27
Apr
2018

Player Pianos and the Origins of Compulsory Licensing – Some Details of its Origins

/
/
0
/
27 Apr 2018

The Music Modernization Act just passed the House as I’m writing this, and it seemed apropos to look at the origins of mechanical licensing in the 1909 Copyright Act.  The story has been told before (although not in a dedicated article or book), but I’ve found a number of aspects of the story that I believe have been largely forgotten, along with a few documents that I’ve scanned which are pretty cool to see.  Accordingly I’ve compressed parts of the narrative and expanded the parts where I have something new to add.  One of these days I’d like to write a book on the 1909 Act, or at least put this stuff into an article

The first sound recordings that could be played back were made in the 1870s, and the technology was commercialized effectively beginning in the 1890s.  In these early days the legal aspects of the various early sound reproduction systems were generally lumped in with another technology that developed at around the same time, the player piano,1 and in 1888 Harry Kennedy brought suit in the U.S. Circuit Court in Massachusetts, arguing that the inventor John McTammany had prepared Kennedy’s song Cradle’s Empty, Baby’s Gone, copyrighted in 1880, on a piano roll for use in a player piano (the song is a serious downer, fair warning).  The casefile from the National Archives  (which I originally shared last year) shows that in 1882 the Automatic Music Paper Company paid Kennedy a license fee, and published an authorized piano roll indicating that the song was used with permission.  In fact, although the reported decision does not make it clear, the Automatic Music Paper Company was Kennedy’s co-plaintiff in the case.2  However, in a brief decision the Court held that programming a paper reel with punches for use in a player piano did not infringe the copyright of the musical composition which was played back from that paper reel.  The matter was appealed to the U.S. Supreme Court, but in 1892 the Supreme Court dismissed the case without an opinion.  145 U.S. 643 (OT 1891 No. 168)

It would be another decade before another reported decision on mechanical reproduction, but at the beginning of the twentieth century one George Rosey was accused of selling sound recordings of two songs on wax cylinders – “Take Back Your Gold” and “Whisper Your Mother’s Name.”   I’ve scanned  the appellate case file of the DC Court of Appeals, and the briefs show the conflicting positions taken by the parties to the case, where Rosey based his argument on prior caselaw – the 1888 Kennedy decision and the then-recent English case of Boosey v.Whight, 1 Ch. 836 (1899) and 1 Ch. 122 (1900).  By contrast the music publishers of Joseph W. Stern and Edward B. Marks looked to broad principles of copyright to assert that recording a copyrighted song constituted infringement.  In 1901 the DC Court of Appeals (now the DC Circuit Court of Appeals) held similarly regarding sound recordings, finding that a sound recording did not infringe the copyright in a musical composition.

As I mentioned (I’m not sure others have noticed this), as early as 1882 the Automatic Music Paper Company was paying royalties to songwriters for mechanical reproduction, in spite of the lack of clear legal precedent saying they had to.  Later in that decade the Automatic Music Paper Company merged with the Mechanical Orguinette Company, and the resultant company would be named the Aeolian Company.  Aeolian would continue its predecessors practice of compensating songwriters for use of their songs.  Its motivation for starting this practice is unknown (at least to me – feel free to educate me in the comments) – was it alruism and a sense of duty, did they believe they had a legal obligation, or, as critics increasingly charged, did they intend to establish such a right once they already held licenses to most popular music, setting themselves up to dominate the player piano market using copyright?

When the White-Smith Music Publishing Company sued the Apollo Company to assert that Apollo had sold piano rolls that infringed their copyright in the musical compositions, it was generally understood that Aeolian was actually behind the litigation.3  The case slowly worked its way through the Courts, and while this was happening, in 1905, the gears began turning on a major revision to the copyright laws.  In June of 1906 Congress held hearings, and the transcript shows that Aeolian’s opponents showed up in force to argue that Aeolian was trying to create a monopoly in the player piano using its licenses to musical compositions.  Aeolian was not represented at these hearings, but at their behest a young Nathan Burkan published a pamphlet with the unwieldy title of The Charge That The Passage Of The Copyright Bill, Senate Bill 6330, Will Create A Monopoly In The Manufacture Of Automatic Musical Devices Is False.  The pamphlet on the one hand acknowledged that Aeolian had, in 1902, made contracts with many of the major music publishers for mechanical rights, assuming a court found such rights existed.  Aeolian also agreed to fund the litigation, and according to Burkan they had already expended $50,000 in legal fees, the equivalent of over $1.2 million today.  However, they argued that Aeolian had not secured contracts for all popular music, and further that such contracts explicitly did not include rights for sound recordings, which were likewise ascendant.  Nonetheless, the response does leave a certain lingering impression that Aeolian acknowledged that they would have at least a dominant position in the player piano market if a law giving the owners of copyrights in musical compositions an exclusive right of mechanical reproduction.

After the case arrived before the Supreme Court, but before argument, the defendants moved to dismiss the case, arguing that the amount in controversy was not significant enough for the Court to have jurisdiction.  This motion is found at the end of the case file held by the National Archives, and has not to my knowledge been previously discussed.  It’s not clear why Apollo would move to dismiss at this point if the amount in controversy was so minor anyway – they’d come so far, one would think they would want to be vindicated.  One possibility is that, if the case was a test case with a willing defendant, Aeolian grew concerned that the winds of the case were blowing against them, and tried to get out before the Supreme Court issued a decision.

Sure enough, in 1908 the Supreme Court affirmed the earlier decisions regarding mechanical reproduction in White-Smith v. Apollo, and held that the 1870 copyright law did not recognize mechanical reproductions as “copies.”  However, the Court made clear that Congress could step in and remedy this inequity, and in response the 1909 Copyright Act provided compulsory licensing provisions for “mechanical” reproductions of copyrighted musical compositions, including sound recordings.  It was widely understood that the compulsory nature of the licensing was meant to prevent another entity like Aeolian from dominating the market using exclusive licenses.

Roughly ten years later, the Federal Trade Comission investigated Aeolian for various antitrust claims, including price-fixing, leading to a cease and desist order (at pg. 124).  However, Aeolian survived far longer, and long after player pianos had left the mainstream, going through a series of mergers and finally declaring bankruptcy in 1985

  1. There were actually different technologies and names for mechanical organs at the time, but the details are only incidentally relevant and I’ve chosen to simply describe everything as player pianos and piano rolls. ↩
  2. The casefile is pretty interesting in general, and the answer filed by the defendant includes a detailed description of how the player piano worked. ↩
  3. The rolls were actually made by QRS, which is still in business today as the last maker of piano rolls.  This section is going to be compressed a bit, but you can read a longer version of the story of White-Smith v. Apollo at pp. 31-33 of my article Common-Law Copyright. ↩
25
Jan
2018

Opinions of the US Supreme Court – Does Word Count Correlate to Citations?

/
/
0
/
25 Jan 2018

As part of a symposium on forgotten cases in intellectual property with the Syracuse Law Review, I recently wrote a short history of the US Supreme Court’s 1879 decision in Perris v. Hexamer, entitled How Perris V. Hexamer Was Lost in the Shadow of Baker V. Selden.  Perris is essentially forgotten today, but it has somewhat similar facts and holding to Baker v. Selden, decided the following year.  The decision is less than 1,000 words, so it’s pretty easy to give it a read, but essentially the Court held that the legend/key to a fire map showing what different symbols meant was not protected by copyright, and that using the same colors and symbols as a competitor’s map did not constitute infringement either.  In some ways this decision is even more relevant than Baker (concerning the copyrightability of accounting ledgers) to the copyright questions raised in cases like Google v. Oracle, among others.  However, Baker keeps being cited (hundreds of times in the past few decades alone), while a citation in 2016 by the 9th Circuit was the first citation to Perris in 3+ decades.  I wanted to figure out why.  Given that Perris is a fairly short opinion and Baker is an average-length opinion, I figured perhaps just the length of the opinion led Perris to be ignored, as presumably other short opinions would be ignored.

This question naturally led me to try to answer a broader question than I actually needed to.  I assumed that there must be public databases out there of the number of words in an opinion, along with the number of citations to that opinion.  However, while there has been some scholarship on the question, no public database of this sort exists.1  Accordingly, with help from my law school classmate Corey Mathers, I decided to try to assemble it.  Paid databases like Westlaw and Lexis were not options, but the website Courtlistener.com (a project of the Free Law Project) does have the entire US Reports, along with citation tracking.  Accordingly, we decided to build our database from that site.

Accordingly, our data is here (zipped CSV), with data on every Supreme Court decision, including word count, number of citations to authority in the opinion, and number of citations to the opinion (as well as caption and year).  We removed cases that have less than 200 words, which are typically not real decisions but are rather summary orders like grants of certiorari.

However, the data has some weaknesses, and should really be considered more of a first step than a definitive resource.  By far the most significant weakness stems from the holdings of Courtlistener – while it has many recent decisions and all Supreme Court decisions, it is missing most caselaw from before 1950 or so.  Accordingly this creates a bias in favor of more recent cases, but at least the bias is uniform across all cases.  Ideally the data could be recreated from a database with more decisions.  Also, the word count for pre-1880 decisions frequently includes lengthy arguments of counsel as well as the actual decisions.  There are several other smaller issues as well, all of which could be resolved by re-running the query we ran on a database with full coverage of federal and state decisions.

Usually these posts have been a chunk of information, but this one is really more open-ended, since I know the data we created is deeply incomplete – it was acceptable for purposes of my paper but could be improved.  What would be good next steps for developing the data?  What other data should we be trying to generate (only data that can be done automatically, nothing that would require manual review beyond error-checking)?

  1. There is the Supreme Court Database, which has extensive information on cases, and is of special interest to political scientists and legal historians.  However, it does not contain this information. ↩
7
Dec
2017

Archival Adventures: The First Louisiana Copyright

/
/
1
/
07 Dec 2017

As folks to this blog likely noticed, I engaged in a project to bring together all pre-1870 copyright records, which culminated in a post earlier this year. As part of that project I was able to find a few records from Louisiana from 1837 and 1838. These were located in the National Archives Regional Division at Ft. Worth, but I wasn’t able to actually go there myself to see if there were any more. So when I had a chance to go to Ft. Worth to present at the IP Scholars Roundtable in October, I figured it would be a chance to double check the record books myself and see if there were more copyrights there.

Aside from the 1837-1838 records I found last year and a single registration held by the Library of Congress from 1835, there were known no copyright records from before 1851. This is a major gap in the records since New Orleans was a major city from the Louisiana Purchase onwards, and the copyright records would be a useful source for understanding the legal and cultural milieu of the time. In fact, aside from possibly Maryland, where the copyright records for the state are missing from before 1831, this is likely the largest gap in the pre-1870 copyright records.1

So, I went to the Archives, and decided to photograph the record books for the pre-1851 time period as well as looking through them (recall that the Library of Congress has copyright records for the federal court in New Orleans for 1851 on.  I figured I might find more than the student who looked at the books originally had found.  The National Archives catalog entry says that the series of record books is case papers, copyrights, and appointments, but in reality it’s mostly bonds and such documents related to pending litigation.  I’ve uploaded my scans the two volumes to The Internet Archive, and everyone is free to look themselves, but there are no copyright records for years other than 1837 and 1838.

However, I realized that the copyright records had to be somewhere, and I knew I wouldn’t have a chance to be in Ft. Worth again for a while.  Accordingly, I decided to look at the Minute Books of the Louisiana District Court on microfilm, which is where I found this:

1810 Copyright Registration for a City Directory of New Orleans from the District Court’s Minutes

It’s a little difficult to read, but this indicates that on October 11, 1810, Thomas H. Whitney (fixed, I originally thought it was White) registered a book for copyright entitled “New Orleans & Natchez Directory and Louisiana Almanac.”  I’ve also included a link to the introduction to the microfilm and a reproduction of the complete page containing this registration here.  The volume has a mark on the first page indicating that there is a copyright registration on page 257, along with a few other indications of highlights in the volume but no other mentions of copyrights, strongly suggesting that this was the first copyright in Louisiana.

I haven’t taken the time to go through the remainder of the minute books on microfilm, which would be extremely laborious, but this at least suggests that there may be more copyright records there, pointing to a resolution for where at least some of the missing records are.  It’s also instructive for the “lost” records from the early Republic, suggesting that just because we don’t know where they are, it doesn’t mean they’re necessarily lost forever.

  1. Another contender for largest gap is Virginia, but even though the original records from Richmond are missing until the 1860s, a transcription exists for 1790-1844, and the “second set” of records held in the Library of Congress should cover much the rest of the gap until the Civil War.  The question of how that early transcription is made is an interesting one, but it might be too esoteric even for a blog post unless I can bundle it with a discovery or two. ↩
26
Oct
2017

Another Season, Another Common-Law Copyright Opinion

/
/
0
/
26 Oct 2017

To be clear, this is my opinion only, and the views expressed in this post, and indeed this blog, should not be imputed or otherwise associated with anyone else.

This one from the Supreme Court of Florida, finding that Florida common law does does not recognize an exclusive right of public performers for the holders of common-law copyrights in sound recordings made before February 15, 1972.  The 11th Circuit certified a series of questions to the Florida Supreme Court, namely:

  1. Whether Florida recognizes common law copyright in sound recordings and, if so, whether that copyright includes the exclusive right of reproduction and/or the exclusive right of public performance?
  2. To the extent that Florida recognizes common law copyright in sound recordings, whether the sale and distribution of phonorecords to the public or the public performance thereof constitutes a publication for the purpose of divesting the common law copyright protections in sound recordings embedded in the phonorecord and, if so whether the divestment terminates either or both of the exclusive right of public performance and the exclusive right of reproduction?
  3. To the extent that Florida recognizes a common law copyright including a right of exclusive reproduction in sound recordings, whether Sirius’s back-up or buffer copies infringe Flo & Eddie’s common law copyright exclusive right of reproduction?
  4. To the extent that Florida does not recognize a common law copyright in sound recordings, or to the extent that such a copyright was terminated by publication, whether Flo & Eddie nevertheless has a cause of action for common law unfair competition / misappropriation, common law conversion, or statutory civil theft under FLA. STAT. § 772.11 and FLA. STAT. § 812.014?

Instead of addressing these questions, the Court chose to address a reformulated question of its own, “Does Florida common law recognize the exclusive right of public performance in pre-1972 sound recordings?”  The obvious problem with this is that it fails to address whether pre-72 sound recordings are protected under Florida law more generally.  The Court notes (pp. 19-20) that Florida criminal law provides penalties against commercial bootleggers of sound recordings, but those criminal provisions do not impact a range of activity including noncommercial infringement.

This could be excused as judicial minimalism if it wasn’t central to the case – Flo & Eddie sued in Florida specifically because SiriusXM has servers there, and alleged that copying was ongoing on those servers in violation of their exclusive right of reproduction.  The Court later in the opinion reasoned that because the criminal statute exempts copying made in the course of “as part of a radio, television or cable broadcast transmission,” so the copying would not trigger civil liability under common-law copyright, regardless of whether any such rights exist.  But of course this analysis doesn’t properly follow – whether certain actions do not give rise to criminal liability does not mean the same actions give rise to civil liability.  One only need to look at federal copyright law (or, indeed, nearly any body of law) for ample demonstration of this – most conduct which gives rise to a civil claim for infringement does not give rise to a criminal claim for the same conduct.

So, in revising the question formulating by the Circuit Court to one it preferred, the Florida Supreme Court effectively ignored one of the main claims being brought by Flo & Eddie.  However, I believe the Court’s reasoning on the performance rights claim is problematic as well.  One key part of the Court’s analysis is that finding performance rights would upset settled expectations and cause wide-ranging impacts not expected by the statute.  Regardless of whether this is true, 1 the Court relies on a second argument as well.  The Court notes that Congress extended federal copyright protection to sound recordings in 1972, but in the same breath denied them performance rights.  The Court thus reasons that Congress could not have intended to take away rights that existed, and thus performance rights must not have existed before that time as well.

The problem with this analysis is that Congress clearly was taking away performance rights, at least for sound recordings from Pennsylvania, where a performance right had existed at common law since the decision of that state’s Supreme Court in the 1937 Waring v. WDAS decision.2  And by the evidence, this was a tradeoff sound recording rightsholders were fine with at the time – the right to federal remedies against bootleggers of recorded music was worth far more than a right in some states to demand royalties for sound recording performance.  This is especially true because record companies did not want to imperial radio airplays by demanding additional royalties at the time.

Regardless, Florida is unlikely to revisit this decision anytime soon.  California is the next stop, where the same issue is now being briefed before that state’s Supreme Court.

  1. I’m skeptical that the Flo & Eddie cases would actually have such an effect since they’re only aimed at services that already pay royalties for sound recordings, and indeed Flo & Eddie and SiriusXM reached a settlement that would resolve the issue, but that’s another story.  By contrast, I do think the ABS series of cases, targeting terrestrial radio stations, may raise broader concerns, regardless of the legal merits. ↩
  2. Whether New York recognized such a right was unclear in 1972 – the 1940 RCA v. Whiteman decision had said no, but in 1955 the same Court (the 2nd Circuit) found that decision to be overruled.  Three states including Florida had expressly abrogated the performance right by statute, and as far as I’m aware 45 states had no precedential decisions or laws on the topic. ↩
24
Oct
2017

Blogiversary, Mark 1: Looking Forward

/
/
0
/
24 Oct 2017

It’s hard to believe that today marks the one-year anniversary of when I first posted to this blog in earnest (I started it a few months earlier but hadn’t posted anything beyond a “hello world” post). In the process I’ve been pleasantly surprised at how many people are interested in what is surely one of the more esoteric blogs about intellectual property law out there.

I have an awful lot in the backlog of stuff to share, including a lot of tidbits about sound recording copyright and the White-Smith decision, as well as a post I really should get out soon on Copyright and the 11th Amendment/Sovereign Immunity.  However, the biggest project I’m working on is an empirical project, bringing together statistics of copyright in America from 1790 through 2015.  Here’s a teaser of what I’m working on (with my coauthor Richard Schwinn, an economist):

All Registrations per capita

This chart shows the number of copyright registrations made, per 100,000 people, per year, with the color of the line representing the cost of registration adjusted for inflation.  This is only the tip of the iceberg – the statistics I’ve assembled are broken down for the type of work for 1870 to the present, and I have rates of renewal for 1909-2005 as well for all classes.  A paper that hopefully lays it out all is in the works.  Here’s to another great blog-year.

10
Oct
2017

The Forgotten Origins of Copyright for Photographs

/
/
0
/
10 Oct 2017

It’s fairly well-known that photographers like Matthew Brady used photographs in unique and important ways during the Civil War to document the conflict like never before.  It’s also known among copyright nerds that 1865 saw not just the end of the Civil War, but the amendment of the federal copyright statute to include photographs.  However, the conventional narrative of this law has always been that the amendment to the law to include photographs was close to a bolt from the blue.  As William Patry puts it:

This Act had a remarkably short legislative history. On February 22, 1865, the Committee on Patents and the Patent Office, which had been studying the issue, reported S.468, which was passed by the Senate the same day. The House passed the bill on March 2, and President Lincoln signed it into law the next day.

Source (internal citations/quotations omitted).  The law is one of Abraham Lincoln’s two main accomplishments on copyright, the other being his appointment of Ainsworth Spofford to be the Librarian of Congress in the same year.  However, the legislative history of the law was longer than has currently been understood, as a bill to include photographs within copyright law had actually been introduced in the House the previous year by Thomas A. Jenckes and committed to that chamber’s Committee on Patents.  However, for reasons that are unclear the House did not order the Bill printed, and as a result it has been all but forgotten until I found a manuscript copy of the bill in the Congressional files at the National Archives.

Finding the bill was a bit of a fluke – when I was writing my article on the origin of performance rights for music in 1897, I went through Thorvald Solberg’s work Copyright in Congress, 1789-1904, in search of any previous bills to provide such performance rights. What I found is that for a number of bills, he describes copyright bills, but provides no details as to the content of the bill.  The Library of Congress’s American Memory – A Century of Lawmaking site does not have a copy of the bill (House Bill 505 from the 38th Congress), so I (perhaps excessively) checked the files of the Committee at the National Archives.  Sure enough, there’s a handwritten copy there.  My scan of the bill is here, and I’ve included a transcription below the jump.

The act that would be passed in 1865 to include photographs in copyright is extremely terse, stating that  the provisions of the copyright law “shall extend to and include photographs and the negatives thereof which shall hereafter be made, and shall enure [sic] to the benefit of the authors of the same in the same manner, and to the same extent, and upon the same conditions as to the authors of prints and engravings.”1  On the other hand, the bill introduced by Jenckes in 1864 created a whole mechanism for deposit of a “memorandum” describing the photograph with the clerk of the District Court (since copyright registration was still at the District Courts until 1870).  Also included, seemingly added later, were two final sections establishing limited trademark protection for the marks of photographers (six years before the first law providing for federal trademark protection).

I don’t currently know the connection between the 1864 Jenckes Bill in the House and the bill a year later in the Senate which became law, but the introduction of the Jenckes Bill gives an explanation of how Congress moved so quickly on the issues – even if the public record doesn’t make it clear, the House Committee on Patents had been considering a bill to include photographs within copyright since the previous session of Congress.  There’s probably the necessary information to link these bills in the Papers of Thomas A. Jenckes; one of these days I hope to be able to tell the whole story.  But in the interim, this seemed a nugget of information worth sharing.  The bill text follows below the jump.

Read More

  1. the 1865 Act then includes several paragraphs reestablishing the requirement of copyright deposit with the Library of Congress, a requirement which has been retained ever since ↩
14
Sep
2017

Mini-Post: 1947 Copyright Office Report on Editing Music

/
/
0
/
14 Sep 2017

It’s been too long since I made a proper post – I’ve been busy putting together a piece I’m calling An Empirical Study of 225 Years of American Copyright Registrations (title subject to change, natch), which I fully expect to be the bees knees, all while being no bigger than a breadbox. But there’s still a lot I want to share, and this seems as good a time as any to share a report the US Copyright Office issued in 1947 which has been almost completely forgotten.

Until 1947, the US Copyright Office would not register editions of public-domain music for copyright.  Obviously the music could not be reinserted to copyright protection, but the Office held that merely editing public-domain music was insufficient for copyright protection.  However, in that year the Office issued a report (link to my scans) that gave the history of Office rules for registration of music, and an argument for how they should be changed to include edited versions of public-domain works.  The Office changed its position based on this report and allowed such editions of music to be registered for copyright.  The Office had distributed the report to interested outside parties, and copies are available at a number of law libraries, but I haven’t seen it online.

The issue of whether editions of public domain music are nonetheless protected was litigated in England in the Hyperion Records case, involving a small classical music label in a suit by a scholar over his editions of the music of an early 18th century composer.  However, I think the discussions in the report are illuminating for discussing the scope of music copyright protection and infringement more generally, something Joe Fischman also discusses in his upcoming article Music as a Matter of Law, and I of course discuss in my upcoming article Common-Law Copyright (never miss an opportunity to plug one’s own work).

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

My Twitter Feed