I am not a lawyer and this is not meant as legal advice. However, I think everyone would benefit from more information and discussion of copyright law, so please leave comments if you have something to add.
U.S. copyright law applies to writings, or "physical renderings", produced by an author. For this article, we assume the author is a U. S. citizen and the work was produced on U. S. soil. However, a "writing" is not assumed to be human-readable, so, for example, a software program in executable binary form, or "object code", is included [L], section 3.06. The owner of the copyright of a work has the exclusive right for
* reproduce or copy the work,
* prepare derivative works,
* distribute the work,
* perform the work publically,
* display the work publically.
Before explaining these terms, exceptions to these rights, and how these rights relate especially to mathematical works, we discuss works for which copyright law cannot be applied. The law is designed to protect creative written works.
Ideas which are only communicated oraly are generally not subject to copyright, as they have not been "physically rendered." If you tell a friend your unwritten idea of how to solve the Riemann hypothesis, who proceeds to write up the idea and publish it, feel flattered because copyright law won't help you.
An unoriginal work, or a work "mechanically produced", say by a computer program whose use requires no originality, are not copyrightable (more precisely, are not subject to a separate copyright - the program could, for example, output copyrighted elements). For example, the output of an automatic theorem proving program is not copyrightable. On the other hand, the output of
an image processing program which takes an image and applies a de-noising algorithm is a "mechanical" derivation of the original image, so the copyright is the same as that of the original.
Data is not copyrightable. It doesn't matter how much money or man power it took to discover, collect, or obtain it. (However, there are various laws which can be used to protect such intellectual property, such as trade secret laws.) In some cases, a creative arrangement of the data itself is copyrightable, even if the data itself is not.
Works in the public domain (in particular most "official" works by the U. S. government), are not copyrightable. All written works eventually pass into the public domain. Due to the variety of copyright laws which have been passed in the United States over the years, the duration of copyright depends on when the work was written, if it is a joint work (or a "work for hire") or not, and various other factors. However, life plus 50 years is a minimum, according to the Berne convention, so that will apply in most cases.
For the owner of a creative mathematical work, whether it is an article or a piece of software, we explain next what these rights mean.
Reproduction: A reproduction is to fix a copy in a tangible and relatively permanent form, such as a xerox copy or a file on a computer (though a copy stored in your cache is exempted). Aside from non-profit, educational, government, or "fair use", the copyright holder have the sole right to make unlimited copies of your work. For example, if you publish a paper or book, you often sign over your copyright to a publisher. If anyone could make a copy of your article freely, the commercial interest of the publisher would disappear. Similarly, if you wrote a mathematical software program which you wanted to market, you would want to restrict the copies of the program to those who paid for it. A research paper downloaded from the internet and then emailed to a colleague is an example of a reproduction.
However, there is a fair use exception to copyright law regarding copying for personal use if you are a scholar (at a non-profit institute) or the educational use of your students if you are a teacher (at a non-profit institute). These do not apply to commercial think-tanks or to commercial training centers. The guidelines are different for research than for educational use, but the basic idea is to copy no more than is necessary. The guidelines for education are more strict. Generally, 1000 words or 10\% of the material (the minimum of the two) are recommended limits [L], section 10.12.
Derivative works: Only the copyright holder can create a new work which is adapted from the original but which contains copyrightable modifications. For example, if you wrote a mathematical textbook and you retained its copyrights, then only you have the right to create a translation into another language or a second edition. Conversely, if you wrote a mathematical software program which you wanted to give away for free but subject to the open source General Public License (GPL), then you want to restrict the modifications or derivations of your program to those who publically redistribute the modified program under the same open source terms. This is what the carefully crafted legal language of the GPL does for you [F], [W]. (An example of such a project is the SAGE software program [S].)
Distribution: A work is distributed if it is made available to the "public" in some form. For example, a copy in a public library or a file posted on a world-accessible internet site are publically distributed. However, defining the term "public" precisely in this context is a technical legal matter, for which we refer to [L], section 8.13.
Performance and display rights generally refer to works of drama or art, and probably less useful to a mathematician. For these rights, we only mention that a talk on a scholarly paper could be a performance. Legally, such a performance does not constitute a "public distribution" or "reproduction" of the paper itself.
References:
[F] Free Software Foundation, http://www.fsf.org
[K] B. Klemens, Math you can't use,
Brooksings Institute Press, Washington DC, 2006.
[L] M. Leaffer, Understanding copyright law, 4th edition, LexisNexis, 2005.
[S] Sage mathematical software, http://www.sagemath.org
[W] M. Webblink, Understanding Open Source Software,
http://www.nswscl.org.au/journal/51/Mark_H_Webbink.html
2 comments:
A couple of small corrections. Full disclosure: I am a former Mathematician turned IP lawyer. That said, as with your disclaimer, this should not be taken as legal advise.
You state that "[i]deas which are only communicated orally [sic] are generally not subject to copyright, as they have not been 'physically rendered.'" The term of art used is "fixed in a tangible medium of expression." In certain instances, this could be something communicated orally. The concept here goes to immutability. A poem - the words that are used - whether spoken (and recorded), emailed, or written have been fixed somehow. This may be on paper or on some electronic storage device.
Secondly, you state that "data is not copyrightable." This is not correct. Facts - things that are inherent characteristics of something - are not copyrightable. So, for instance, your phone number being what it is or the location of a building are considered facts. Without delving into the possible philosophical issues surrounding mathematics (e.g. the "discovered" v. "created" arguments), formulae, theorems, numerical patterns, etc. are generally considered to be facts, and are thus, not copyrightable. However, you could write a book on a particular theorem and to the extent that book is (1) original, (2) not factually based, and (3) not in the public domain, the words would be copyrightable.
Data is copyrightable in another manner. As a notable case clarified (Rural Telephone), the collection and organization of data can be copyrightable. So, for instance, the set containing all prime numbers would not be copyrightable. The set containing all prime numbers with an inherent characteristic would not be copyrightable. However, if you (for some specific application) selected out and organized elements from that set in such a manner as to be optimal (for that specific application), and this selection and organization was original and not inherently part of that set, you could acquire a copyright as to the original elements of it - e.g. the selection and organization. Such a justification is what is often used to permit copyrighting of geographic data, for instance.
I don't really want to go into it here, but file formats are generally not copyrightable but data structures (in computer science) may be.
By the way, thanks for trying to help educate people on these issues. They're not easy and often misunderstood.
The Leaffer book is very good. For more, you can also grab a copy of "Copyright Law" by Joyce, Leaffer, Jaszi, and Ochoa.
data is not copyrightable. then there are law that say you can copyright certain data. this is illegal copyrighting data is illegal. putting an exceptions kill the idea that data is not copyrightable. under us law laws cannot violate other laws.
Post a Comment