Saturday, November 12, 2005

Dilbert on Intelligent Design

Interesting post by Dilbert on Intelligent Design, which raises some interesting questions of Knowledge and Trust.

He makes the following points.

1. The arguments for Darwinism (and intellectual defences against the "flaws" identified by the Intelligent Design and Creation folk) are complex.

2. Belief in Darwinism depends on trusting the vast majority of scientists working in the field.

3. However, the scientific field relevant to Darwinism is compartmentalized. Scientists are required to trust evidence from other specialisms and disciplines. We are not just talking about non-scientists trusting scientists, but scientists trusting other scientists.

4. Therefore the entire scientific edifice of Darwinism is based on inter-disciplinary trust.

5. If the institution of science is anything like the organizations that Dilbert has made a fortune analysing and drawing, then we have to take seriously the possibility that they've all got it completely wrong.

Of course, the possibility that science has got things completely wrong has been explored by philosophers of science such as Kuhn, Lakatos and Feyerabend. But Dilbert is pointing to a new angle on this - the institutional mechanisms (familiar in large organizations) that permit lots of small bits of evidence to be accumulated and amplified into false knowledge.

Dilbert is making a profound point about the way knowledge is composed from lots of bits of evidence. If much of this evidence is stronger when seen from a distance, and weaker when examined closely, this seems to call the whole body of knowledge into question. (Dilbert doesn't go into the recursive loops of analysis and interpretation, where the evaluation and interpretation of any piece of evidence depends on lots of prior knowledge from elsewhere - on what Bruno Latour calls Black Boxes. But this would add to his argument.)

What Dilbert is rejecting is the theory of repetition, whereby if you repeat something often enough it becomes true, or if you get enough bits of weak evidence from different sources, it becomes strong evidence. This is a theory that is embedded in the way that lots of organizations behave, and in the way that a lot of computer systems produce so-called "business intelligence". Dilbert is all-too-familiar with the ways in which false knowledge can emerge (or should I say evolve?) in complex social settings.

Technorati Tags:

No comments: