Recently, BioLogos posted the concluding Parts 3 and 4 of Robert Bishop’s review of Darwin’s Doubt. Although the discussion at BioLogos is winding down, Bishop’s review raises issues calling for a reply, because the underlying premises of his position are shared by large numbers of science and philosophy faculty at both secular and religious universities. It is likely that, left unchallenged and unexamined, these faulty premises will continue to influence the debate.
One issue in particular deserves extended comment: the standing of methodological naturalism (hereafter, MN) as a rule for scientific inquiry. Evaluating the role of MN can make for distinctly odd bedfellows. One finds theists such as Bishop, or Princeton philosopher Hans Halvorson, arguing in favor of the doctrine, whereas atheists and agnostics — e.g., Maarten Boudry and colleagues (2010, 2012), Sahotra Sarkar (2011), or Bradley Monton (2009) — arguing against it. Bishop’s use of MN in his critique of Darwin’s Doubt reveals the enormous distance between his position and that of ID theorists. Although the distance is great, communication across the divide is still possible. We hope that our response helps to clarify the ID standpoint and indicate how a more thorough analysis of MN can help in the future.
What Methodological Naturalism Is Not (and Never Was)
In Darwin’s Doubt, Stephen Meyer argues that inferences to intelligent causation, while fully warranted by the evidence of the Cambrian explosion, run afoul of the dictum of methodological naturalism (MN). As Meyer defines MN:
scientists should accept as a working assumption that all features of the natural world can be explained by material causes without recourse to purposive intelligence, mind, or conscious agency. (p. 19)
As Meyer later explains (p. 385), the fatal defect in MN is not hard to find: “if researchers refuse as a matter of principle [namely, MN] to consider the design hypothesis, they will obviously miss any evidence that happens to support it.” One cannot evaluate the evidence for or against any hypothesis that has been ruled out a priori. For this and other reasons, ID theorists regard MN as an obstacle to knowledge and hence a methodological rule that we would be better off without.
Bishop cannot see the harm in MN. Quite the reverse, in fact; in his view, “methodological naturalism is the way scientific investigation has been done since before the time of the Scientific Revolution.” The rule of MN — a reasonable and philosophically neutral boundary, as he sees it — simply represents an approach to scientific investigation that seeks to “take the biological phenomena on their own terms to understand them as they actually are.”
Now, who could disagree with that cheerful formulation of MN? No one, really — certainly not Steve Meyer or any other ID theorist. Consider: if the phenomena of the Cambrian explosion in fact implicate intelligent design, then of course we should try to explain those events, to employ Bishop’s phrase, “on their own terms…as they actually are.” Expressed that way, ID and MN would be entirely congruent, and you wouldn’t be reading this article.
But, as the rest of his review makes clear, that’s not at all what Bishop means by MN. Rather, phenomena are to be understood and explained solely via material or physical causes, come what may. Bishop categorically excludes agent causation, or causation by mind, from all biological explanation, and restricts the inference of intelligent agency to human activities. As he argues, “an intelligent agent is a presupposition external to cellular and evolutionary biology; intelligence has to be brought in from the outside.” Thus, if ID proposes agent causation to explain any biological event, it violates the well-defined boundaries of natural science — a violation, Bishop asserts, that “biologists rightly object to.” The rule of MN has been broken.
Notice, first, that Bishop completely misunderstands the basis of Meyer’s case for intelligent design. True, the intelligent agency that Meyer invokes to explain the origin of the information present in animal forms is “external” to the present operation of cells in those animals, just as the intelligence responsible for the design of a laptop computer is external to it. But that does not mean that Meyer “presupposes” that an agent “external to cellular and evolutionary biology” caused the origin of the information that arose in the Cambrian explosion of animal life. Instead, Meyer infers that a designing intelligence external to the features of cells and animals generated that information and he does so based upon our knowledge of cause and effect and information-rich structures present in living systems. Since, as he argues, intelligence or mental activity is the only known cause of the origin of large amounts of functional or specified information, especially when that information is found in a digital form, the origin of the enormous amount of specified information that arose in the Cambrian period is best explained by the activity of a designing intelligence. Intelligence is not presupposed; it is inferred based upon what we know about the cause, indeed the only known cause, of specified information.
Notice too that Bishop’s formulation of MN renders the evidence itself wholly irrelevant. If scientists must provide material or physical explanations for any phenomenon, whatever the evidence, then that is where they must remain, chained to the bench — even if the evidence strongly indicates design. This a priori formulation of MN makes a farce of empirical investigation, because the outcome of any research could never be in doubt: some material or physical cause must be affirmed as the explanation. If you don’t find one, try harder; just keep looking until you do. That is what scientists have (allegedly) always done.
Bad History Makes for Bad Philosophy
But scientists haven’t always done that, nor (as we’ll explain below) do they follow MN today, unless they are keeping their boots firmly planted on the necks of ID proponents. As Bishop’s own scholarly paper on MN shows — see for instance his endnote 36, on Robert Boyle’s view of the intelligent design of animals — leading figures in the Scientific Revolution did not see themselves as bound to strictly material or physical explanations. Isaac Newton, for example, made arguments for intelligent design in both the Opticks and the Principia. In the General Scholium to the Principia, he argued for the intelligent design of the solar system based upon the fine-tuning of the position of planets. As he stated:
Though these bodies may indeed continue in their orbits by the mere laws of gravity, yet they could by no means have at first derived the regular position of the orbits themselves from those laws. Thus, this most beautiful system of the sun, planets, and comets, could only proceed from the council and dominion of an intelligent and powerful Being.
Indeed, as an abundance of historical data affirm, MN became a putative philosophical convention of biology only after a protracted struggle. And one doesn’t have to fight for a doctrine that everyone already accepts.
Moreover, it is impossible to make sense of the Darwinian Revolution if we assert that MN governed scientific explanation for centuries before Darwin’s birth. In the Introduction to the Origin of Species (1859, p. 6), Darwin surveys the landscape of existing scientific opinion — one can almost hear him drawing in his breath with apprehension — about “the view which most naturalists entertain, and which I formerly entertained — namely, each species has been independently created.” Looking around, Darwin sees few if any prominent allies for his materialistic theory.
Were these “creationist” naturalists doing science? The answer is yes, unless one begs the question and identifies science with MN. These scientific contemporaries of Darwin, however, weren’t conducting their investigations under the strictures of MN. In 1859, intelligent design was not only a live empirical possibility, it was generally thought to be the best explanation of the origin of living forms, thereby compelling Darwin to mount his “one long argument” against it. If MN already ruled, none of that could have happened.
The pre- (and post-) Darwinian existence of good science done without the strictures of MN shows that the rule is not necessary for the discovery or systematization of empirical knowledge. At bottom, the only real motivation for holding to MN is to keep the bad guys at bay, as an all-purpose defeater for ideas like intelligent design, especially when the data may not be cooperating. And all-purpose defeaters, while handy in many a difficult moment, eventually reveal themselves to be the cheats they are. Who wants to play in a soccer league where one team always wins, whatever the score on the field?
In short, MN never was the way science was always done. Science — empirical inquiry — pretty much takes care of itself, as long as curiosity, the evidence, and testability are given half a chance.
A Rule Honored in the Breach
Nor does MN govern today, except in official contexts (such as federal courts or statements from national science organizations) where definitions are required for demarcation purposes, to determine whether any idea passes muster as “science.” Above, we noted that MN is a putative rule for biology — “putative” (that is, supposed but not actual) insofar as the content and practice of the science exhibit the widespread use of theological concepts and categories.
It is a little-remarked but nonetheless deeply significant irony that evolutionary biology is the most theologically entangled science going. Open a book like Jerry Coyne’s Why Evolution is True (2009) or John Avise’s Inside the Human Genome (2010), and the theology leaps off the page. A wise creator, say Coyne, Avise, and many other evolutionary biologists, would not have made this or that structure; therefore, the structure evolved by undirected processes. Coyne and Avise, like many other evolutionary theorists going back to Darwin himself, make numerous “God-wouldn’t-have-done-it-that-way” arguments, thus predicating their arguments for the creative power of natural selection and random mutation on implicit theological assumptions about the character of God and what such an agent (if He existed) would or would not be likely to do.
Now, the usual response to criticism of this type of theologically grounded argument claims that evolutionary biology has been forced into its extra-scientific entanglements by stubborn religious opposition to the theory of evolution. The creationists started the fight, this view holds, so it’s not surprising that evolutionary biologists need to push back using the terms and categories of creationists themselves.
Authors such as Coyne or Avise, however, hold that the apparently imperfect or suboptimal features of organisms provide objective evidence for undirected evolution. Presumably the standing of these features as evidence for evolution is not conditional on the presence, somewhere in the room, of a creationist or two saying otherwise. Put another way, Coyne and Avise would offer the same features as evidence to a science seminar populated by intelligent beings (aliens, let’s say) with no concept of God or theology. Scientific evidence doesn’t change its epistemic complexion depending on the audience or rhetorical context at hand.
If so — and Coyne has consistently defended the theological propositions in his book as fully empirical — then the very content of evolutionary theory rests on theological assumptions, borrowed or not. Philosopher of science Steven Dilley has carefully analyzed this situation with respect to one of the most famous texts in 20th-century biology, Theodosius Dobzhansky’s essay “Nothing in biology makes sense except in the light of evolution” (1973).
Although its title is widely cited as an aphorism, the text of Dobzhansky’s essay is rarely read. It is, in fact, a theological treatise. As Dilley (2013, p. 774) observes:
Strikingly, all seven of Dobzhansky’s arguments hinge upon claims about God’s nature, actions, purposes, or duties. In fact, without God-talk, the geneticist’s arguments for evolution are logically invalid. In short, theology is essential to Dobzhansky’s arguments.
Eventually this reality will be grasped by evolutionary biologists themselves, with inescapable consequences for the validity of MN. If Dobzhansky’s essay genuinely belongs to the explanatory patrimony of evolutionary biology, MN is not only descriptively false (as history), but proscriptively unsound — we shouldn’t follow the rule even if we could. MN is a bad philosophy of science on all counts.
In Signature in the Cell and Darwin‘s Doubt, Steve Meyer himself provides an exhaustive refutation of those who would enshrine MN as a normative convention for science. He shows that attempts to justify MN using various demarcation criteria — such as observability, replicability and testability — have failed. He also shows that, in any case, the theory of intelligent design is testable in at least three interrelated ways.
First, he shows that, like other scientific theories concerned with explaining events in the remote past, intelligent design is testable by comparing its explanatory power with that of competing theories.
Second, Meyer shows that ID like other historical scientific theories is tested against our knowledge of the cause-and-effect structure of the world. Following Darwin himself and the geologist Charles Lyell, Meyer shows that scientific historical theories provide adequate explanations when they cite causes that are known to produce the effects in question. These:
considerations of causal adequacy provide an experience-based criterion by which to test — accept, reject, or prefer — competing historical scientific theories. When such theories cite causes that are known to produce the effect in question, they meet the test of causal adequacy; when they fail to cite such causes, they fail to meet this test.
Third, he shows that intelligent design makes a number of specific predictions that differ from predictions made by the materialistic theories of evolution against which ID competes (see his Appendix A in Signature in the Cell for a discussion of ten such predictions). These predictions not only provide another way to test the theory of intelligent design, they have, in several striking cases, already served “to confirm the design hypothesis rather than its competitors.”
For readers unfamiliar with Meyer’s critique of the use of MN in science, we recommend Chapters 18 and 19 in Signature in the Cell and Chapter 19 in Darwin‘s Doubt where he provides a thorough refutation of Bishop’s case for accepting methodological naturalism as a normative rule for science. Indeed, in asserting MN as normative for science, Bishop really doesn’t engage Meyer’s earlier refutation of the need for MN, let alone refute Meyer’s arguments against allowing MN to stand as a rule of method.
A Philosophy of Science No One Needs
MN does nothing for science that science cannot do for itself. Seen in the bright light of day, MN turns out to be little more than an all-purpose defeater for unwelcome ideas — another “Press Button in Case of Emergency” doctrine of the sort that brings disrepute on the philosophy of science. If ID is untestable or empirically empty, as its critics claim, we won’t need MN to establish that. ID will fail on its own terms.
If ID is testable, however, as Meyer convincingly argues, then MN can only be a philosophical obstacle shoved in the way of the empirical possibility of design, for reasons having nothing to do with open-ended scientific inquiry. Either way, MN is a pointless rule.
And science will be better off without a rule that no one needs, that few actually obey, and that limits the freedom of scientists to follow the evidence wherever it leads. On this final point, let’s give Meyer (2009, p. 437) himself the last word:
[A]llowing methodological naturalism to function as an absolute “ground rule” of method for all of science would have a deleterious effect on the practice of certain scientific disciplines, especially the historical sciences. In origin-of-life research, for example, methodological naturalism artificially restricts inquiry and prevents scientists from exploring and examining some hypotheses that might provide the most likely, best, or causally adequate explanations. To be a truth- seeking endeavor, the question that origin-of-life research must address is not, “Which materialistic scenario seems most adequate?” but rather, “What actually caused life to arise on earth?” Clearly, one possible answer to that latter question is this: “Life was designed by an intelligent agent that existed before the advent of humans.” If one accepts methodological naturalism as normative, however, scientists may never consider this possibly true hypothesis. Such an exclusionary logic diminishes the significance of any claim of theoretical superiority for any remaining hypothesis and raises the possibility that the best “scientific” explanation (according to methodological naturalism) may not be the best in fact.
Avise, John. 2010. Inside the Human Genome: A Case for Non-Intelligent Design. Oxford: Oxford University Press.
Boudry, Maarten et al. 2010. How Not to Attack Intelligent Design Creationism: Philosophical Misconceptions About Methodological Naturalism. Foundations of Science 15:227-44.
Boudry, Maarten et al. 2012. Grist to the Mill of Anti-evolutionism: The Failed Strategy of Ruling the Supernatural Out of Science by Philosophical Fiat. Science & Education 21:1151-65.
Coyne, Jerry. 2009. Why Evolution is True. New York: Viking.
Darwin, Charles. 1859. On the Origin of Species. London: John Murray.
Dilley, Steve. 2013. Nothing in biology makes sense except in the light of theology? Studies in History and Philosophy of Biological and Biomedical Sciences 44:774-86.
Dobzhansky, Theodosius. 1973. Nothing in biology makes sense except in the light of evolution. American Biology Teacher March, 125-9.
Meyer, Stephen. 2009. Signature in the Cell. New York: HarperOne.
Monton, Bradley. 2009. Seeking God in Science: An Atheist Defends Intelligent Design. Peterborough, ON: Broadview Press.
Sarkar, Sahotra. 2011. The science question in intelligent design. Synthese 178:291-305.