Software: Work or Labor?
Is software work or is it labor, asks David M. Berry in the Free Software Magazine. He refers to Hannah Arendt’s distinction of labor and work.
Following the Ancient Greek tradition, Arendt defines labor as something done of necessity. Labor used to be the duty of slaves, now it has become the life of the masses, something that docile bodies do to “make their living”. Work is seen as an activity that creates something that lasts. Work is important because it makes politics possible:
This is important because it is only by escaping necessity (i.e. the constant requirement to produce things we need) that we can begin to communicate and become human as political animals. For Arendt, Work is a prerequisite for the possibility of Action – the realm of great deeds and great words.
Berry argues (like Marxists and the critical theory before him) that in today’s society, we’re all caught in a spiral of laboring and consuming and always alienated from the product of our work. Free software is promising in this respect because the products of free software activities are work - something that lasts - and also commons, i.e. something everybody has access to.
Free software certainly has liberating potential but what makes hacking possible, economically? Free software might turn labor into work, but at whose expence? While we build an information society here in the North, we’re at the same time “outsourcing” the labor into the South in the form of production of food, clothing and electronics. And often on not so fair terms.
Let me a little bit critic with this point of view. Software is “something that lasts”, but it is known for many years now that software also ages, a concept coined by David Parnas in his Software Aging paper around ten years ago. This has been also the matter of study for many researchers, among others Lehman and his findings known as the laws of software evolution (see here) which applies to E-type software which is almost all software. The first law one (known as ‘Continuing Change’) says that a system must be continually adapted else they become progressively less satisfactory in use. So, even if software is mathematically perfect, but it becomes in time less and less useful (that’s why yo probably use one of the most recent versions of GNOME or KDE and not the ones from 1998 for instance). If you consider this, I don’t know if we should put software as labour or as work. My two cents ;-)
oh, sorry, got my HTML wrong in the previous post. I’m really sorry.
grex, thanks for your cents and links. The paper looks interesting. I also fixed your markup.
I share most of your criticism. Although I agree with many Marxist ideas, I’m not so sure about the alienation theory. On the other hand, there is a difference between work done for a proprietary software company and the work done for the joy of hacking. The latter may lead to active life, politics etc. And although particular versions of KDE become less useful, the KDE software as an idea (or codebase or community) is less prone to aging because it is maintained by the community. This can happen with proprietary software as well, but it will stop on the day the company goes bankcrupt.
thanks for fixing what I messed up, Niklas.
I agree with the idea the libre software should be theoretically less prone to aging. We have done some empirical work on this for some libre software projects, but haven’t data to compare it against proprietary software.
On the other hand, I want to point out that there is not that much difference for working on a libre software project or on a proprietary one for the ones who actually work on them. For both of them, and with current laws being the copyright holders means to own the software in the same way.
My point is that the big difference is for the ones who receive the software not for the ones who create it, so that’s why I don’t see the labor vs. work argument being of importance in this area.