Newton’s blunder regarding planetary motion is usually given as the classic “God of the gaps’ error. Newton knew that in addition to the sun’s gravitational attraction, the planets’ affected each other. Yet the planets orbits seemed too stable to account for this, and he suggested that God sometimes intervened to smooth the orbits out. It was later discovered that the regularity of planetary motion can be accounted for based on a more rigorous application of Newton’s own equations. Thus, the problem with a “God of the gaps” argument is that as scientific understanding advances, the phenomena previously ascribed to God continually narrow. We find that it was an error to “cover up,” so to speak, our ignorance by resorting to direct acts of God to explain the data, and we should not make that same error the next time we come up against a hard problem.
ID proponents are often accused of making a similar “God of the gaps” error. They are accused of reasoning along these lines: “Living things are too complex to have evolved through natural processes; therefore God must have done it.” Is there a difference between Newton’s error and the ID position? No one will be surprised when I argue there is.
In the Principia Newton made a design inference, but his design inference was faulty as can be seen from an application of the explanatory filter.
Planetary motion should have been filtered out at the first step (Is it contingent?). In other words, Newton made a simple “snowflake error.” What is a snowflake error? Consider the following photograph taken by a National Geographic photographer.
One might look at the beautifully intricate pattern of the structure depicted in the photograph and conclude that it had been designed. But one would be wrong. This snowflake and the countless quadrillions of other snowflakes that have fallen can be accounted for through purely natural regularities. The shape of the snowflake is “determined” by natural laws. In other words, while it is somewhat complex, it lacks “contingency,” and the explanatory filter excludes it.
Newton’s intuition was the same sort of intuition that might lead someone to believe a snowflake is designed. He observed an orderly pattern in the solar system that is like the orderly pattern in the snowflake. He failed to understand that the orderly pattern he was observing was not contingent and therefore likely caused by natural law.
Now, how are the snowflake intuition and Newton’s intuition different from a biological design inference from, for example, the DNA code? The answer is that codes, like the DNA code, are always contingent. This is another way of saying there is no known natural law that can account for a semiotic code like that found in DNA.
I can already hear the objections. “But Barry, isn’t that just another instance of the same problem that Newton ran into? Doesn’t the phrase “known natural law” give the game away? After all, Newton did not know of a natural law that could account for the data, so he ascribed it to God. Isn’t the ID proponent doing the same thing? Even if the current materialist explanation of ‘random errors sorted through a fitness function’ is wrong, there may be some other natural explanation that, like Newton, we have just not yet found.”
No, the ID proponent is not doing the same thing. There are two key differences. First, natural law could, in principle, explain planetary motion. It cannot, in principle, explain a super-complex semiotic code. Second, Newton’s intuition was based on a strictly negative inference (“no known natural reason”). The ID approach is based on both a negative inference (“no known natural reason”) and a positive inference (“there is a known cause of CSI”).
To flesh this out, Newton erred because he ascribed to God the “smoothing out” of the planetary orbits when it should have been plain to him that, in principle, the application of natural laws could account for that phenomenon in the same way that natural laws can account for the intricate patterns in snowflakes, and he should have investigated further (as future scientists in fact did). ID is different for two reasons. First, ID posits that natural forces are not, in principle, capable of generating the complex specified information found in living things (more precisely, ID posits that the chances of CSI arising through natural means is so vanishingly small as to be considered impossible as a practical matter). Therefore, it is not just a matter of science not yet having found the answer. ID argues for good reason that looking for a natural cause of a super sophisticated digital semiotic code is a fool’s errand. Second, while natural forces cannot generate CSI, we do know of a cause that produces CSI routinely – intelligent agents. Indeed, intelligent agents are the ONLY known source of CSI where the provenance of the CSI has been actually observed rather than inferred. And while our digital code is not as sophisticated and elegant as that found in DNA, it is the same sort of thing. Therefore, concludes the ID proponent, intelligent design is the best explanation for the data.
The application of this type of reasoning is universally unobjectionable in certain areas of inquiry (e.g., forensics, archeology). In other words, when an archeologist finds CSI in an arrowhead and concludes the patterns were created by an Indian and not natural weathering, his colleagues do not ridicule him for making an “Indian of the gaps” argument. Only in biology is a straightforward inference based on sound premises ridiculed. Why? It is not the data. Is not the reasoning from the data. It is a Lewontinian insistence on a priori materialism – and that, ladies and gentlemen, is not science. It is a faith commitment to a metaphysical position (which is often called “religion”).