The Scope of CRMS-US

A volume was a candidate for CRMS-US if it matched the following criteria:

• Rights status of “ic/bib” (“in copyright by virtue of bibliographic data,” a default status assigned by the system based on bibliographic metadata)
• Bibliographic format of “bk” (book)
• Published between 1923 and 1963
• Published in the United States (i.e., not a foreign work)
• Written in English
• Not a US federal government document
• Not a translation
• Not a dissertation

For CRMS-US, we focused on reviewing books published 1923–63 in the United States for the following reasons:

• Books were the focus of our review in order to leverage the Stanford Copyright Renewal Database, a resource geared toward “Class A” materials (mainly books), without which a review of a book would currently be a much slower proposition.
• Published works were important because unpublished works may receive a different copyright term and further research is often required when there is an underlying question regarding publication.
• 1923–63 (inclusive) was the time range when US copyright law required renewal of copyright.[11] If a work was first published in the United States during this time period but not renewed twenty-eight years after publication, then it is properly considered to have entered the public domain for the purposes of US law.

The Scope of CRMS-World

A volume was a candidate for CRMS-World if it matched the following criteria:

• Rights status of “ic/bib” or “pdus/bib” (“public domain US only”; both are default statuses assigned by the system based on bibliographic metadata)
• Published in Australia, Canada, or the UK
• Published between the following spans (see paragraph below)
• 1873–1943 (UK)
• 1893–1963 (Australia or Canada)
• Written in English
• Not a translation
• Single publication/copyright date

When CRMS-World was developed, we decided to focus on volumes first published in the UK, Canada, and Australia. We did this for the following three practical reasons:

• For the relevant date range, these three countries represented a candidate pool of appropriate scale: approximately 170,000 works fell into this category
• They were English-language works, which promised to make the review process less complicated for our US-based reviewers
• The legal regimes of these three countries were sufficiently similar to form a coherent project

CRMS review projects, including any review encompassed by CRMS-World, do not consider works published more than 140 years ago. Because international copyright terms typically persist for the life of the author plus a defined number of years after the author’s death, we consider 140 years as an appropriate threshold. Here is our logic: a hypothetical twenty-year-old author writing and publishing in the UK in 1875 would need to have lived to age ninety (1945) for their work to still be in copyright in the UK in 2016. If that same author were twenty-five when publishing in 1875, they would need to have lived to age ninety-five for the same to be true, with more remote scenarios emerging for older authors publishing in 1875.

On the other end of the spectrum, CRMS-World does not currently review UK works published after 1944. We made this decision because the UK bases its copyright duration for most published books on the life of the author plus seventy years. A book published in 1945, where the author died that very year, would be protected until January 1, 2016 (1945 + 70 = end of 2015). From our perspective, it is likely that authors survived the publication of their books by a few years. We sought to maximize our resources by focusing on reviews of books more likely to be in the public domain.

An Alternate Approach: Author-Based Scoping

Though we did not implement it in CRMS, we did consider the potential benefits of an “author-based” approach to copyright review.

The central data point for most non-US copyright determinations is the death date of the author. With a death date, a reviewer could easily make a copyright determination for anything written and published by a given author. From our perspective, scoping your candidate pool to include only works for which an author and death date are known would be more efficient, given that reviewers would not have to approach each work by the same author “fresh” each time.

The challenge with any author-based approach is that books often contain contributions from multiple authors, so your project must be sensitive to the possibility that a given author death date may not be determinative for all works in which that author has contributed material.

With the above caveat recognized, we believe that a properly designed author-based approach may yield substantial gains in efficiency. We also find that an author-based approach lends itself to the identification of high-return death-date research projects. When an author is tied to many works, and his or her death date cannot be located, that information gap can prevent a large number of copyright determinations. Arguably, when we know that the identification of an author death date would provide clarity for a great number of volumes, investing the resources to locate that death date becomes worthwhile.

Another Approach for US Works: Copyright Notice–Based Review

In the United States, affixing a copyright notice on a published work was a formal requirement of the law until March 1, 1989. From January 1, 1978 through March 1, 1989, failure to affix notice to the work could be remedied by registering copyright in the work within five years.[12] Prior to 1978, however, this remedy was not available—virtually all pre-1978 US works published without a copyright notice entered the public domain by operation of law.

At an early stage, we made the policy decision for CRMS not to review volumes for the presence or absence of a copyright notice alone. Instead, for two related reasons, CRMS-US focused on renewal records in the review process. First, the early planners of CRMS-US saw value in leveraging the Stanford Copyright Renewal Database for identifying renewal records. Stanford could be searched quickly and efficiently, making it an ideal tool for copyright determination research and one that nearly matched the speed of checking for a copyright notice in the work itself. Our second concern was the possibility that the scans we were reviewing for our determinations might have had missing pages. This conservative stance was taken to reduce the likelihood of mistakes, and it is one that has resulted in arguably fewer public domain determinations. Today, we have greater confidence in the quality and completeness of scans, lending support to copyright review of US works based primarily on the presence or absence of copyright notice.

Application: US State Government Documents

Our work with US state government documents is one example of scoping a project around the copyright notice formality. Copyright notice review allows reviewers to focus on the volume alone and does not require extensive use of additional research tools.

Our focus on US state government documents is based on a recognized need. Researchers from other institutions depend on state banking reports and similar state documents to perform valuable historical research. It is also based on evidence that many states often did not intend to assert copyright in their publications. When a publicly supported state government document was published without a copyright notice during the time range when the copyright law required such a notice, we see a good opportunity for review.

With regard to US state government documents, the presence or absence of copyright notice is sufficient to make public domain determinations for volumes published from 1923 to 1977. Arguably, review for the presence or absence of notice could be applied to state government documents through 1989, but a project reviewing through 1989 would risk a possible uptick in the number of works that did not bear a copyright notice but were registered within the five-year window.[13]