Patenting of Computer Software
Everyone knows that you cannot patent computer software. Right ? Wrong !
The South African Patents Act of 1978 contains essentially the same exclusion on the patentability of computer software as is found in, for example, the British Patents Act or the European Patent Convention. Section 25(2)(f) of our Act states :
"Anything which consists of ... a program for a computer ... shall not be an invention for the purpose of this Act".
Subsection (3) qualifies subsection (2) by making it clear that it is only a computer program as such which is prevented from being treated as an invention. Taken at face value, this prohibition simply states that a computer program as such is not considered to be an invention, and is therefore not patentable. But what is a computer program as such ? The Patents Act itself does not define "computer program", forcing us to look elsewhere for help. The Copyright Act of 1978 offers this definition:
" ´computer program´ means a set of instructions fixed or stored in any manner and which, when used directly or indirectly in a computer, directs its operation to bring about a result".
In other words, a computer program comprises a set of instructions which are fixed or stored in some way, for example, on a floppy disc or in an EPROM chip. Clearly then, a disc or chip on which is stored a computer program cannot be the subject of patent protection. But is there not more to computer software than a mere list or file of coded instructions stored on a disc or chip, or printed on paper ?
Conceptually, computer software may be thought of as a set of instructions which are designed to control the operation of a computer or a computerised machine or system in a desired way, by interpreting and responding to various inputs and by then controlling the operation of the computer or machine accordingly, effectively following "rules" or a logic structure embodied in the software.
One can perhaps think of computer software as being analogous to a play: In one sense, the play is nothing more than a sequence of words written on paper, but in another sense it amounts to a set of instructions which, interpreted by a director and actors, brings to life a performance which is only suggested by the printed text. This immediately hints at the way to avoid the exclusion on the patentability of a computer program as such. Instead of trying to patent the actual program, one looks at what the program is used for. Does it effectively cause a computer to carry out a method or procedure which is new and inventive, or does it perhaps result in a computerised machine which is able to carry out certain unique functions? If so, it is the method or the machine which one can attempt to patent, just like any other patentable invention.
What does this mean in practice? Firstly, it must be understood that different countries tend to approach the patentability of computer software in slightly different ways. Since South African patent law is very similar to British and European law in this area, and since there are as yet no South African Court decisions concerning the patentability or otherwise of computer software, we must be guided by the approach of the British and European patent offices and the relevant Courts. In a nutshell, the approach followed in Europe and the UK when a patent application for a software-related invention is examined is to look for the "technical contribution" or "technical advance" which the invention brings to the field in question. If the invention provides a solution to a technical problem, the mere fact that it is embodied in a computer program should not prevent it from being patented. Conversely, if the computer or machine running the software in question does nothing more than perform an automated version of an otherwise entirely conventional process, there will be no invention.
The recent British Fujitsu case has been interpreted as being a setback to the cause of software patentability. However, a closer look at this case suggests that the general principles set out above are still applicable in the UK.
Fujitsu´s patent application was for a software-related invention which allowed an operator to display the crystal structure of various chemicals on a computer screen and to join the images of different structures together in order to study the combined structure. The system was analogous to the making of physical wire-and-ball models and manipulating them manually. The Court took this into account, and held that the display of the molecular images was wholly conventional at a technical level. In addition, the Court took the view that the invention in question was unpatentable because it amounted in substance to a scheme or method of performing a mental act (another category of unpatentable subject matter). This was because the software in question left it to the computer operator to select what data to work on, how to work on it, how to assess the results and which, if any, results to use. In other words, the software simply provided an aid which was very much analogous to the wire-and-ball models, but did not offer anything more than that.
In the Court´s judgment, a number of important points were made. Quoting with approval from the earlier Vicom case, the Court stated: "generally speaking, an invention which would be patentable in accordance with conventional patentability criteria should not be excluded from protection by the mere fact that for its implementation modern technical means in the form of a computer program are used. Decisive is what technical contribution the invention as defined in the claim when considered as a whole makes to the known art."
To put the matter in perspective, many thousands of patents have been granted by the British and European patent offices for software-related inventions. Very few cases have been rejected on the grounds that they lack technical content. In an interesting article published in Patent World, April 1997, Keith Beresford lists a number of patents which have been granted in the field of software-related inventions, including patents for image processing and computer graphics, word processing software, speech recognition and a data management procedure for a database, to give but a few examples.
In conclusion therefore, instead of assuming that software-related inventions are not patentable and thus abandoning the possibility of obtaining patent protection for such inventions, it makes more sense to assume that such inventions are patentable and that it is only in a minority of cases which do not constitute a technical advance that patentability is dubious. Of course, the absence of any South African case law it makes it impossible to predict with certainty exactly what approach our Courts would take, but it is reasonable to assume that they would be guided by the European and British decisions, which have for some years now followed a fairly consistent line. Also, bearing in mind that the target market of our local clients may well not be restricted to South Africa, it is arguable that South African software companies should as a matter of policy file patent applications locally for new software-related inventions, and that a careful decision should then be taken in due course whether to finalise such patent applications, and in which countries.
Spoor & Fisher