Bayesian = Science?

A few weeks back, Bilbo wrote the following at Telic Thoughts:

Contrasted with this is Mike Gene’s own approach, which he labels Inductive Gradualism. Here, one may begin by only suspecting that a certain thing is designed, and then look for further evidence that may strengthen the suspicion, eventually making that suspicion plausible, then probable, and perhaps finally near certain.

A critic/scientist with the handle Raevmo replied:

How is that different from normal scientific practice? Sounds almost Bayesian to me: update the prior with new evidence to adjust the plausibility of the hypothesis (or “suspicion”).

One problem. Just because my approach is Bayesian does not mean it is science. Let me easily demonstrate this.

Abiogenesis expects life to be multifarious.

In the last posting, I noted the Spawning Story leads us to expect that life should be multifarious.

The idea of messy simplicity entails that there are many, many ways to skin a primordial cat. That is, one envisions that life-like functions could be rather easily recovered from the prebiotic soup not only because the soup was rich with all sorts of potentially useful chemicals, but also because such functions could be carried out by a wide assortment of these chemicals. If there are so many ways to build a self-replicating homeostatic system, one would think the soup was continuously turning out a dizzying array of proto-organisms. Recall Stanley Miller’s list of prebiotically produced amino acids. If we kick off our story with messy simplicity, the “choice” of the ten amino acids used by life-as-we-know-it are only one of many, many possibilities. Various other permutations of amino acids, including both longer and shorter lists, might have been used to produce hundreds or thousands of different types of self-replicators (or peptides that helped the self-replicators). As such simple entities gradually became more complex, each one could spawn another set of permutations, tapping into the functional promiscuity that comes from exploiting all those sloppy interactions as entailed by the Spawning Story. A variety of means to synthesize polymers, perhaps involving all sorts of difference codes, could likewise evolve from each ancestral group built from different molecules, using its own set of co-opted parts and sequences it just happened to stumble upon. In other words, if we are to explain the origin of life through a gradual series of spawning events, the potential for permutations exists at every step, leading to an explosive cascade of diversity. The only theme that would unite all this diversity is that they all reproduced, showed catalytic activity, and performed some type of metabolism.

Of course, when we survey the living world, this level of deep diversity is completely missing. And this is a serious problem for proponents of non-telic abiogenesis. Put simply, the living world does not look like it was spawned from a set of messy, simple conditions.