Zvi S. Rosen, The Lost (and Found) Copyright Records
Before July 1870, copyright registrations were at each federal district court. Form of oath dictated by statute; also deposited title page. Deposit records; assignment records; indexes; misc. other records. Process: register by signing oath at local district court as author/proprietor; have to deposit in DC. Until 1831, had to provide notice by publishing in local paper for 4 consecutive weeks. 1802: added notice on works. Almost no one did the newspaper publishing, as far as he can tell. Most records held at rare book room of Library of Congress. Many missing records.
B/c title page deposit was completed before publication, it was often the only part to survive if the book was never published. About 30% of the deposits are these ghost books. For some jurisdictions, only the title pages and not the other records survive. DC: Good records of deposits exist for much of the period, but most books weren’t deposited, so it’s a spotty record.
List of non-standard registrations up to 1861, being product labels etc.—handwritten. Many records were assumed lost, but ended up in regional divisions of National Archives. Digitized all and will host at GW law.
Goals: locate old records; crowdsource local transcription; combine transcribed data to get a continuous record of registration for 1790 to now.
RT: What did they think they were doing with records of registrations? Was there a larger plan? Especially interested in that non-standard registration list—were there associated proposed reforms. But overall: Compare to property records, contract doctrine about reading terms: what is the importance of requiring notice when the required steps will generally not serve to provide actual notice? Symbolic notice: a statement that I know what I’m doing? Rare chance that someone will actually search the records before acting? W/real property title, you’ve incentivized the creation of intermediaries that actually perform the necessary checks—doesn’t seem to have happened in © for many uses.
A: notice was always half-baked. They knew newspaper notice wasn’t working and kept it in the statute for 40 years.
Q: got rid of newspaper notice in 1831 for new registrations, but kept it for renewals.
A: technically kept until 1909. Not much legislative history about 1831, but interesting to know why they kept it.
Silbey: Institutional memory in © registers—whether the people who worked there worked there for a really long time, as civil servants tend to do.
A: it was the clerks of district courts, which was a patronage job, about ten years. Sometimes a deputy clerk if there were multiple courthouses. The administrative head of the court, but that might be the only person who worked there, maybe not even a fulltime job. NY had a whole infrastructure, including a dedicated © person, with printed forms. Delaware had 50 © registrations for 70 years—they just handwrote them.
Very little © litigation before 1830s.
Will Slauter, Copyright Law and Registration Practice: The Case of News, 1870-1918
Example: Wanted registration as a book, not a print, because they wanted to sue for infringement even though it was a weekly. Political and cultural resistance to the very idea of owning news—18th century to today. Registration records are used more by literary historians than lawyers—filled in gaps where no copy survives; teach us about the use of © by particular authors; trends in publishing.
Registration had to precede publication, which was hard for news. Also, what do you register—a dispatch? These registrations, if done, were of questionable validity for a long time. 1829: Clayton v. Stone, no copyright for price current (updates on market news)—suggested no © for newspapers, but not clear. In 1884, effort to legislate 8 hours © automatic; defeated. 1886: Harpers Weekly sued; newspaper could qualify as a book under the statute, but considered an illustration so was less useful for written news. 1900: Tribune v. AP: can register, but lots of stuff in a newspaper isn’t covered anyway. 1909 Act: newspaper copyrightable, but only the copyrightable matter—not the news.
1850s: seeing more registrations of individual articles, and whole Harper’s Weekly every week. Serial fiction; history; politics; biography—but what about news? 1830s, attempt to legislate special © for news in Britain; not a response to tech but to political economy, since up until that time newspapers had a stamp duty that artificially kept prices high. Entrenched papers invoked fear of flood of pirate papers. 1880s in US: rise of cooperative press associations, news © proposed in US as national news organizations began to integrate. AP, and United Press—thinking about a direct rival, AP leadership pushed for special © and didn’t get it. That didn’t stop them from registering news articles as books. Colorado News & Press Ass’n found its news copied by a news thief, a slick fellow who repacked their news for his own customers. They threatened to sue, but he said news couldn’t be ©.
INS v. AP: Citing Nat’l Tel. v. Western Union, AP argued news couldn’t be copyrighted. But what hasn’t been noticed before is that the AP had used © in the past, such as Spanish-American war, and did so again in months leading up to INS. Registered 16 selected news stories—seems like effort to set up test case. You might say this doesn’t matter. Didn’t come up in the proceedings, even though much of the debate was about whether news, once printed w/o ©, was abandoned to public. If anyone had noticed that AP was registering and arguing that © was impossible, that would have seemed hypocritical, but the majority didn’t even accept the argument that © was impossible—it was possible for the literary form, not the underlying facts, which is what the AP wanted.
Why do these records matter? Complement to history of legislation and case law—how historical actors sought to use © and to what effects.
Silbey: administrative agencies often drive debates or reify lines drawn from inside out; this is not a standard story of bureaucracy.
A: there’s probably a change after 1898, when the new register does want to impose a bit of order on the chaos.
Zvi S. Rosen & Saurabh Vishnubhakat, An Empirical Study of U.S. Copyright Registration and Renewals, 1870-1977
Until 1897 there were statistics in annual reports; then they moved to a new building. 1891: Catalog of Copyright Entries. That included statistical info from 1903; 1909 started listing renewals separately. 1947: richer statistical information. Computerized since early 1970s; form basis of public database of all copyright registrations and renewals filed since 1978.
Authors almost always registered around the time of publication. This can be checked by comparing registration date to renewal date. Over 95% registered b/t 27 and 29 years before renewal was filed. In 1992, dropped but only to above 90%, and then in the last year a bunch of people tried to register and renew at the same time. When renewal became optional, absolute #s of renewals dropped.
Are the annual reports statistically effective for comparing registrations to renewals? All based on the fiscal year, not the calendar year, for registrations, even though all the renewal data is per calendar year.
Large volume categories—books, periodicals, music, artwork. Artwork peaks 1909, and only got back to that level by the 1970s. Books peaked in 20s, dipped, then steadily up. Music and periodicals: steadily up. Small volume: dramatic works and lectures go up, then dip a lot. Photographs: people registered motion pictures as photos until 1920s, but still huge drop from 1000s to under 100 in the 50s. Panics in economy create small dips.
Renewals: books go steadily up; motion pictures small but still show up. Total renewal rate by registration year goes somewhat steadily up until renewal becomes optional, which is a cliff. Music: renewal rates 35-40%; motion pictures very high—the only one over 50%. Books peak for original registrations in the 40s. Periodicals go up and then plateau when the renewal term becomes 47 years. Artwork: definite jump that steadies when renewal term becomes 47 years. Music is volatile, but high for works created in 1910s and 20s; goes back up for works in 40s and 50s.
Renewal rate for 28+28 term, 12%; extending 15%; to +47 up to 18%; then downwards when optional, even when renewal term increased to +67.
Can do event studies to see if new Register or legal changes map onto anything.
D.R. Jones, Edicts of Government: Copyright in State Legal Materials
Compendium: policy, won’t register a gov’t edict issued by state in other jurisdiction, including legislative enactments, judicial decisions, administrative rulings. States are nonetheless making claims of copyright. Some discussion in runup to 1976 Act; decision not to put specific coverage in though some people claim that’s b/c it’s covered already.
Open access, right to access law as part of access to justice. But what does access mean? Is searchability required? Access to justice literature talks mainly about forms, but the law itself is also important. State has duty; doesn’t need incentive to publish. But states induce publishers to publish by claiming ©/exclusive copyright. This is the law: a factual issue. The concern is one for authentication.