[sword-devel] Copyrights and derivative works

Chris Little sword-devel@crosswire.org
Sat, 18 Jan 2003 05:15:50 -0700 (MST)

On Sat, 18 Jan 2003, Daniel Russell wrote:

> NONE of these examples listed by the copyright webpage are like a 
> concordance.

The examples are not exhaustive.  They are, rather, I suspect, the most
common examples.

> A concordance is a precipitation of facts, not a remolding of a work. In 
> the examples, the works are merely translated, rearranged (what 
> instruments play what parts, etc in a musical score), reproduced 
> (copied), abridged (same text, but shorter), condensed (similar to 
> abridgement; this is not what a concordance does).

A concordance is the quotation of an entire work.  It doesn't matter how
you re-order it or in what manner you change the text, it is still
derived.  Furthermore, the examples given are specifically of types of
derivative works that are eligible for copyright.  Since a concordance is
non-creative, it is not eligible for copyright on this basis.  It is
nonetheless a derivative work.

If you make B out of A, then B is a derivative of A.  If you make a
concordance of the NIV, it necessarily requires use of the NIV--that is
words themselves--so it is a derivative.  If you took an ASCII text of the
NIV, gzipped it, uuencoded it, fed the result to a printer, and bound the
printed papers without telling anyone what it really was, it would STILL
be a derivative work.

> This is such a preposterous argument that any judge that has an I.Q. 
> above 70 is going to throw it out of court. Just because something 
> *could* possibly, maybe, sorta be reverse-engineered, does not mean that 
> it is practical or likely that it will be reverse-engineered. And 
> frankly, i don't think a concordance can be reverse-engineered to 
> produce a text that is more similar to its object than some other, 
> independently copyrighted translations are. For one thing, a concordance 
> does not record the order of each *instance* of a word, and generally, 
> concordances do not contain pronouns, and other mundane words, which 
> make up a large portion of the essential grammatical structure of passages.

Actual ability to reproduce the original is irrelevent, but of some
importance is the fact that a lot of information from the source text is
being copied, without license, into the derivative.  It is enough to glean
a lot of details about how the NIV translates certain words without
needing an NIV.  For example, if you know that a certain word in a certain
verse is by some translations rendered "love" and by others "like", you
could easily do a search on that verse to retrieve all words it contains.  
By investigating that list for the words "love", "like", and similar, you
can probably figure out how the NIV renders that particular word.

> In short, the examples listed in the U.S. copyright definition of 
> "derivative work" are there to show us that only works that are 
> basically the same as the original with some recasting count. Works that 
> are *related* to the original work, or that refer to it, are not 
> derivative works; unless of course you have some extremely dishonest and 
> talented lawyers in a courtroom of some senile judges.

Being "[b]asically the same as the original" isn't necessary.  A painting
based on a scene from a novel isn't even remotely similar to the
expression in the novel, but it's a derivative work.  Works that cite an
original work are not derivative, but that's not what concordances do.  
Concordances quote the entire work.

But, hey, don't believe me.  Call UBS; ask them if they would consider it
a derivative work.  Hire a copyright lawyer; ask him if it would legally
be considered a derivative work.