , spurred the Copyright Office to act, and the Copyright Office issued this on May 19, 1964, the ancestor of the modern Circular 61:
After extensive consideration the Copyright Office has taken the position that copyright registration for computer programs is possible under the present law.
Copyright Registration for Computer Programs – Announcement from the Copyright Office, reprinted at 11 Bull. Copyright Soc’y U.S.A. 361 (1964)
Speaking very generally, a computer program is either a set of operating instructions for a computer or a compilation of reference information to be drawn upon by the computer in solving problems. In most cases the preparation of both of these types of programs involves substantial elements of gathering, choosing, rejecting, editing, and arranging material. Some types of programs also embody verbal material which is written by the programmer and could be considered literary expression.
The registrability of computer programs involves two basic questions:
(1) Whether a program as such is the “writing of an author” and thus copyrightable, and (2) whether a reproduction of the program in a form actually used to operate or be “read” by a machine is a “copy” that can be accepted for copyright registration.
Both of these are doubtful questions. However, in accordance with its policy of resolving doubtful issues in favor of registration wherever possible, the Copyright Office will consider registration for a computer program as a “book” in Class A if:
(1) The elements of assembling, selecting, arranging, editing, and literary expression that went into the compilation of the program are sufficient to constitute original authorship.
(2) The program has been published, with the required copyright notice; that is, “copies” (i.e., reproductions of the program in a form perceptible or capable of being made perceptible to the human eye) bearing the notice have been distributed or made available to the public.
(3) The copies deposited for registration consist of or include reproductions in a language intelligible to human beings. If the only publication was in a form that cannot be perceived visually or read, something more (e.g., a print-out of the entire program) would also have to be deposited.
The detailed practices of the Copyright Office in this area will have to be evolved over a period of time, on the basis of experience.
The language about why software registrations was allowed is interesting, and matches what Cary said in his article:
There thus seemed to exist a reasonable doubt whether the principles of Baker [v. Selden] and White-Smith [v. Apollo] were properly applicable to a computer program in the form of punched cards or tape.
George D. Cary, Copyright Registration and Computer Programs, 11 Bull. Copyright Soc’y U.S.A. 362, 365 (1964)
This doubt was such that the Office considered it justifiable to resolve the doubt in favor of registration, in keeping with its policy of giving the applicant the benefit of the doubt wherever possible.
The citation to White-Smith v. Apollo (a 1908 case holding that sound recordings were not properly the subject of copyright on statutory grounds because they were not readable by humans) is interesting – the Copyright Office was still rejecting attempts to register sound recordings for copyright on those grounds and would continue to do so for another eight years. Regardless of this, shortly thereafter all three programs were registered for copyright. And more software programs came in the years thereafter, although it was hardly an avalanche, as the statistics in the annual reports of the Register of Copyright reveal.
In the late 1960s there began to be stirrings of what would become the internet you’re all reading this on, as well as questions of how copyright would handle such a medium. The internet was ultimately based on Arpanet, a Defense Department project, but at the same time there was active conversations in a coalition called EDUCOM, which planned to built a network based on educational values and needs called EDUNET. The back issues of EDUCOM’s Bulletin have some fascinating conversation about the future of copyright on the nascent internet, and the issue of April 1967 is especially interesting, with the entire issue dedicated to copyright in software, and how the proposed EDUNET would interface with copyright law. I’ve uploaded it here, and it’s worth looking at.
Some proponents of software copyright argued that EDUCOM was calling for a moratorium on software copyrights, and characterizing EDUCOM and its allies as calling for a moratorium on software copyrights until what would become the CONTU commission could be convened. CONTU was not convened until 1976 (folks might remember that I shared the transcripts here), and opponents of a moratorium successfully argued that a moratorium would effectively kill software copyright and the software industry.
Here we are, 55 years later, and the software sector has prospered with copyright protection, and seems to have found norms that work (ranging from open-source to commercial distribution). However, we’re still registering computer programs copyright as if they’re printouts of source code, as was required at the time.
There is one other article on the subject of software copyrights worth noting (actually quite a few from the era, but this one most of all) – future Supreme Court Justice Stephen Breyer’s 1970 article The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 Harv. L. Rev. 281. Breyer was surveying much of this as the background of his article and it’s worthwhile to be aware of when considering his arguments. Breyer’s piece was and remains somewhat controversial, siding as it does firmly on the “minimalist” side of copyright policy, but it’s an important piece to be aware of. In reading Breyer it’s also worth re-reading Barbara Ringer’s 1974 speech The Demonology of Copyright, which is in part a response to Breyer.
I’ve included one of John Banzhaf’s copyright application’s below – it’s fascinating to see the origins of what’s become a multibillion dollar industry.
Pingback : Mostly IP History » Registering Generative AI Works for Copyright Under the Rule of Doubt