On 5/6/98, JQ Johnson <jqj[_at_]darkwing.uoregon.edu> wrote:
>
> Bob Stock <bstock[_at_]ucla.edu> wrote:
> >
> > Terry's conclusion is a fair one because even though Feist never even
> > mentioned the word "database" in its decision, if a physical database
> > has nothing but facts in it, it is unprotectable by copyright.
>
> I have to disagree with Bob Stock's characterization of Feist and
> databases. As Terry notes, his remarks are limited to "a database of
> the kins at issue in Feist." This reframes the problem as determining
> what sorts of databases are in fact of this kind. I don't believe that
> the test is whether the database has "nothing but facts in it."
>
> Unlike a hardcopy alphabetically arranged telephone directory, the
> creation of a physical (computerized) database requires quite a number
> of creative decisions even after the complete set of raw data has been
> identified, and independent of the particular data values loaded into
> it. For example, consider a relational database with a SQL interface.
> What tables/relations should be included? How much normalization should
> be done? What access tools (search forms, etc.) will be provided? What
> security policies will be implemented? The list of decisions the
> database implementer has to make goes on and on, and is the subject of
> text books and courses in the art of database design. The labor in
> these decisions isn't protected, but the originality is. Since the
> creator of the database has to make so many original decisions, I see no
> reason why the result can't be copyrighted. I'd expect a court to rule
> in favor of the creator of the database if the issue were one of copying
> the physical database (or substantial portions of its design), but in
> favor of the user if the issue were one of extracting the underlying
> data and copying that.
Normally a database (or compilation) is only protected by the originality of the selection, arrangement, and coordination of the unprotectable facts. So, if one determines the expressive layout to be copyrightable and the person copies that layout, he will have infringed. You, though, are arguing that the database structure (as if it were a program) is copyrightable. It would seem to me, if one puts aside the resulting compilation for a moment, that the only way *that* would be copyrightable is if you argue that there are literary components to the database structure and that those components, just like program instructions, are copyrightable.
I would think that a computerized database is made up of two major components, the database itself and any software needed to manipulate the database. If we put aside the software and just look at the database, I would think that the structure contains very little except the data and certain surrounding identifiers. If one were to copy that structure, one would be copying the data (which for the moment we'll assume to be uncopyrightable) and that minimal amount of surrounding stuff. Would you really argue that that stuff is copyrightable? The software interface may indeed be copyrightable but I don't see how the database can be, at least not as a result of any pointers or identifiers imbedded within it. It is essentially a complex, multi-dimensional file. It has no literary components to it as source code does.
This archive was generated by hypermail 2.2.0 : Mon Mar 26 2007 - 00:35:30 GMT