The conversation on BioLogos about the implications of Lenski’s E. coli experiment continued, between Richard Buggs and Joshua Swamidass, after I wrote my piece on it here. It turns out that, after discussion, they agreed that, in contradiction of my conclusion, the situation with respect to human genetics is less unpredictable, rather than more, as I suggested there, because of the highly mutable nature of bacteria in comparison to mammals. Lenski therefore seems to have been dropped from the discussion as irrelevant. As Joshua writes, “Retractions are good”!
Joshua asked in a comment here why Richard had introduced the Lenski work at all, and I’d suggest it was simply to illustrate his agreement with GJDS and me that models can indeed be seriously misleading when projected beyond their validated limits. But it seems as if that particular data does not speak to the question of the thread, that of an ancestral human bottleneck of a single couple, so I was wrong there.
Still, another observation I made in a comment to Joshua here remains true: that in a mainly non-biological forum like BioLogos, if the highly technical to and fro between “experts” does not lead to their sweet agreement, as it hasn’t in what is at the time of writing a 530 post thread, then what ordinary people conclude about the question in hand is not actually going to be decided “on the science”, but on some non-scientific basis likely posing as “the science.”
In the specific situation (a BioLogos thread), there has been a strong element of rooting for “Team BioLogos” against the suspicious masked interloper Buggs (is he actually not a scientist at all, but an IDist or even, given his unnatural interest in a human bottleneck, a Creationist??). So it can be pretty much guaranteed that, in the event of an impasse, not only will the original opinion of Venema be preferred, by those people, to that of Buggs, but in their minds this will be taken as the voice of Science.
Those seeking to be more dispassionate, yet equally unable to follow the technicalities, will still inevitably judge on non-scientific grounds. These could be, for example, the human qualities of the participants. Unlike some other participants, Joshua Swamidass has been very quick to acknowledge his errors – as, in fact, Richard Buggs has been ready to admit his own misapprehensions. The very existence of such errors and misapprehensions is an indicator that the matter in hand is highly technical, non-routine and, therefore, fairly unlikely to be capable of resolution beyond all doubt currently. But if the punters have to reach some decision and even the experts can’t agree on the science, perhaps the perceived character of the participants may carry weight. Right or wrong, that’s not a scientific judgement, but a moral one. It’s good to admit errors… but perhaps only bad scientists make them in the first place, and the good ones are right all along. We weigh, we choose who to believe, and in the end it’s a subjective judgement.
I’ve already suggested that there is a “Team BioLogos” effect in operation, crediting the home side as the voice of science. And of course, BioLogos is by no means a scientific consensus, but a small interest group based round a particular model of theistic evolution. But in the unlikely event that the discussion became of sufficient interest to the world biology community that everyone pitched in, lay people could draw their conclusions based on what appeared to be “the scientific consensus”, which perhaps might end up lending weight to Dennis Venema’s “heliocentric certainty” on the matter. But, once more, such a conclusion would not be a conclusion based on science, but on another argument from authority, in this case, the authority of a democratic vote of experts.
The trouble with that, of course, is that at the time of Copernicus, and for a century or more thereafter, the scientific consensus was very much against the certainty of heliocentrism. There were some good scientific reasons for this, as well as scientific reasons for it to change eventually, but it also depended on many non-scientific factors – philosophical, theological, historical, sociological – and personal, like the antagonism between Galileo and the Pope that colours the understanding of the matter to this day. To throw your weight in with a consensus, knowing from the historical perspective that it may well turn out wrong, is more a political act of support for the scientific establishment than a rational conclusion about a particular issue like Adam and Eve. But the layman has little choice, perhaps, not being mentally equipped to decide on the science itself, unscreened by scientific authorities (though, as the BioLogos thread has proved, even science itself is screened by the very questions scientists have chosen to answer).
This distinction between an under-determined science and non-scientific decision-making is actually better seen on another BioLogos thread, in which Dennis Venema and other BioLogians
cheer on discuss his recent critique of ID molecular biologist Doug Axe’s book, Undeniable. All the same motivations and strategies mentioned above have already emerged on the thread, but the “science v nons ciense” trope is easier to see, because rather than being a discussion up in the technical stratosphere, as on the Buggs thread, this one is able to flow in down well-worn channel of the “no true Scotsman” fallacy – Undeniable isn’t science at all, but some kind of natural theology, so Axe is not a proper scientist, though on paper far better qualified than Venema on the subject of protein evolution. Any layman reading can thus easily conclude on which side the authority of Science™ lies.
Since my subject here is how lay people actually decide things on non-scientific criteria (just as Undeniable suggests is the legitimate thing to do!), I won’t enter into that debate too much, except for a couple of examples to show how metaphysical assumptions hide behind “science”. In his original critique, Venema cites evidence that “new genes that code for novel, functional proteins can pop into existence from sequences that did not previously encode a protein.”
Assuming the validity of his evidence, in fact all it shows is that functional proteins do pop into existence. It does not show that they “can” because of some particular ability, because the causes are unknown. The whole question is whether they do it by chance, ie by unspecified natural processes, or by teleological intent. And science simply cannot answer that: if wine “pops into existence” at a wedding in Cana, bystanders will actually judge on some non-scientific basis whether it happened “naturally” or not. The fact that Jerry Coyne might say, “I call that a trick, or a fluke, not a miracle” carries no more intellectual weight than Mary saying “This was the work of God through my son.” It shouldn’t need to be said, but sadly it does, that until you know the cause, you don’t know the cause.
If a scientist wants to attribute the wine’s appearance to chance, he really has to do some legwork to prove it (a difficult job since chance is not a cause at all). In Axe’s phrase, “After all, the only adjudicator on questions of chance is probability.” It’s not enough, as Venema writes at BioLogos, to say that real science is all about overturning ones intuitions. It is only that when it establishes why those intuitions are wrong – and more often than not, in reality, it confirms and quantifies that what already seemed self evident was true.
It’s been a common critique of design arguments, and especially of Axe’s rule-of-thumb “How likely is that?” in Undeniable, that actual mathematical probabilities have not been calculated, so that excluding “natural causes” (undefined!) is invalid. That, however, is how real life works. For example, when Venema claims, as he often has, that similiarities between the chimp genome and the human genome prove common descent beyond reasonable doubt, he’s using a non-mathematical probabilistic argument. Such similarities, he’s saying in effect, could not plausibly have arisen by chance. Like Axe, he quotes no actual probabilities.
Actually, there are many examples of molecular convergence in the tree of life that are routinely said to be evidence that X or Y evolved twice, or three times, or even more. So to distinguish homology from homoplasy in the chimp and the human, one is as much obliged to provide statistical calculations as one insists is the case in proposing, say, irreducible complexity. But nobody, in practice, even attempts to answer the question, “How likely is that protein to have evolved twice by chance?”, still less to compare it mathematically with alternatives, “How likely is convergent evolution in this case in comparison with divergence from a common ancestor?”, and make an informed conclusion.
Here’s the punchline – decisions on such basic matters are usually made, even by scientists, on gut feelings, and not even on the basis of calculations with Bayesian priors. One such gut feeling is that “things can pop into existence”. Another is “No-way – it has to have been designed.” The more interesting, and important, question, is what underlying world-view commitments (not what scientific findings) determine those different responses.