Wednesday 30 January 2013

SAS v WPL - programming languages not protected by copyright

 Yesterday the High Court handed down its decision in SAS Institute Inc. v World Programming Ltd in which it applied the CJEU's decision about which aspects of a computer program may be protected by copyright (which Jeremy reported on back in May of last year, here).

A reminder of the facts

SAS developed the SAS system, which enables data processing and analysis tasks. A key aspect of the SAS system is that users can write and run their own applications to use the system to manipulate data. These programs had to be written in the SAS language, a programming language developed by SAS meaning that users were then tied to the SAS system to run their own applications.

Along came World Programing Limited (WPL), creators of World Programming System, a system which replicated the functions of the SAS components. Crucially, World Programming System was compatible with the SAS language meaning that users were no longer tied to SAS and could use their own applications with World Programming System instead of the SAS system.
WPL created World Programming System by studying a Learning Edition of the SAS system (supplied under licence) and by consulting SAS's manual. There was no allegation that WPL had access to, or copied, SAS's source code.

The CJEU's answers

In responding to the High Court's questions on the interpretation of the Software Directive and the Information Society Directive, the CJEU held that:

- the functionality of a computer program, programming languages and the format of data files used in a computer program cannot be protected by copyright;
- users may observe, study or test the functioning of a program in order to determine the ideas and principles which underlie that program, notwithstanding contractual restrictions imposed by the owner of the program; and

- reproduction, whether in a computer program or in a user manual for that program, of certain parts of the user manual for another computer program may constitute infringement of copyright subsisting in that manual.

As Ben observed at the time, this was all fairly unremarkable.

The High Court's decision

So, the case was referred back to the High Court where Arnold J was required to determine whether WPL had infringed SAS's copyright. The short version is, that WPL had not, except where it had copied the SAS Manual. Arnold J held that the WPL Manual contained a substantial part of the SAS Manual so WPL had infringed copyright as set out in his first judgment (see here). Apart from that WPL had not infringed copyright in the SAS system.
The longer version is that, despite saying that SAS could not claim copyright in the SAS language, (because they had not pleaded so in the first place, and it was too late to re-amend their re-re-re-Amended Statements of Case) Arnold J went on to consider in some detail whether a programming language could be protected by copyright. He was of the view that confusion could arise when considering fixation. A programming language can be a way of fixing a work: source code can store an artistic image or a literary work. However, he said, the technical means by which fixation is achieved is irrelevant. Fixation should not be confused with the work itself.

He compared a programming language with a conventional spoken/written language saying that:
"A dictionary and a grammar are works which describe a language. Such works record, and thereby fix, the elements of the language they describe: the meanings of its words and its syntax. It does not follow that the language is a work. Rather, the language is the material from which works (including dictionaries and grammars) may be created… Programming languages such as the SAS Language are no different in this respect."

This seems uncontroversial: no-one would argue that the English language should be a copyright protected work over which one person should have a monopoly; in Arnold J's view it was irrelevant that the SAS language was a manufactured language as it would evolve with time in the same way as any other language.
What is interesting about Arnold J's analysis is that in considering the originality of a programming language he used the "intellectual creation" test, rather than the "skill, labour and judgment" test. Assuming a computer language is not a computer program, this implies that the test for originality is now harmonised for all works rather than just for computer programs, databases and photographs (see here and here).

More interesting still is that Arnold J went further, and said that even if he had found that a programming language could be an intellectual creation, it did not follow that it had to be a work. There has been some concern that the CJEU, in cases such as Infopaq, BSA, FAPL and SAS v WPL, have not only implied that the test for originality has been harmonised but also that all works that are their author's own intellectual creation are protected. This concept does not sit comfortably with s.1(1) CDPA. Arnold J seems, for the time being, to have clarified that the CJEU's broader definition of a copyright protected work does not apply in the UK.
Arnold J went on to hold that there was no evidence before him that the SAS data file formats were original so he could not find that they were works in which copyright had been infringed.

Arnold J made a further clarification when consider whether a substantial part of a work can be reproduced by elements which are themselves not protected by copyright. He relied on Infopaq to say that that was not possible. Therefore because the functionality, the programming language and the data file formats of a computer program are not protected by copyright, they are not relevant to the question of whether a substantial part of a work has been reproduced.

Finally, as regards the Learning Edition which WPL had used, under licence, Arnold J held that such use was within Article 5(3) of the Software Directive. Further he said that if such use was contrary to the licence terms those were null and void by virtue of Article 9(1) of the Software Directive. Therefore WPL did not infringe by using the Learning Edition.

So in summary, Arnold J confirmed that copyright cannot protect the look and feel of computer programs, which will not come as a great surprise to most. In this blogger's view the most interesting part of this judgment is the analysis concerning programming languages which explicitly cannot (and should not) be protected, and Arnold J's clarification that a work is not automatically protected simply because it is an "intellectual creation".

3 comments:

Unknown said...


As a student of Computer Science in the 80's, it was always our understanding that copyright would give some protection to a piece of software. What that protection was few of us ever needed to discover.

From reading this we must conclude that the only protection that can be properly gained on a piece of original work is a patent on a process which we may have devised, assuming that it is novel and new.

However I must ask the question by way of analogy. The ring tone or corporate gingle is a valuable asset and subject to copyright.

This ruling if applied to the music industry would seems to imply that should a composer come up with the same string of notes on a different set of instruments would not infringe the copyright?

Or am I just off key?

Andy J said...

@Ruan.
I think you may be off-key. If a musical work contained sufficient notes to get over the de minimis threshold, it would probably gain copyright protection, irrespective of the instrument(s) it was played on. Indeed I seem to recall obiter in a case, the name of which escapes me, that even transposing the key in which the tune was played would still infringe copyright, assuming a substantial part was copied. Incidentally, it would probably much easier to use trade mark law to protect a corporate jingle, as I believe Microsoft and Intel have done in the past with several of their jingles.
Also I wonder if you are not conflating software with the programming language it is written in. Mr Justice Arnold was clearly talking about language, since the actual software WPL wrote did not copy SAS's software, it was merely written so as to interoperate with scripts already written in SAS's programming language.

Anonymous said...

Please can someone explain two issues:
(1) if I am found by Trading Standards in possession of software for which i do not have a valid license, do i now have a defence if i allege i am only using the software to study its underlying ideas & principles? Isn't this the same as a user at WPL accessing the software without a valid license?
(2) substantial copying - where in the CDPA is it a requirement to consider the originality of the work and also the originality of the part of the work copied? If the work is original and if a subtantial part of it in terms of quality is taken, you have prima facie infringed copyright. If you have taken the core of the product i.e. how it functions, have you taken a substantial amount of the quality of that work?
(3) where does this decision leave the software industry?