1 Publish a mathematically rigorous definition of CSIWe challenge Dembski to publish a mathematically rigorous definition of "complex specified information" (CSI) and a proof of the Law of Conservation of Information in a peer-reviewed journal devoted to information theory or statistical inference, taking into account the criticisms in Elsberry and Shallit (2003) and elsewhere.
2 Provide real evidence for CSI claimsHere is a brief catalog of some of the things Dembski has claimed exhibit CSI or "specified complexity":
- 16-digit numbers on VISA cards (Dembski 1999: 159);
- Phone numbers (Dembski 1999: 159);
- "All the numbers on our bills, credit slips and purchase orders" (Dembski 1999: 160);
- The "sequence corresponding to a Shakespearean sonnet" (Dembski 2002: xiii);
- Arthur Rubinstein's performance of Liszt's "Hungarian Rhapsody" (Dembski 2002: 95);
- "Most human artifacts, from Shakespearean sonnets to Durer woodcuts to Cray supercomputers" (Dembski 2002: 207);
- Scrabble pieces spelling words (Dembski 2002: 172-3);
- DNA (Dembski 2002: 151);
- Error-counting function in an evolution simulation (Dembski 2002: 217);
- A fitness measure that gauges degree of catalytic function (Dembski 2002: 221);
- The "fitness function that prescribes optimal antenna performance" (Dembski 2002: 221);
- "Coordination of local fitness functions" (Dembski 2002: 222);
- What "anthropic principles" explain in fine-tuning arguments (Dembski 2002: 144);
- "Fine-tuning of cosmological constants" (Dembski 2002: xiii);
- What David Bohm's "quantum potentials" extract in the way of "active information" (Dembski 2002: 144); and
- "The key feature of life that needs to be explained" (Dembski 2002: 180).
3 Apply CSI to identify human agency where it is currently not knownThus far CSI has only been used to assert design in two classes of phenomena: those for which human intervention is known through other means, and those for which a precise step-by-step causal history is lacking. We challenge Dembski or other intelligent design advocates to identify, through CSI, some physical artifact - currently not known to be the product of human design - as an artifact constructed by humans. After this prediction through CSI, provide confirming evidence for this conclusion, independent of Dembskian principles.
Along similar lines, apply CSI to identify a suspicious death, currently thought to be from natural causes, as foul play. Furthermore, also provide confirming evidence for this conclusion, independent of Dembskian principles.
We note that Dembski himself has stressed the importance of independent evidence (Dembski 2002: 91).
4 Distinguish between chance and design in archaeoastronomyThe Anasazi, or ancestral Puebloans, occupied what is now the southwestern United States from about 600 to 1300 ce. Several of their buildings - including those at Chaco Canyon, Hovenweep National Monument, and Chimney Rock - have been interpreted as astronomical observatories, with alignments correlated to solstices, equinoxes, lunar standstills and other astronomical events (Malville and Putnam 1989). Using the techniques of The Design Inference, provide a rigorous mathematical analysis of the evidence, determining whether these alignments are due to chance or human design.
Similar challenges exist for the claimed astronomical alignments at Stonehenge (Hawkins 1965; North 1996) and Nabta (Malville and others 1998), and the enigmatic drawings at Nazca in southern Peru. Which of the proposed alignments were designed, and which are pure coincidence?
5 Apply CSI to archaeologyAnother interesting question about the Anasazi is the presence of large numbers of pottery shards at certain ruins. Some archeologists have interpreted the number of these shards as exceeding the amount that could be expected through accidents. Use CSI to determine if the pots were broken through accident, or human intent (possibly in support of some religious ritual).
Archeologists have developed methods for determining whether broken flints cracked due to human intervention or not (Cole and others 1978). Attempt to re-derive this classification, or prove it wrong, using the methods of CSI.
Provide a useful means of applying CSI to distinguish early stone tools from rocks with random impact marks.
6 Provide a more detailed account of CSI in biologyProduce a workbook of examples using the explanatory filter, applied to a progressive series of biological phenomena, including allelic substitution of a point mutation. There are two issues to be addressed by this exercise. The first is that a series of fully worked-out examples will demonstrate the feasibility of applying CSI to biological problems. The second is to show that assigning small-scale changes to "chance" and "design" only is indicated for much larger-scale changes or systems already noted as having the attribute of "irreducible complexity". It is our expectation that application of the "explanatory filter" to a wide range of biological examples will, in fact, demonstrate that "design" will be invoked for all but a small fraction of phenomena, and that most biologists would find that many of these classifications are "false positive" attributions of "design".
7 Use CSI to classify the complexity of animal communicationAs mentioned in Elsberry and Shallit (2003: 9), many birds exhibit complex songs. We challenge Dembski or other design advocates to produce a rigorous account of the CSI in a variety of bird songs, producing explicit numerical estimates for the number of bits of CSI.
Similar challenges can be issued for dolphin vocalizations, as in providing a definitive test of the "signature whistle" hypothesis (Caldwell and others 1990), and estimation of information of a dolphin biosonar click (to be compared to the information measure suggested by Kamminga ).
8 Animal cognitionApply CSI to resolve issues in animal cognition and language use by non-human animals. Some of these outstanding issues include studies of mirror self-recognition (Gallup 1970, 1982) and artificial language understanding in chimpanzees (Savage-Rumbaugh 1993), dolphins (Herman and others 1993), and parrots (Pepperberg 1993). We note the use of examples in Dembski's work involving a laboratory rat traversing a maze as an indication of the applicability of CSI to animal cognition (Dembski 1998, 1999, 2002).
These, we feel, are reasonable challenges that Dembski, or others who wish to pursue "intelligent design" as a scientific research paradigm, ought to be eager to meet.
AcknowledgmentsWe are grateful to Anna Lubiw, Ian Musgrave, John Wilkins, Erik Tellgren, and Paul Vitányi, who read a preliminary version of the longer paper from which this article is derived (Elsberry and Shallit 2003) and gave us many useful comments. We owe a large debt to Richard Wein, whose original ideas have had significant impact on our thinking.
[Adapted with permission from section 12 of Wesley Elsberry and Jeffrey Shallit, "Information theory, evolutionary computation, and Dembski's 'complex specified information'". Available on-line at http://www.talkreason.org/articles/eandsdembski.pdf.]