Hugh Ross is an old-earth creationist – so he thinks the earth is billions of years old, but God created all life and evolution is bunk. He is desperately trying to paint his version of creationism in a scientific light, but despite the fact that he has a PhD in astronomy, he demonstrates a dramatic and fatal lack of understanding of science. His latest attempt at dressing up creation as science is to frame it as a “testable model,” but all he succeeds in doing is demonstrating his lack of understanding of what “testable” means in the context of science. But he does offer a good example to illustrate this important point.

The most damning criticism of so-called creation science and intelligent design theory is that they are not scientific theories. This fact has been central in denying creationism access to public science classrooms, and so apologists like Hugh Ross are trying to rectify this. In the past their attempts at fooling judges informed by real scientists have all failed, and this new strategy will fail also. They fail for a simple reason – they are wrong, and when you are wrong it’s impossible to formulate a sound argument to defend yourself. A sound argument (with factually correct premises and valid logic) cannot lead to a false conclusion, so creationists must employ false premises and/or invalid logic. They have no choice (unless of course they wish to abandon creationism and embrace evolution).

Here we are dealing with a central and necessary feature of all scientific hypotheses – they must be testable. It must be possible to conduct some experiment or make some observation that will argue for or against the hypothesis, and that will help distinguish it from other competing hypotheses. Creationism/ID fails at this core challenge because it relies upon a supernatural omnipotent agent. Therefore, whatever we find in nature, and whatever the outcome of any experiment, it can simply be argued that God intended it to be so. Creationism predicts nothing, and worse it is compatible with every possible outcome – therefore it is not falsifiable, and it is not science.

Ross is now trying to argue that creationism (his model) does make predictions. But he makes a classic pseudoscientist mistake – he thinks that because he can explain things that we already know with his model, that his model therefore predicted those things. This is identical conceptually to an astrologer explaining an event that has already happened through an astrological reading, fitting one of Nostradamus’s predictions to an historical event after the fact, or spinning an elaborate conspiracy theory to explain a large number of unconnected facts.

What pseudoscientists like Ross fail to understand is that people are really good at finding patterns and correlations. It is actually quite easy to start with a conclusion and then work backwards; to take a model or theory and then fit the evidence into it. Such retrofitting can seem very compelling to the scientifically naïve. We also fall prey to psychological factors such as subjective validation to reinforce the apparent fit.

Real science does not work this way. Well, it may begin with such an analysis – can a new theory explain what we already know? But then, to count as science, it must make specific and testable predictions about what we do not know.

Here is Ross’s top 20 list of “predictions” he claims were made by his creation model:

1. transcendent creation event
2. cosmic fine-tuning
3. fine-tuning of the earth’s, solar system’s, and Milky Way Galaxy’s characteristics
4. rapidity of life’s origin
5. lack of inorganic kerogen
6. extreme biomolecular complexity
7. Cambrian explosion
8. missing horizontal branches in the fossil record
9. placement and frequency of “transitional forms” in the fossil record
10. fossil record reversal
11. frequency and extent of mass extinctions
12. recovery from mass extinctions
13. duration of time windows for different species
14. frequency, extent, and repetition of symbiosis
15. frequency, extent, and repetition of altruism
16. speciation and extinction rates
17. recent origin of humanity
18. huge biodeposits
19. Genesis’ perfect fit with the fossil record
20. molecular clock rates

Ross, however, did not predict any of these things with his model. He is just looking at what we already know and then pronouncing that, yep, his model predicted that. Some of the things on the list are not predictions at all, like “cosmic fine-tuning.” This is just a statement of the anthropic principle, which states that the laws of the universe are compatible with human life. Well of course they are. They would be if evolution were true and creation false. Some are false premises, like “Genesis’ perfect fit with the fossil record.” Talk about shoe-horning. He predicted molecular clock rates and mass extinctions? Where was that published?

Many of the items on the list are just descriptions of what we have already found – the Cambrian explosion, the fossil record, the recent origin of Homo sapiens, etc. Creation doesn’t “predict” any of these things. Couldn’t God have created man farther in the past, couldn’t God have created more advanced multicellular organisms than those that appear in the Cambrian explosion (which represents the first appearance of multi-cellular organisms in the fossil record)?

What Ross needs to do is use his model to make an actual prediction – something that is not already known but can be found out. I have yet to hear that from any creationist.

Well, actually, they make predictions all the time without intending to. They imply that there are gaps in the fossil record that cannot be filled in – until we fill them. They argue that biological designs are irreducibly complex – until we reduce them. They claim that specific evolutionary pathways to current biochemical functions are unknowable – until we discover them. Creationists like Ross have failed in all the actual predictions that they have unwittingly made. But we know what creationists do when that happens, they just ignore the evidence and move over to more dumb predictions – nudging back the goal post a far as it takes.

It is also instructive to note that while Ross is trying to portray his model as testable, it is specifically formulated to be untestable. He claims that God created life in stages. In what stages? Well, in whatever stages we find in the fossil record. New species appear over time in the fossil record not because of evolution, but because God created new species over time. God, apparently, constantly experimented with his designs and tweaked them. Is there a possible piece of evidence in the fossil record or in biology that would not be compatible with Ross’s model. Nope.

How about evolution? Creationists like to argue that evolution does not make predictions (again they demonstrate their own ignorance of science/intellectual dishonesty here). What they mean is that evolutionary theory cannot predict what organisms will evolve in the future, but that is not what is meant by a scientific prediction. You can’t predict the future of a chaotic system, like evolving life. But evolution does make predictions about what we will find when we look at life and the remnants of life.

Evolutionary theory predicted (off the top of my head, not a complete list):

1 – a medium of inheritance that retains mutations and is not diluted.
2 – the future discovery of fossil extinct species that will be morphologically transitional between extant species.
3 – biochemical relatedness will closely match morphological similarity
4 – genetic variation will show a pattern of relatedness that will follow an evolutionary pattern
5 – the fossil record will be contiguous and progressive (i.e. new designs will not appear out of nowhere or out of order)
6 – developmental biology will follow a pattern reflecting evolutionary history (still true despite being overstated in the past)
7 – the existence of derived characteristics (i.e the panda’s thumb)
8 – the existence of vestigial anatomy, even at the genetic level (i.e hen’s teeth)

What future observations could falsify evolution? Horse fossils in Cambrian deposits. Complex species without any evolutionary connections to other species. Genetic code incompatible with an evolutionary pattern of relatedness. There have been thousands of experiments (at least) in the last 150 years that could have falsified evolution. Evolutionary theory survived them all.

But I have to admit, I love creationism (in a morbid sort of way). It is the best textbook of pseudoscience available. Creationists have thoughtfully crafted numerous great examples of every aspect of pseudoscience. So I should thank Ross for this fine example – a great illustrative case of confusing pattern recognition and retrofitting for scientific predictions.