Friday, March 17, 2006

Software, Information, and Mental Processes

Working on a paper pointing out that the Supreme Court has never reversed, or even questioned, its holding in Gottschalk v. Benson that computer programs--algorithms--cannot be patented and contending that that holding is correct, I find myself struggling once again to explain matters that, unlike the law, I know little about.

The Court, for example, said in that case that algorithms are not patentable because, although most processes, if original enough, are patentable, algorithms are mental processes. That, of course, raises questions about the nature of mental processes, questions that years of studying the law have hardly prepared me to answer. Other ontological questions raised directly or indirectly by Gottschalk are: "what is an idea?", "what is a number?", "are mathematical truths laws of nature?", and "what is information?".

And they, in turn, raise the ultimate question as to how one can possibly explain such issues to an audience of lawyers and, worse, legal academics.

4 Comments:

Anonymous Laurent GUERBY said...

In the USA, I believe you could make an economics based argument. The USA constitution says that patents are a mean "to promote the Progress of Science and useful Arts". My opinion is that the nocivity of patents in the area of "mental" processes comes from the fact that even relatively small algorithm or software refer to hundreds if not thousands of other "ideas" and the pre-investment in capital is low or zero, unlike their "material" counterparts where the number of connected inventions is smaller plus you'll likely need capital.

Given the huge transaction costs (in time and money to negociate contracts and make your way out of long-running lawsuits) introduced by "mental" ideas patents, there is no way they can be seen as "promoting" anything.

Plus there is now abundant evidence now thanks to the free software experiment that progress is still present in this field without the benefit of patents.

What do you think about this way of putting the problem?

1:00 AM  
Anonymous Laurent GUERBY said...

Another thing, the Supreme Court accepted to review a "mental" patent case but brought on technicality, I asked here if the Supreme court can decide by itself to review the meat of the issue, any idea?

1:05 AM  
Blogger Seth Finkelstein said...

FYI, have you seen the material here:

http://lpf.ai.mit.edu/Patents/patents.html

8:48 PM  
Blogger Peter Junger said...

There are many good arguments agaist software patents that do not turn on the nature of software as information or mental processes, but those or not the arguments that the Supreme Court used to explain why
algorithms are not patentable.

5:05 PM  

Post a Comment

<< Home