Copyright Doctrines, Abstraction and Court Error

Copyright protection can be divided into ?ve levels: subject matter, level of abstraction, exceptions, term limit, and restricted acts. Although copyright exceptions, in particular the fair use doctrine, and term limit have been subject to signi?cant economic analyses, studies on protection and the limits of protection of subject matter, and level of abstraction in copyright are still fairly scarce. Furthermore, the dominant model for optimal copyright protection is problematic for it requires a standard-based copyright doctrine to achieve what was postulated. Since copyright doctrines in respect of protection based on the level of abstraction are more rule-based in nature, an alternative explanation is in order. In a recent article titled "Copyright as a Rule of Evidence", Douglas Lichtman (2003) hinted such an approach where evidence plays a role in explaining this set of doctrines. In this paper, we use an abstraction and a probabilistic model to explain copyright doctrines. Copyright doctrines such as the idea-expression dichotomy, the originality requirement, de minimis rule, substantiality requirement, merger doctrine, and the scènes á faire doctrine, have the effect of creating a protection divide. Doctrines such as the causal connection requirement, independent creation defence, and the objective similarity requirement, further create an inference divide. We show that the protection and inference divides are relevant in protecting the literal and non-literal dimensions in a copyrighted work. Furthermore, we ?nd that between the protection divide and the inference divide, there is a region of non-strict liability protection. All these three regions, and the related copyright doctrines, are explained by an evidence theory of minimising the risk of court error in deciding infringement cases.