It has already been noted above61 that the aim of copyright protection for software is an important element in the consideration of non-literal copying, and it is worth investigating more fully.
Kindermann considers the aim of any system of protection to be the provision of “an environment that allows recovery of program package development investment so as to permit further reinvestment.”62 There is an inherent logic in this approach – in a capitalist economy, survival of an entrepreneur is only guaranteed where sufficient income is generated to cover necessary expenses, and is thus desirable. The greatest danger to the software publisher in this respect is software piracy – if too few consumers pay for the program they have acquired, the publisher will not receive sufficient income to cover their costs63.
In Whelan, it was suggested that copyright was a reward for “sweat of the brow.” Karjala64 rejects the idea of copyright as an incentive system on the grounds that computer software is a technology, and thus utilitarian in nature. Technology is otherwise protected by patent and – in theory – only where significant advances have been made65. Patents are thus an incentive system and a system of reward – copyright, which arises upon creation, regardless of the degree of creativity involved, is not. The reason why copyright has been chosen to protect software is the comparative ease with which object and source code is reproduced, unlike other items of technology, and the need, therefore, for immediate legal protection.
While software is utilitarian, the classic recipients of copyright protection (music, art and literature) are made valuable by their aesthetic content or the information they convey. Karjala66 argues that variety is therefore desirable – there is no social benefit in copying another’s novel, only to make minor alterations or to change names. Whether this is true or not, it is plain that variety is not necessarily desirable in software. Utility is not automatically increased by the fact that a task is completed in one out of a range of possible ways67, especially when the number of ways available is limited. Software should therefore be seen as being similar to “works descriptive of technology, rulebooks, histories or legal forms, in which the scope of protection is ‘thin’.”68 This point of view stands in conflict with the approach of the courts in Whelan and Computer Associates, where it was thought that the existence of a variety of expressions meant that the one chosen was protected. In the light of Karjala’s argument, it appears that the courts’ approach – while based on the theory underlying copyright in the “classic” recipients – failed to take into account the utilitarian nature of software69.
It is apparent, then, that the aim of copyright is solely to protect against piracy70, thus to prevent inflicted loss71 leading to market exclusion. With this knowledge, it is possible to consider the actual measures required to make copyright effective.
First suggestions as to the necessary content of copyright protection in the U.S. were made in the Final Report of the CONTU72. It was considered that lawful users of software should be allowed to make their own adaptations73 – that is to say, to alter the way in which software works to accord with their needs. However, such adaptations should not be passed on to third parties without the copyright owner’s authorization. It is clear that the latter would effectively be piracy causing loss to the copyright owner – unless the first user merely passed on a program which altered the program code to implement the adaptations, rather than passing on an adapted version of the adapted program code itself74. The actual act of alteration causes no loss and should not be prohibited. Under current U.K. legislation, such an act nevertheless constitutes infringement (see above)75.
It is plain that the main concern of the CONTU Report was the act of literal copying – “one is always free to make the machine do the same thing as it would if it had the copyrighted work placed on it, but only by one’s own creative effort rather than by piracy.”76 Dietz confirms this in his commentary on the Report – “[copyright] should not… prevent others from contributing to the development of the art, and grant no more economic power than necessary to the developer.”77 Seen in this light, the protection of non-literal program elements becomes questionable. The CONTU Report envisioned no such thing, presumably because it was not considered necessary. Karjala78 points out that no other form of technology has such protection by default. The mere fact that software is comparatively easy to reproduce cannot justify protection beyond the literal. There is, of course, a danger in allowing others to copy “structure, sequence and organization”79 – it may potentially cut short the development time of a rival program, allowing that program to enter the market earlier and thus possibly alienate part of the original program’s market. However, the same may be said of any unpatented technology – and, given the nature of programming, there is no guarantee whatsoever that the program will enter the market any earlier as a result80.
Copyright protection not only includes copying of the SSO, but has also been found to subsist in screen displays81 and user interfaces82. Bainbridge suggests that copyright in the former may arise because displays are works in their own right, because of the copyright in preparatory materials or if the display is “considered to be a photograph or a film.”83 Yet none of this is persuasive – if it is part of the program’s function to produce a screen display, why should that be any more protected than, say, the digital display on a clock? It may be argued that the screen display required more effort to produce, but if the manner in which that display has been generated is protected – i.e. the program code – that should surely be sufficient84. In the case of user interfaces, a similar argument applies, but with the added consideration that uniformity of application interfaces is desirable. Users are not interested in having to learn a new method of operation for each program they encounter, and certainly not where that program fulfils the same functions as another program they have used.
Protection against public performance and display, mentioned above, seems even less justifiable than that of screen displays. Given that the performance or display of the output of a piece of software is merely transient, it seems impossible to argue some form of “fixation.” Unlike music, the software’s benefit does not lie in its very existence but in the functions it performs – hence the need to protect the code rather than what the code does. It also seems inconceivable that the public display of a piece of software, or the software’s output, should cause the copyright to suffer economic loss.