The standard recommendations from Europeana and DPLA is to release content under one of these machine readable license statements, when access or another right may be restricted: http://rightsstatements.org/en/ . The documentation section of that website has two White Papers, which are the consensus among GLAM + copyright experts in this kind of copyright situation on both sides of the pond -- and its specifically designed for the works that we simply, just don't know about or where some type of contract restricts use of the digital object. 

With one of these statements, it gives reusers the chance to clearly justify reuse or challenge the GLAM placed license on a situation by situation basis, while reducing the liability greatly for the original institution. 

Cheers, 

Alex Stinson

On Fri, Sep 9, 2016 at 7:40 AM, Andrew Gray <andrew.gray@dunelm.org.uk> wrote:
Hi Fae,

Under UK law, the technically correct but unhelpful answer is that no
such cut-off exists - because of the deeply silly 2039 transitional
rights for unpublished works. You can, of course, advise them to do
the sensible thing and ignore this, & the BL's Illuminated MS page is
a good example of how to do so:
https://www.bl.uk/catalogues/illuminatedmanuscripts/reuse.asp (and a
convenient precedent should they want to justify taking the same
position)

For published works...

The ARROW study the BL did in 2011 -
https://www.arrow-net.eu/sites/default/files/Seeking%20New%20Landscapes.pdf
- noted that the extreme cases can be very extreme - during 2010, at
least one work from 1859 was still known to be in copyright under
life+70, and during 2011, at least one from 1865 - 151 and 146 years
since publication. These are both now PD but it demonstrates the
challenge of setting a 'safe' hard cut-off. It is quite possible that
there is some notionally copyrighted pre-1874 material still out
there; that's 142 years, less than these two extremes.

If I was to make a wild guess at the edge of a "reasonable presumption
of safety", covering almost the full range of material but leaving off
the extremes, I'd say 140 years - work published at 20, lives to 90,
life + 70 takes it to 140. This sounds like the 1874 threshold you
mention.

The Wellcome published a report looking at digitisation of both print
and archive material -
http://www.create.ac.uk/wp-content/uploads/2013/12/CREATe-Working-Paper-No.10.pdf

While they didn't use a cut-off for considering books PD - they
assessed rights on an item-by-item basis - they did decide to post
some material on a "we'll see if anyone complains" basis, and this was
given a basic date-triage:

"The WL requests for permission to publish online which did not
illicit a response will be made available in three batches – the
first, which included titles published up to 1930; the second, from
1930 until 1970; and a third batch, from 1970 up to the present day,
which caused the greatest concern to the Wellcome Trust’s legal team.
The most recent batch is regarded as most problematic because the
rightsholders will almost certainly still be living, and therefore
could be more likely to object to publication without permission."

For archive material they took a harder line, and (given the subject
material and dates) presumed it all to be in-copyright and requiring
consent. This may not be relevant for your situation...

Andrew.

On 9 September 2016 at 10:33, Fæ <faewik@gmail.com> wrote:
> Hi, I could do with some quick views to advise a UK GLAM institution
> as to how to make part of their digital archives of images "No
> copyright known", based on the date of creation of works (a proportion
> may have year of publication instead). To make this an easy step for
> the GLAM, there needs to be some evidence of why any date chosen is no
> risk, or extremely low risk, so that a case could be made to the
> institution's management; and lawyers.
>
> I'm aware that Google Books ran into a lot of lawsuits for releasing
> book scans a bit too liberally, while at the British Library there was
> a very conservative date of 1874 (I think) where everything before
> than could be considered public domain as a default.
>
> The tricky part is to choose a generic date that can apply regardless
> of the nature of the item digitized, or its country of
> creation/publication.
>
> I have created a thread on-wiki at -
> https://commons.wikimedia.org/wiki/Commons:Village_pump#For_an_institution.2C_can_we_advise_on_a_single_year_before_which_all_works_are_public_domain.3F
> - any links or advice from publications would be greatly appreciated.
>
> Thanks,
> Fae
> --
> faewik@gmail.com https://commons.wikimedia.org/wiki/User:Fae
>
> _______________________________________________
> GLAM mailing list
> GLAM@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/glam



--
- Andrew Gray
  andrew.gray@dunelm.org.uk

_______________________________________________
GLAM mailing list
GLAM@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/glam



--
Alex Stinson 
GLAM-Wiki Strategist
Wikimedia Foundation
Twitter:@glamwiki/@sadads

Learn more about how the communities behind Wikipedia, Wikidata and other Wikimedia projects partner with cultural heritage organizations: http://glamwiki.org