On 07/21/13 2:20 PM, Lars Aronsson wrote:
On 07/21/2013 07:55 PM, Amgine wrote:
If I recall correctly, under US copyright case law a word's definition cannot be copyrighted. There are only a limited set of ways to express the definition of a term, and it is recognized this puts an undue burden on new works.
However, a *collection* of terms may be copyrighted under the database portion of the copyright rules.
(I have a dual role here, both as user:LA2 in Wiktionary and as the founder of Project Runeberg, where that dictionary was digitized.)
According to http://en.wikipedia.org/wiki/Database_right "database rights" exist in the EU, but not in the U.S. Such a bill was introduced, but rejected, http://en.wikipedia.org/wiki/Database_protection_bill
When EU introduced the catalog protection in the mid 1990s, they borrowed the concept from the Scandinavian countries, which in the late 1950s jointly renewed their copyright laws (Denmark 1958?, Sweden 1960, Norway 1961). It was designed to give a short protection, 10 years after the first publication, to a catalog, table or listing, hence the name "catalog protection" (Danish: katalogbeskyttelse, Swedish: katalogskydd). Later, this has been extended to 15 years, but it's still based on the year of publication, and not the death year of the author.
The catalog is a collection of facts or short entries which cannot be copyrighted each on their own. However, it is not clear how long such an entry can be before it is eligible for copyright.
In each case, we have (at least) two different issues to discuss: 1) Are the individual entries copyrightable? 2) Is the entire collection protected by database rights?
The 2nd issue is the easiest: Database rights, where they are applicable, only last for 15 years. So anything older than 15 years is free from database rights.
The harder issue is whether individual entries are copyrightable. Nobody gives a clear definition.
When digitizing books in Project Runeberg, I have made the assumption that entries in spelling dictionaries, translation dictionaries, and who-is-who books are not copyrightable. Based on this assumption, I have digitized a large number of such works from Sweden, Norway, Finland, and Denmark which are older than 15 years but younger than 70 years. This started three and a half year ago.
So far, nobody has protested. Since the publishers of these works are still around, and I'm not trying to hide, they are free to express their dissatisfaction, but they have not.
(My assumption does not extend to etymological dictionaries where each entry is a little story in its own right. For such dictionaries, I respect the common life+70 copyright.)
In practice, the who-is-who entries are never copied from Project Runeberg to Wikipedia, but completely rewritten in standard Wikipedia prose, using the digitized book as a source reference, as in this case, http://sv.wikipedia.org/wiki/Gunnar_Blix
The source reference is a 1977 who-is-who, http://runeberg.org/vemardet/1977/0139.html
To reuse a scanned dictionary in Wiktionary, similar transformations are needed.
Database rights is a feature unique to the EU, and is distinct from collective copyrights. As I understand it it only applies to electronic databases, and without regard to any copyrights on the material included. Thus if a database right applies to a scanned dictionary, a separate and independent scan of the same material would not infringe that database right.
Collective copyrights are the rights that apply to a work as a whole, and in the absence of a clearly defined individual being responsible for the whole, are the property of the publisher. Here in Canada, that right would expire 50 years after publication. The rights of individual authors (if identified) in a collective work are distinct from this, and are mostly not subject to the rules for joint works. New material in these collections will mostly have copyrights that survive longer than the collective rights. Publishers can continue to publish the material, but only in the same (or similar?) context. My extended interpretation is that when the publisher's collective rights expire those rights become the property of the public; the public may then reproduce the collective works, but not the still protected items as separate works.
To the best of my knowledge there is no specific legal provision exempting individual dictionary definitions from copyright. What we do have is the principle that information is not copyrightable; only the way that information is expressed is copyrightable. If there is only one way to express an idea it is not copyrightable, because that would prevent that Idea from being expressed. Add to that the fact that many dictionary entries are themselves taken from earlier dictionaries that are already in the public domain, and one must draw the conclusion that very few dictionary definitions are copyrightable. Those few may be within the range of fair dealing.
I have not been very active lately in the English Wiktionary. When I was the community had taken the view that definitions from other dictionaries could not be used in Wiktionary; I don't know if that is still the policy. From an epistemological perspective this can be a big problem. It implies that each definition must be original! That certainly has an impact on reliability.
Ray