Jay
It seems to me that the “impossibles” are more often attributed to ID arguments by their opponents, than claimed by themselves.
Off the top of my head, for example, Dembski’s “universal probability bound” is just that - a question of probabilities that become far less plausible than design, even if one generously reduces them by orders of magnitude (which, incidentally, Dembski usually does compared to non-ID probabilities he cites from the literature). Now, of course, that does not mean the starting assumptions are right, but that’s a question for discussion. But he’s completely in agreement with the argument that no set of information is impossible (and that any string is as likely as any other) - but that the set of functional information is so much smaller that “cheating” is hugely more likely.
Likewise, Meyer’s most discussed work overtly stresses “inference to the best explanation” methodology, usually by describing the problems with the alternatives already discussed in the existing literature. I can’t recall him using the language of “impossibility” at all.
The one exception amongst the big boys of ID, in my limited experience, is Behe. That seems to be because the form of his argument for “irreducible complexity”, however well or otherwise his examples match it, is one of deductive logic, where there appears to be a logical contradiction in positing a stepwise evolutionary pathway.
Incidentally, most replies to him have proposed hypothetical possible partial solutions that weaken the logical force by suggesting possible loopholes. That certainly makes “impossible” less tenable in a deductive argument, but of course does not provide an actual chain of efficient causation, nor address the probability question.
Regarding IC, Sy Garte gives a good example that struck him forcibly back in his college days: the transcription of DNA requires tRNAs, of twenty different flavours, each matching one coding triplet to an amino acid. That’s twenty separate evolutionary stories required, to make twenty related, but highly specialised, RNA molecules for the twenty amino acids constituting life.
But those tRNAs have to be synthesised by RNA polymerase III, an enzyme complex (which is itself part of a more complex transcription apparatus) - which is, of course, a protein made up of the very amino acids that require tRNA to transcribe them…
It’s actually more complicated than that, of course. Now Sy is not an ID proponent, nor does he say its evolution is impossible, but there is a chicken-egg problem there which rather invites such a description.
So much for “impossible”. But I agree with you that the design argument is cumulative (though remember that, for me, it is axiomatic: unlike ID I’m less interested is demonstrating design to unbelievers than seeing how Christian creation doctrine applies in the world of science).
I’d say, though, that opposition to ID’s cumulative argument (IC [Behe], UPB [Dembski], weakness of non-design mechanisms [Meyer], fine tuning at both cosmic and smaller scales [Denton], metaphysical arguments for First Cause [Torley], form [Sternberg], biosemiosis [Johnson], near-universality of “design instinct” [Axe]…) comes through attacking each aspect by every means available from careful refutation to ad hominems and misrepresentation, without addressing the cumulative case much.
That may be inevitable - it seems ID proponents often similarly attack individual aspects of evolutionary naturalism rather than the cumulative case… though I have seen frequent discussions of the weakness of its underlying Epicurean assumptions, far more than I’ve seen critiques of the actual metaphysical assumptions underlying ID (I must add that “Epicurean”, however, as a term, seems to be N T Wright’s preferred pejorative rather than that of any actual ID people I have read).