Discussion of Definitions

Commercial recording

A single recording of a selection or selections by an artist, issued for sale in the United States to the general public during the period specified and generally meeting the definitions in Title 17, U.S. Code, sections 101–104. For popular music, this generally means a single “track,” whether it originally appeared on a cylinder; on one side of a 78- or 45-rpm disc; on a multitrack LP, extended play (EP), or tape; or in some combination of these or other fixed formats. Multipart recordings of a single extended selection (e.g., a symphony or an opera) are considered to be one recording. Alternate takes of the same selection, made at or about the same time, and rerecordings made solely for duplication purposes, as was common in the early years of the industry before mass-duplication technology became available, are not considered separate recordings.1 However, recordings of different lengths made for different recording formats (for example 7- and 10-inch 78-rpm discs) are considered separate.

Master recordings may have originated from any source, domestic or foreign, as long as they were or are controlled in the United States by a rights holder subject to U.S. law. However, the recording must have been originally intended for issue as a commercial sound recording; recordings taken from broadcasts or film soundtracks are not included. Copyright for those recordings presumably resides with the broadcasting or film entity.

Protected/nonprotected status

Whether or not a recording is currently protected (i.e., controlled by a rights holder) is in the judgment of the compilers. Our approach was to replicate the determination that a reasonable person would make, after a reasonable amount of diligent research, if that person, or his or her institution or association, wished to reissue the recording legally. We did not make use of legal counsel, or expend substantial amounts of time and money, to try to establish with 100 percent certainty the status of each individual recording. Rather, our goal was to determine whether it was probably protected or not. Since our ultimate purpose was to calculate the proportion of recordings from each period that is protected, the exact status of individual recordings was relatively unimportant.

We used three tests to determine whether a recording is probably protected.

  1. Corporate lineage. Can the entity that made the recording be traced forward, either directly or through mergers and acquisitions, to a present-day rights holder?
  2. Marketplace evidence. Who has asserted ownership in the years since the recording was made (a minimum of 40 years in this study), either through legal claims or “authorized” issues or reissues? If the original recording company has disappeared, who has reissued the recording and under what circumstances?
  3. Consultation with experts. The project director and contractor for this report are recording-industry historians, and we were able to resolve many cases. For the most problematic cases, we asked a number of experts with years of experience in the field of reissues their opinion regarding current ownership of the labels involved. These experts are named in the acknowledgments.

None of these tests is infallible. Corporate lineage would seem to be definitive proof of ownership, but it is not. Without access to documents specifying ownership at the time when the recording was originally made and to documents associated with each subsequent change of ownership of the original record label, we cannot be sure to whom ownership of the recordings passed. Press reports that a record company was “acquired” by another can be misleading. For example, scholars were long uncertain about the relationship between the Indestructible Record Company (IRC), a cylinder manufacturer, and the Columbia Phonograph Company between 1908 and 1912. The trade press at the time said that Columbia had purchased Indestructible “lock, stock, and barrel,” and Columbia itself called it an “acquisition.”2 This suggests that Indestructible cylinders made during this period were owned by Columbia and that they would currently be the property of Columbia’s successor, Sony BMG. However, documents recently have surfaced showing that IRC and Columbia simply had a distribution agreement. Since IRC owned the recordings, later went bankrupt, and had no known successor, the cylinders are in fact not protected.

In some cases, ownership is fairly obvious. For example, recordings originally made by Columbia, Victor, and their subsidiaries are now controlled by Sony BMG; Decca recordings are controlled by Universal Music. Products of the small record companies of the 1890s that quickly went out of business and have no known ownership chain to the present day are presumed to be nonprotected. However, ownership of many small labels of the 1920s and beyond is extremely unclear, because of mergers, alliances, exchanges of matrices, bankruptcies, and the like. The Emerson Phonograph Company was founded in 1915, went bankrupt in 1920, operated for a time in receivership, and then suspended operations, was sold and reactivated in 1922, was sold again in 1924, was sold again in 1926, and was discontinued in 1927 (Sutton and Nauck 2000, 76–9, 269, 277–81). Along the way, Emerson masters were released on many other labels, although whether they were sold to or leased by those labels is unknown. The company also spun off a radio division that has lasted to the present day and that may or may not have an interest in some Emerson recordings. So who owns the rights to Emerson recordings today?3

In most cases of past corporate changes, legal documents spelling out the terms of sale are not available. In their absence, one test used to determine transfer of masters during a change of business status is whether the successor label continued to press and sell the predecessor label’s back catalog. If it did, that is a strong indication that the successor did acquire rights to the predecessor’s recordings, not just its trademarks, physical plant, etc. If it did not, that is at least a suggestion that rights to the recordings were not part of the transaction. For example, after the Indestructible Record Company severed its relationship with Columbia in 1912, IRC continued to operate as an independent company for 10 more years, manufacturing and selling the recordings made between 1908 and 1912, among others. This is a clear indication that ownership of the 1908–1912 recordings resided with IRC, not Columbia.

The second test is marketplace evidence. Who has asserted ownership? Liberty Music Shop (LMS), a New York retailer, produced its own recordings, including those of many important stage and cabaret artists, from 1933 to 1942. The store went out of business during the 1970s. It is unclear who, if anyone, now owns the majority of its masters (a few were sold). They have been reissued by a variety of labels, including those of such reputable organizations as the Smithsonian Institution and the Metropolitan Opera Guild, without clear credit. (When a label is unable to find an owner, it sometimes holds money in escrow in case one should emerge.) With no one known to be asserting ownership, our panel of experts believes that, with a few exceptions, LMS masters are currently not protected. Likewise, the Newark, New Jersey-based Manor label, which produced important jazz and R&B recordings in the 1945–1949 period, appears to be in limbo. One of our informants indicated that two reissue producers “tried hard to find ownership of Manor, but the trail went cold.”4 They proceeded with their reissues, in one case putting money in escrow, but no claimant ever emerged.

On the other hand, Sony BMG has asserted ownership to the earliest products of its predecessor companies-including 1890s cylinders of the Columbia Phonograph Company of Washington, D.C., and early Victor and even Berliner discs of the 1890s and early 1900s-and has occasionally reissued such recordings as its own. Given the bankruptcies and numerous changes of ownership that have occurred over the years, it might be difficult for Sony BMG to prove chain of title in court, but custom has been to assume that a company in such circumstances does have such rights, and we have made that assumption here.

One notable component of the public domain for recordings is the body of recordings made by the companies of Thomas A. Edison between 1890 and 1929. Legal title to those recordings did pass to a successor company; however, in the 1950s that company conveyed title to the U.S. government as part of an agreement with the U.S. Parks Department. Thus, Edison recordings can today be used without restriction. No other major company is known to have donated its rights to the public in this manner.

Some 400 labels are represented in this study. Most were individually identified as to protected/nonprotected status. However, even after extensive effort, we could not identify the status of about 100 small labels percent of the total. For those labels, we used a statistical process known as ascription to assign a status.5 With more time and the engagement of legal experts, it might be possible to track down the status of additional labels (and possibly change the assignment of some we did categorize). However, we believe that even with substantial expenditures, a large number—perhaps not much less than the 25 percent that we were left with—would remain unknown. In addition, many of those that were identified as to ownership would be “probable,” rather than “definite,” identifications, owing to the lack of a clear legal paper trail. This is an illustration of the confusion and uncertainty caused by the lack of federal copyright protection of pre-1972 works and related prescribed copyright and ownership markings. The elimination of mandatory registration and marking in the 1976 law will make future research of more- recent publications equally difficult and open to question.

The assumptions made here are not intended to prejudge legal determinations that may be made in the future.

Reissue availability

Current availability is defined as reasonable availability of a new copy to an ordinary person, through normal commercial channels (e.g., store, mail order, Internet). The giant online music databases of Allmusic, Amazon, and MUZE were the principal sources that we used to determine current availability. It was not our purpose to track reissues of extremely limited availability or of availability only to predetermined or exclusive parties, such as club members. Availability had to be in the form of a CD or other currently produced physical format. This study does not consider online access. There is no evidence of legal Internet distribution of pre-1965 recordings not simultaneously offered to the public as compact discs.

Availability from the rights holder means issued by that entity or by its licensee (as indicated on packaging) and legally distributed within the United States. It does not include illegal reissues or reissues available only from foreign sources; these are enumerated separately. It is presumably not the purpose of U.S. copyright law to force those who wish to reissue historic recordings to operate illegally or move overseas.


1 In the early 1890s, technology for duplicating cylinder recordings was quite primitive. Most cylinders sold to exhibitors and to the public were “original” recordings, individually made (or made in small groups) by the artist. These rerecordings are considered production copies, i.e., made for the sole purpose of maintaining stock, and were not considered separate reissuable recordings for the purpose of this study. While mechanically made duplicates of cylinders became more common as the 1890s progressed, mass duplication did not become the norm until cylinder moulding was introduced in 1902. Likewise, early press runs of disc records were limited by technology to a few hundred duplicates of each original recording, after which the artist had to rerecord the selection to maintain a stock for sale. A system for the mass duplication of discs was also introduced in 1902. Virtually no original masters of cylinders or discs survive from the 1890s, and even copies sold to the public are rare, so the reissue of alternate versions would be difficult in any event.

2Talking Machine World, October 15, 1908, 8, 51. As a result of these statements, later histories of recording tended to be somewhat vague about the Columbia-IRC relationship. The widely used From Tin Foil to Stereo by Read and Welch, states that Columbia said that it was “taking over sales” of IRC products (p. 100). Another basic source, Gelatt, The Fabulous Phonograph, asserts that Columbia “took over the entire output of [the Indestructible] factory” (p. 165). Columbia marketed the cylinders as its own, calling them “Columbia-Indestructible” cylinders.

3 Since the last known owner of Emerson, the Consolidated Record Corporation, went out of business in 1929 and had no known successor, we have assumed that Emerson recordings are unprotected today.

4 Correspondence from Dan Morgenstern to Tim Brooks, September 24, 2004.

5 In ascription, a portion of a sample for which a characteristic is known is used to predict the incidence of that characteristic in a similar portion of the sample for which that characteristic is not known. In this case, the sample of small, post-1940 record labels was separated into two groups: those for which protected status is known, and those for which it is not known. The proportion of the group that is protected was then assumed to be true for the other group.