Eric Anderson, over at Uncommon Descent, gives what is IMO a very concise and lucid account of what ID does and does not do:
You may be disappointed that ID doesn’t identify the designer. You may complain that ID doesn’t provide a bright-line definition of mind, or consciousness or other things that scientists and philosophers have struggled with for centuries. You may be frustrated that ID doesn’t provide more answers than it does. But these are frustrations borne of your own desires and expectations, not a problem with ID itself.
ID is not a theory of everything. It has never claimed to be. ID does not seek to identify the designer. It is not a philosophy of mind, or consciousness, however interesting those topics may be. We may wish to pursue those topics as follow-up questions to ID, but it is not a failing of ID that ID does not have answers to everything. I’ve said it and will say it again: ID is a very simple and limited inquiry, and can be understood and explored and addressed with a very basic common every-day understanding of what we mean with words like ‘design’ and ‘plan’ and ‘intent’ and ‘purpose’. There is no need for definitional plays or semantic deep-dives. If someone wants to go beyond that and ask philosophical questions or speculate about mind, consciousness, the nature of reality, and so forth, great. Those are valuable inquires in their own right.
However, by the same token it seems to me to highlight what, IMO, is wrong with ID, without muddying the water with extraneous claims.
It is certainly not a failing of ID that “not a failing of ID that ID does not have answers to everything”, any more, I would say, that it is a failure of Darwin’s theory that it does not have answers to everything, not least being an answer to the question of how life began.
Nor is it a failing of ID that it does not “provide a bright-line definition of mind, or consciousness or other things that scientists and philosophers have struggled with for centuries”. But where Eric is wrong, I think, is where he says:
ID is a very simple and limited inquiry, and can be understood and explored and addressed with a very basic common every-day understanding of what we mean with words like ‘design’ and ‘plan’ and ‘intent’ and ‘purpose’. There is no need for definitional plays or semantic deep-dives.
Yes, there is such a need, if ID is to be regarded as a testable scientific hypothesis. In order to test a hypothesis in science you need to operationalise it, and to operationalise it, you need an operational definition of the terms used to express that hypothesis, in this case, words that include “design” and “plan” and “intent” and “purpose”.
And to quote the linked definition on Wikipedia:
An operational definition defines something (e.g. a variable, term, or object) in terms of the specific process or set of validation tests used to determine its presence and quantity. That is, one defines something in terms of the operations that count as measuring it.
And that, to me, is where ID falls down. Dembski, to his credit, does provide a reasonable operational definition of intelligence (“the power and facility to choose between options“), but does so in a way that excludes “intention” (specifically, in fact) and so inadvertently covers natural selection (a point that I made, not originally, in my first doomed foray on to UD). But if IDists do invoke “intention” as a necessary attribute of “intelligence”, it seems to me that their argument no longer holds anyway, as “choosing systems” that are nonetheless not “intentional” can be readily demonstrated to produce precisely the kinds of patterns that Dembski, ironically, rightly, given his definition, attributes to “intelligence”.
Moreoever, Eric’s post also reveals what I think is a more fundamental flaw in ID: the fact that it is a default conclusion without a specific hypothesis. It is arrived at (at least in Dembski’s formulation) by rejecting a null. Which would be fine, if what is usually called “H1” was defined. But it isn’t, not even using “very basic common every-day understanding of words like design”. It is the default conclusion to be inferred if a pattern cannot be accounted for by “Chance”. The problem here is that not only is H1 not defined, but H0 (the null), namely “Chance” is not defined either. The implication seems to be that “Chance” stands for “drawn from an equiprobable distribution of patterns”.
But the theory of evolution (in any formulation) does not state that complex life forms equipped with all kinds of fancy gadgets that promote not only their own persistence in a environment full of resources and threats, but the persistence of their kind, came about because certain configurations of organic compounds happened to be drawn from some cosmic hat. It provides a whole series of mechanisms (involving both choosing mechanisms and mechanisms that provide options to choose between) by which such forms will tend to arise, given some starting conditions.
And this is why ID is essentially vacuous. If all it does is what Eric rightly says it does, then it fails. It can only succeed if it tries to do what Eric rightly claims it does not try to do: actually propose mechanisms by which a designer might not only have designed biological life, but implemented it, and derive testable hypotheses from that proposal.
This is why I was delighted to see Genomicus’s series of posts (here, here and here) on “frontloading”. I think his ideas are a long way short of a viable program still, but he’s on the right lines. But then he’s also going well beyond what Eric claims is ID’s remit.