ID supporters continue to send me emails about Josh Swamidass. The latest hammers on a comment I made in 2008 at Uncommon Descent, namely: “I’ve pretty much dispensed with the EF [Explanatory Filter]. It suggests that chance, necessity, and design are mutually exclusive. They are not. Straight CSI is clearer as a criterion for design detection.”
I would not write that now. In my view the filter is just fine and it neither conflates nor falsely differentiates the three modes of explanation (chance, necessity, and design). My comment back then should be seen as an unnecessary concession to critics, not as undercutting the filter per se.
To properly use the Explanatory Filter, it is vital to identify what exactly one is trying to explain. Take a rusted automobile. In Jonathan Waldman’s wonderful book Rust: The Longest War, one reads that an average car can lose 10 pounds of weight per year. So, if one is trying to explain why a car after ten years has lost about 90-100 pounds, that is the result of rust. It’s a high probability event.
A Probabilistic Phenomenon
Is it chance or necessity? Rust, as with the kinetic theory of heat, is at base a probabilistic phenomenon, but when averaged over huge numbers of molecular events, the probabilities come so close to one that necessity becomes a natural explanation as well (it’s no criticism of the filter if it allows chance and necessity to bleed into one another for events with probability extremely close to one). What about the sagging of the car’s shocks over time? It seems that can readily be explained by necessity. What about the pattern or rust on the doors? It seems chance is as good an explanation here as any.
What about the structure of the chassis or the differential? Put that into the filter, and you’ll get design. So let me reiterate: I really still like the Explanatory Filter, and any “backing off” I may have done reflects an unnecessary concession to critics.