Thursday, August 05, 2010

Open Source and Science: A match made in pragmatism.

Note: This is an issue I care a lot about, and so I've written about it before. This is kind of a "take two," where I hope to expand on previous writings.
I am an idealist of a peculiar kind, in that one of my highest ideals is that of pragmatism. For instance, I posit that society should not spend immense amounts of resources in efforts that evidence has shown to be futile. As I wrote about in my post on science and faith, this is one of the quintessential features of science. So much so, in fact, that one may define science as the set of those means of learning which expand human knowledge.

This pragmatic ideal is what determines much of the way that we do science. We have found that, as the sphere of human knowledge grows, it has quickly transcended the capacities of any one human brain. Thus, in order to continue to do science, we have recognized that science must be a social enterprise. This then requires that we have some means for accepting that what another scientist tells us accurately reflects reality.

It is this point at which many will claim that science must base itself on faith; specifically, faith in the goodwill and honesty not simply of our peers, but of all who came before us. Through pragmatism, however, we see that this is not the case. Rather, necessity has driven a complicated system of social protocols for communicating science by which everything is reproduced and verified such that errors due to misplaced trust are minimized. A key aspect of this social system is that science is done in the open. While truly taking seriously the ramifications of such a principle is an effort still in its earliest stages, we have long recognized that science cannot go beyond that which is communicated (put differently, science cannot exist in a vacuum). Thus, secrecy has no place in the development of scientific knowledge. In order to truly succeed in the sciences, we must wholly embrace the social and open nature of science.

Of course, science in the abstract is not the only place that we find such concerns. Consider, for instance, a computer. It was not too terribly long ago that a single person could in principle understand every aspect-- perhaps even every circuit component-- of a computer. Despite their immense physical size, computers of this age were small in the sense that they fit into the human mind. Now, however, there has been so much technological progress that it is ludicrous to think that a single person could design a modern computer from first principles. Rather, the development and manufacture of computers is a social enterprise, and not just to the extent that it overlaps with science as we have discussed it so far.

To make the discussion still more concrete, we can consider the enormity a modern operating system. The Linux kernel alone has grown from 10,000 lines of code to about 13,000,000 lines, representing far more work than any one individual can master. Such a task is undertaken with open collaboration and communication, whereby each contributor can focus on some subset of the immense whole that is the Linux kernel. This is to say nothing of all of the other parts required to make up an operating system, such as a desktop environment and low-level userspace utilities. The modern operating system must be a true community effort, if only due to the proportion of the task.

In order, then, to develop software commercially, one must create within their company a microcosm of this sort of community. Undoubtedly, this can be done, as the evidence exists in the form of closed and proprietary software. Science, however, serves in this instance to show us the value of an open flow of ideas. We spend an immense amount of effort in the sciences on facilitating communication, utilizing everything from conferences to telecommunications as tools to do so. It is, in essence, an openness born primarily of a pragmatic ideal which can be readily seen to apply in society more generally.

The story of openness, however, is far from being a constant push from academia to the rest of society. Indeed, much of the current open science movement relies on the open source movement for its inspiration. There is, in fact, a healthy community on the boundary of the open science and open source movements. This community is a wonderful example of the more general realization that pragmatism can and should drive forward the open exchange of ideas.

I would be remiss if I did not take the opportunity to share at least a few good links on the subject. In particular, I have provided below links to the work of a small sampling of the people who comprise much of my view of the open science movement.

No comments: