To understand the problems with the naturalism-only mandate, one must understand just what is being mandated. Some evolutionists appear to make the bare assertion that only rigidly naturalistic explanations can qualify as genuine science. As wiith most bare assertions this one doesn't work very well. For instance, it is easily countered by the bare assertion that rigidly naturalistic explanations are not required for genuine science (so there!). Some design critics, such as Taner Edis, warn evolutionists against making such an assertion for it does not allow for design theory to be rejected by according to the evidence.
But more thoughtful evolutionists, such as Joe Felsenstein, provide an underlying reason for the mandate. Felsenstein explains:
what he has just done is to admit that the hypothesis of a Designer is not science, as it predicts every possible result. If you predict every possible outcome, the ones that are seen and the ones that are not, then you have not predicted anything! ...
If there are none, then the Design he speaks of is an infinitely flexible hypothesis that predicts nothing, and thus is really not a scientific hypothesis at all … which is what I originally said.
In other words, in order to qualify as legitimate science a theory must distinguish between different outcomes. Naturalism is needed because otherwise each outcome is equally probable and the theory is not true science.
Deciding what does and does not qualify as legitimate science is notoriously difficult. There seem to be exceptions to every rule. But perhaps Felsenstein's criterion is reasonable. Shouldn't a scientific theory say at least something about the probabilities of what we might observe in the data?
Evolutionists say this test shows their theory to be the perfect model of true science. Consider, for example, the recurrent laryngeal nerve. As evolutionist Jerry Coyne explains:
The reason why the recurrent laryngeal nerve, for example, makes a big detour around the aorta before attaching to the larynx is perfectly understandable by evolution (the nerve and artery used to line up, but the artery evolved backwards, constraining the nerve to move with it)
In other words, historical contingencies and constraints play an important role in evolutionary explanations. Today's designs are not independent of their history. When we see obvious similarities between species evolutionists make the historical connections.
But special creation and design theory have no such basis from which to draw. Designs are independent. God could have created the species in any fashion, so there is no way to distinguish outcomes. Are not all designs equiprobable? As Coyne explains, the recurrent laryngeal nerve
makes no sense under the idea of special creation ... No form of creationism/intelligent design can explain these imperfections
Evolutionists use such examples to argue that their theory is scientific whereas creation and design are not. Now let's look at the facts.
These claims of evolutionists are false and hypocritical. Evolution does not distinguish different outcomes for the recurrent laryngeal nerve. When incredibly fantastic, mind boggling designs are found, they ascribe the wonder to natural selection. When similarities between species are found, they ascribe them to historical contingencies.
And evolutionists have a seemingly never ending list of mechanisms they use to explain everything in between. Whatever we find in biology, evolutionists say it must have evolved. Their predictions and expectations are often falsified and they have to patch their theory repeatedly. And there is no distinction between a new, fantastic design and a repeated design--both are equiprobable under evolution.
If a new, fantastic design appears such as the trilobite eye, then evolutionists ascribe it to natural selection. If similar designs are found in different species, then it is ascribed to common descent. If later cousin species are found to lack the design, then common descent can be dropped as an explanation and the design can be said to have evolved independently. The evolutionary explanation is extremely flexible.
If distinguishing between outcomes is the hallmark of true science, then evolution is the theory that doesn't qualify.
As for design theory, while it does not rule out historical contingency as a possible explanation it, in any case, looks for a rationale for what we find in nature. The more probable outcomes are functional designs that require planning, foresight, mechanisms, and so forth. Design theory tries to figure out how nature works rather than viewing it as a fluke and accident of history. And the history of science is squarely on its side. Over and over evolutionary expectations that nature is a fluke are overruled by the evidence. Over and over we find function and fantastic designs which make no sense under evolution.
How do you want to do science? Do you want to constrain every explanation to an improbable theory, or do you want to figure out how nature works without a priori theoretical constraint?