The above is a schematic of the Intelligent Design community’s explanatory filter, an epistemic proposed by ID guru William Dembski for detecting the presence of Intelligent Design. Other versions of this epistemic can be found on the internet but they all have multiple issues. In fact from a theological point of view the worst problem of the filter is that it prompts a God vs. Nature dualism when the intelligence is identified with God. As I have remarked many times before if one runs through this procedure and finds that life can, after all, be generated by “natural” law and disorder processes then the filter might lead one to conclude “Life isn’t designed”, a conclusion which ultimately threatens theism. As a consequence this epistemic has helped create strong polarization of the debate: The ID community stake all on the belief that “natural processes” (i.e. physics) simply can’t generate life and they use bogus arguments to prove this to themselves: See for example their misuse of the 2nd law of thermodynamics, a point of view that they are committed to defend at all costs. Conversely, atheists have a huge stake in finding ways whereby OOL and evolution might work without recourse to "black box" intelligent activity filling any gaps.
As well as theological problems I have a couple of smaller niggles with Dembski’s procedure: Firstly, his separation of law and chance; in my experience physics tends to form one seamless law and disorder (L&D) package. Secondly, I’m growing more and more uncomfortable with the IDist dichotomy of intelligence vs. L&D processes. After all, it appears that the functioning of the human mind has a considerable point by point conformity with neural operations, operations which in large measure at least follow known L&D principles. This polarization of L&D and intelligence into two separate categories may be fundamentally flawed.
The dichotomies championed by the ID movement influences the whole way they think about the subject. Their parceling “intelligence” into a separate black box category seems to be the reason why some of them, of late, have been digging themselves into another hole: Viz they are declaring that it is sufficient for their purposes to simply identify an object to be the product of intelligence with little or nothing being claimed about the character of the intelligence involved. To my mind this is clearly an error which potentially undermines the whole ID project: Since it is necessary to have at least some inkling about the nature of intelligence in the first instance in order to identify and anticipate its works, this strategy entails a contradiction. But in actual fact one finds that attempts by the ID community to use their ideas to make predictions do depend at least on a modicum of implicit understanding about the nature of the intelligence involved!. In particular, the prediction by IDists about there being no junk DNA is posited on the basis of an assumption about how an intelligence may prefer to work.
However, although I believe that Dembski’s broad brush epistemic procedure falls over when it is invoked in disciplines like theology, philosophy, and the nature of intelligence, it is nevertheless true that something like it does apply in many everyday situations where we try to determine whether or not an event or object has an “artificial” or “natural” origins. Moreover, the epistemic is also relevant to archeologists and imaginary scenarios like 2001 Space Odyssey. But in an attempt to fix some of the filter's problems and improve its robustness and durability in more nuanced disciplines I have designed this procedure:
In the above procedure we get hold of some “object” to be investigated (top box) and then pass it on to a process called “Get Object Origins”. This latter process may well contain something that looks like Dembski’s explanatory filter, a procedure which attempts to explain an object either in terms of L&D or intelligent agency – but you will notice that I make no explicit reference to this dichotomy because it is not as sharply defined as the ID community make out. Also, in the flow diagram above things don’t end there. We go onto to create an “origins object” and then after outputting the results we ask the question is this object complete? By that I mean that some decision has to be made as to whether the explanatory activity has found an object which is deemed to be a complete explanation of the cosmos. I’m in no way suggesting here that explanatory completeness is something that is either easily attributed or can be done so without controversy – this decision box clearly hides plenty of philosophical cut and thrust. However, if ever the explanatory project is deemed complete the process halts. If on the other hand the epistemic procedure doesn’t halt the “origins object” is submitted back to the “Get object origins” process which is in affect now attempting to get the origins of the origins. This means that if OOL and evolution should, within acceptable scientific standards, be demonstrated to be a product of known physics that physics may then itself be submitted to origins research. I emphasize "may" here: Who knows; if biology is shown to be a product of physics some people may experience no philosophical unease making them feel there are further questions to be answered. For theists on the other hand this would be unacceptable: They will likely feel that even the most complete L&D physical regime calls for more origins research. Moreover, some atheists may also be inclined to resubmit their physics to origins research, perhaps hypothesizing some kind of L&D multiverse. So, in short my procedure caters for both philosophically minded atheists as well as theists.
The belief that the detection of the presence of intelligent agency can be sharply distinguished from questions about the nature of the intelligence concerned appears to be another of the ID community’s false dichotomies. But this dichotomy does not sit well with the comments of one of their own gurus, Granville Sewell. I applauded Sewell’s comments and made reference to them in this blog post. In that post I suggested that Sewell was starting to open up what these other IDists treat as a black box in which sits some kind of localised alien homunculus. In contrast Sewell is starting to discover the consequence of the fact that intelligence is not an intensively located homunculus, but an extensive property spread over more than one subsystem, and which probably involves, I submit, non-locality. The problems introduced by thinking there is a sharp (and artificial) demarcation between intelligence and L&D is once again foisting upon theists a Pandora’s box of problems.
Below is my enhanced explanatory filter expressed in C-like pseudo code. I have made it recursive in order to give it an elegant look.
Fundamental = FALSE;
FindObjectOrigins(Object) /*Submit object being researched to function */
OriginsFlag; /* Flag for origins type */
OriginObject; /* Data about origins */
If (Fundamental) then Halt; /* Halt if object is fundamental */
OriginsFlag = GetOriginsOfObject(Object, OriginsObject& ); /*Is object generated by ID or L&D?*/
Output(OriginsObject, OriginsFlag); /* Output results */
If (CompleteObject(OriginsObject)) /* Is object a complete explanation? */
Fundamental = TRUE; /* If so flag as fundamental */
FindObjectOrigins(OriginsObject); /* Find origins of origins object */