Return of the God Hypothesis: Breakthroughs in Physics, Cosmology, and Biology Seeking Evidence for the Existence of God

by Stephen C. Meyer

Scientists in many fields recognize the connection between intelligence and information and make inferences accordingly. Archaeologists assume that a scribe produced the inscriptions on the Rosetta Stone. The search for extraterrestrial intelligence (SETI) presupposes that specified information imbedded in electromagnetic signals coming from space would indicate an intelligent source.52 As yet radio astronomers have not found any such information-bearing signals. But closer to home, molecular biologists have identified specified information-rich sequences and systems in the cell, suggesting, by the same logic, the past existence of an intelligent cause for those effects. Our uniform experience affirms that specified or functional information—whether inscribed in hieroglyphics, written in a book, encoded in a radio signal, or produced in an RNA-world “ribozyme-engineering” experiment—always arises from an intelligent source, from a mind, not a strictly material process.

Late nineteenth-century scientists knew nothing, of course, about the importance of information to living systems. They assumed that the universe consisted of two fundamental entities: matter and energy. But during the 1950s and 1960s molecular biologists discovered a third fundamental entity at the foundation of life—information. Moreover, the functional digital information in the “machine code of the genes,” as Dawkins put it, does not seem—based upon our experience—“to be the kind of quality we should expect to observe” if there was “no design, no purpose . . . nothing but blind, pitiless indifference”53 at work in the origin of life.

For the mathematically minded scientists at Wistar, doubts about the creative power of the mechanism of random mutation and natural selection stemmed from the elucidation of the nature of genetic information and the confirmation of Francis Crick’s sequence hypothesis during the early 1960s. The discovery that DNA stores information as a four-character digital code raised questions about the efficacy of random mutational changes in producing such information—or at least enough of it to produce a novel protein structure and therefore any major innovation during the history of life. Murray Eden (Fig. 10.3), one of the MIT professors who convened the event, emphasized that in all computer codes and written texts specificity of sequence determines function. Thus, random changes in sequence consistently degrade function or meaning. Indeed, no computer programmer wants random changes introduced into a program that he or she has written. Such changes will inevitably degrade and ultimately destroy the function of the existing program long before a new program would emerge through such a process. As Eden explained, “No currently existing formal language can tolerate random changes in the symbol sequences which express its sentences. Meaning is almost invariably destroyed.”9 He suspected that the need for specificity in the arrangement of DNA bases would also render any random mutational “search” for a new functional gene or protein inevitably unsuccessful as well.

Many now repeat an old aphorism affirming that mutation and natural selection can account for “the survival of the fittest, but not the arrival of the fittest”13—that is, small-scale variations, but not large-scale innovations in biological form.

According to neo-Darwinian theory, new genetic information arises as random mutations occur in the DNA. “Random” means that mutations occur without respect to the functional needs of the organism—mutations have no inherent directionality. Nevertheless, natural selection can only “select” what random mutations first generate.14 And for the evolutionary process to produce new forms of life, random mutations must first generate—at the very least—new genetic information for building novel proteins. And that, Denton told me, was the problem. When it comes to producing new genetic information, the neo-Darwinian mechanism, with its reliance on random mutations, faces a needle-in-a-haystack dilemma, or what mathematicians call a “combinatorial problem.”

Denton explained why the mechanism of random mutation and natural selection also faces a combinatorial problem.15 He did so by drawing an analogy to English text. As Denton noted, linguists have estimated that for every meaningful sequence of English characters 12 letters long there are one hundred trillion (i.e., 100,000,000,000,000, or 1014) corresponding gibberish sequences of the same length—effectively a lock with fourteen dials and ten digits but only one combination.

In English there are vastly more ways “to go wrong than to go right”—that is, for any sequence of any given length, there are more combinations of English letters that will not produce a meaningful phrase or sentence than combinations of those same 26 letters that will generate a meaningful sentence. Indeed, the number of nonfunctional gibberish sequences dwarfs the number of functional combinations. Consequently, random changes in letters are overwhelmingly more likely to “find the gibberish,” or degrade meaning, than to generate a new meaningful sentence, especially as the number of changes to the original meaningful sequence increases. Moreover, as the length of the required phrase or sentence grows, the number of possible letter sequences of that length grows exponentially, and grows much faster than the number of possible meaningful sequences, so that the probability of finding a functional sequence via a random search diminishes precipitously with necessary sequence length. Denton noted that whereas for every meaningful sequence of English letters 12 letters long there are one hundred trillion (or 1014) corresponding gibberish sequences, for every meaningful sequence of English letters 100 letters long there are 10100 corresponding gibberish sequences, an unimaginably large number. Mathematician David Berlinski (Fig. 10.6) has dubbed this the problem of “combinatorial inflation” in his seminal 1996 essay titled “The Deniable Darwin.”16

random mutational changes were overwhelmingly more likely to degrade biological function than to generate a new functional gene or protein.

Even a relatively short protein of, say, 150 amino acids represents one sequence among an astronomically large number of other possible sequence combinations—approximately 10195. That is an enormous number, the digit 1 followed by 195 zeroes. Intuitively, this suggests that the odds of finding even a single functional sequence—a working gene or protein—as the result of random genetic mutations may be prohibitively small, even taking into account the time available to the evolutionary process.

it isn’t just the total number of possible combinations in the amino-acid sequence space that determines the difficulty of a random search for a new protein structure. Ultimately, it’s the ratio of functional to nonfunctional sequences that determines the difficulty.

for every one DNA sequence that generates a short functional protein fold of just 150 amino acids in length, there are 1077 nonfunctional combinations—combinations

the difficulty of a mutational search for a new gene or novel protein fold is equivalent to the difficulty of searching for just one combination on a lock with ten digits on each of seventy-seven dials (Fig. 10.10)!

As I show in more detail in Darwin’s Doubt, a long and painstaking search for such a cause, by some of the best minds in evolutionary biology, has failed to turn up a cause capable of producing the information necessary for genuine innovation in the history of life. Yet intelligent agents routinely produce vast amounts of specified information in order to communicate and to build a variety of new structures. Thus, only intelligent design meets the requirement of causal adequacy. In other words, our uniform experience of cause and effect shows that intelligent design is the only known cause of the origin of large amounts of functionally specified information.

In biology, survival depends upon maintaining present function. Natural selection, therefore, cannot look forward or devise plans that anticipate future needs or desirable outcomes.

What unguided evolutionary mechanisms lack, intelligent design—purposive, goal-directed selection—provides. Rational agents can arrange matter and symbols with distant goals in mind. They also routinely solve problems of combinatorial inflation. In using language, the human mind routinely “finds” or generates highly improbable linguistic sequences to convey a preconceived idea. In the process of thought, functional objectives precede and constrain the selection of words, sounds, and symbols to generate functional (and meaningful) sequences from a vast ensemble of meaningless alternatives.27 Similarly, the construction of complex technological products, such as bridges, circuit boards, engines, and software, results from goal-directed constraints on the possible arrangements of matter or symbols.28

repeated abrupt appearances of new biological form and information in the history of life are not at all what we should expect to observe if a purely materialistic evolutionary process lacking “design” and “purpose” was at work.29 Instead, if a purposive intelligence had acted periodically during the history of life on earth, we might well expect— given our experience of intelligent agents generating information—to find evidence of episodic bursts of new information in the biosphere.

different answers to this ultimate, or “prime-reality,” question. “Naturalism” (or materialism) views matter and energy and the laws of nature as the prime realities. “Pantheism” asserts an impersonal deity present in matter and energy as the prime reality. “Theism” affirms a personal, intelligent, transcendent God who also acts within the creation. And “deism” affirms a personal, transcendent, intelligent God who does not act within the created order after its initial origin (Fig. 11.3). These four worldviews represent four possible ways of answering three basic questions about ultimate reality: Does God exist? If so, is God personal or impersonal? If personal, does God act only at the beginning of the universe or also after it within the created order?

FIGURE 11.3 Philosophers recognize several main worldviews with different answers to the “prime reality” question. Theism affirms a personal, intelligent, transcendent God who also acts within the creation. Deism asserts a personal, transcendent, intelligent God who does not act within the created order after its initial origin. Naturalism (or materialism) affirms matter and energy and the laws of nature as the prime realities. Pantheism asserts an impersonal deity present in matter and energy as the prime reality. In these diagrams portraying these four great systems of thought, the circles represent the physical universe, the drawings inside the circle depict various living and nonliving entities within the universe, the pendulum represents the laws of nature, and “the big G” represents God. Notice that in Theism, God is depicted as separate from but also active in the universe; in Deism, God is depicted as separate from but not active in the universe; in Naturalism or Materialism, God is portrayed as nonexistent, and in Pantheism, God is shown as present in, or “co-extensive” with, every aspect of the material universe but not existing in any way separate from it.

in the spirit of the inquiry that Dawkins has advanced, I will evaluate whether the discoveries in cosmology, physics, and biology discussed in the preceding chapters might provide a good reason for believing in God, or even a better reason for believing in God than in naturalism or materialism, for example. As I’ve already suggested, recent scientific discoveries concerning biological and cosmological origins might be “just what we should expect” if a transcendent and intelligent designer acted to produce life and the universe. Since these same observations of nature may not be what we would expect assuming scientific materialism (or other nontheistic worldviews), the God hypothesis could in principle provide the best metaphysical explanation of the relevant scientific evidence. In other words, even if we can’t prove God’s existence with absolute certainty, we may have better reasons for affirming a theistic view of the “nature of nature” and the “prime reality” than for affirming other metaphysical systems of thought. The next three chapters explore this possibility in light of the evidence already presented.

the science Shermer admired had done a great job of explaining how the universe and life operate, but that it had not offered adequate materialistic explanations for the origin of life, mind, or the universe. But that suggested to me that materialism as a worldview lacked significant explanatory power. Is it possible, instead, that a theistic worldview—a God hypothesis—might help explain what materialism or naturalism has not—or perhaps cannot? If so, would the explanatory power of such a God hypothesis provide a reason for favoring theism over other competing metaphysical hypotheses?

If sometime in the finite past, either the curvature of space reached an infinite and/or the radius and spatial volume of the universe collapsed to zero units, then at that point there would be no space and no place for matter and energy to reside. Consequently, the possibility of a materialistic explanation would also evaporate, since at that point neither material particles nor energy fields would exist. Indeed, since matter and energy cannot exist until space (and probably time) begins to exist, a materialistic explanation involving either material particles or energy fields—before space and time existed—makes no sense. As I used to tell my students, “If you extrapolate back all the way to a singularity, you eventually reach a point where there is no matter left to do the causing.”

Sean Carroll, one of the most prominent proponents of naturalism, has acknowledged that naturalism has not explained the origin of the universe, precisely because it can offer no cause capable of producing it. He suggests, however, that the origin of the universe does not necessarily require a causal explanation; it might “just be.” Nevertheless, because the evidence indicates that the universe has not existed infinitely, but instead began to exist, it would seem to require—by the principles of causality and sufficient reason—a cause. Saying otherwise undermines one of the basic presuppositions of scientific investigation and indeed of reason itself, namely, that “whatever begins to exist must have a cause.” Philosopher William Lane Craig (Fig. 12.4) brought this point home forcefully in an interview I conducted with him several years ago. He pointed out that saying the universe might have popped into existence uncaused for no reason at all is no different from saying that a freight train or a Bengal tiger might have done so. Though, he pointed out, there was no way to disprove such possibilities, reason and the scientific investigation of the world depend upon the opposite assumption—that all such material events do have causes. He said those who deny this principle and who yet also decry the reasonableness of theism “should, therefore, forever be silenced.”18

any entity capable of explaining the origin of the universe, to which these indicators attest, must transcend the space and time, matter and energy of the universe. Naturalism fails to explain the origin of the universe because it denies the existence of any entity external to nature, but theism postulates the existence of precisely such a transcendent entity as a cause.

An infinitely existing primeval atom would have afforded an infinite number of opportunities for such a change to occur, and any one of these opportunities could have occurred an infinitely long time ago. Why then would a sudden change occur only a finite time ago if there had been an infinite number of opportunities for such a change of state to occur over an infinite time? As physicists Anthony Aguirre and John Kehayias have noted, “It is very difficult to devise a system—especially a quantum one—that does nothing ‘forever,’ then evolves. A truly stationary or periodic quantum state, which would last forever, would never evolve, whereas one with any instability will not endure for an indefinite time.”25 Invoking personal agent causation again resolves these puzzles.

to achieve various goals or objectives, intelligent agents typically must choose among (or constrain) a range of possibilities to actualize an otherwise improbable outcome. The act of choosing among options is what is meant by “fine tuning,” and it is precisely what intelligent agents with free will do. (The Latin roots, inter lego, of the English word “intelligence” mean “to choose between.”) Consequently, we have good reason to expect that, if an intelligent agent acted in the past to design the universe to be a life-sustaining place, we ought to expect to observe something like the cosmological fine tuning.4

Expelled: No Intelligence Allowed, featuring the actor, lawyer, and economist Ben Stein.

Simply asserting that life arose somewhere else out in the cosmos does not explain how the information necessary to build the first life, let alone the first intelligent life, could have arisen. It merely pushes the explanatory challenge farther back in time and out into space. Indeed, positing another form of preexisting life only presupposes the existence of the very thing that all theories of the origin of life must explain and have yet to explain—the origin of functional biological information. Beyond that, panspermia certainly does not explain the origin of the cosmological fine tuning. Since the fine tuning of the laws and constants of physics and the initial conditions of the universe date from the very origin of the universe itself, if intelligent design best explains the fine tuning, then the designing intelligence responsible for the fine tuning must have had the capability of setting the fine-tuning parameters and initial conditions from the moment of creation. Yet, clearly, no intelligent being within the cosmos that arose after the beginning of the cosmos could be responsible for the fine tuning of the laws and constants of physics that made its existence and evolution possible.

even if we concede as a logical possibility that an immanent intelligence might explain the origin of life on earth (since such an entity could possibly precede it), the panspermia hypothesis does not explain either the ultimate origin of life in the universe or the fine tuning of the universe—to say nothing of the origin of the universe itself. Instead, if intelligent design best explains the fine tuning of the universe, then the kind of intelligence necessary to explain the fine tuning of the universe must in some way preexist or exist independently of the material universe. Indeed, any designing intelligence responsible for the cosmological fine tuning must have had the capability of setting the parameters and initial conditions from the beginning. Since both theism and deism conceive of God as having an existence independent of the material universe—either in a timeless eternal realm or in another realm of time independent of the time in our universe—both can account for (a) the origin of the universe in time (i.e., at a beginning) and (b) the fine tuning of the universe from the beginning of time. In other words, since both theism and deism posit the prior (either ontological or temporal) existence of a transcendent intelligent agent, the creative and causal act of such an agent in choosing to design the universe with a specific suite of life-permitting parameters would explain the origin of the fine tuning from the beginning of the universe.

Unfortunately for proponents of naturalism, the laws of physics do not, and cannot, explain either the fine tuning of the constants of proportionality within the laws of physics or the fine tuning of the initial conditions of the universe. Indeed, the fundamental laws of physics cannot, in principle, explain why the constants of proportionality have the values that they do. As I explained in Chapters 7 and 8, (1) the structure of the laws allows them to have other values and (2) the specific values of the constants represent features of the laws themselves, not aspects of nature that the laws could conceivably explain. Similarly, the laws of physics do not explain why the universe had the precise set of initial conditions it did. The laws apply to those material conditions, and the laws must presuppose them to describe the universe accurately or to make definite predictions about what would have happened after the beginning of the universe. But the laws do not explain their origin.

naturalism, with its commitment to explaining everything by reference to the fundamental laws of nature alone, cannot explain the fine tuning of the universe—at least not without presupposing exquisite prior unexplained fine tuning of other contingent parameters.

Appealing to some as yet undiscovered law to explain the fine tuning of the physical constants seems implausible for another related reason. Natural laws by definition describe phenomena that conform to regular or repetitive patterns. Yet the idiosyncratic values of the different physical constants and initial conditions constitute a highly irregular and nonrepetitive ensemble. It seems unlikely, therefore, that any more fundamental laws could explain why all the fundamental constants have exactly the values they do—why, for example, the permittivity constant in Coulomb’s law should have the value 8.85 x 10−12 Coulombs2 per Newton-meter, and the electron mass should have exactly the value of 9.11 x 10−31 kilograms, and the electron charge to mass ratio is exactly 1.76 x 1011 Coulombs per kilogram, and the fine-structure constant should have the value .00729, and the Higgs field vacuum expectation has the value of 246 GeV.17 These constants specify a highly irregular array of values that describe either ratios between completely different types of quantities or completely different quantities (e.g., mass, charge, and energy) or quantities without units at all. As a group, they do not exhibit the kind of regular pattern that is at all likely to be subsumed under a single physical law. Considerations of causal adequacy reinforce the above arguments. In our experience, we have often observed intelligent agents producing finely tuned systems, whether in a digital computer, an i...

scientific naturalists have typically attempted to explain the fine tuning not by positing a known law or process, but instead by positing alternate explanations that attempt to reduce the surprise associated with the discovery of the fine tuning. Both the weak anthropic principle and the strong anthropic principle provide good examples of this strategy. Nevertheless, as we saw in Chapter 8, both of...

suffice to say that basic naturalism, or materialism, clearly fails to explain the fine tuning, because none of the popular proposals for doing so—neither brute chance, nor the laws of physics, nor the weak and strong anthropic principles—meets the test of causal adequacy. Indeed, since our experience affirms that finely tuned systems arise from intelligent activity, and since naturalism denies the existence of an intelligent agent before the beginning of the universe, basic naturalism lacks recourse to an entity with the causal powers to produce the effect in question. By contrast, theism and deism affirm the existence of a transcendent intelligent agent prior (either ontologically or temporally) to the beginning of the universe. Thus, theism and deism posit an agent with causal powers—the “right skill set”—to produce the fine tuning of the universe from its beginning, whereas naturalism does not.

naturalists might simply treat the values of the constants as givens. Yet even if one defines the laws as including the constants, the laws certainly do not contain information about the initial conditions of the universe. Yet those conditions also exhibit extreme fine tuning that the laws don’t explain. Moreover, simply defining constants as part of the laws of physics still does not explain why the physical constants themselves have the exact values they do. And since nothing in fundamental physical theory explains why the constants have these precise values, the values still require explanation—an explanation that the laws of physics can’t provide, since those laws can’t explain themselves (or their own contingent features).

Richard Dawkins’s claim that the universe has just the properties that we should expect if “there was no purpose, no design . . . nothing but blind, pitiless indifference.”19 In fact, however, the hypothesis of materialism does not lead naturally to the expectation of a finely tuned universe capable of sustaining life. Since in our experience fine tuning results from intelligent agency, and since naturalism denies the existence of any intelligent agent preexisting the universe, philosophical naturalists should not expect to observe a universe in which life depends upon exquisite fine tuning. Instead, they should expect a universe in which all phenomena can be explained by reference to the fundamental laws of physics. But, as we have seen, those laws themselves do not explain either the fine tuning of the initial conditions of the universe or the contingent features of the physical laws

the range of possible values and initial conditions compatible with the basic laws of physics allows for vastly and quantifiably more sterile universes than life-permitting universes. Since naturalism affirms the laws of physics as the ultimate realities and explanatory principles, the probability of a sterile universe given naturalism is vastly and quantifiably greater than that of a life-conducive universe. Thus, we have a quantifiably greater reason to expect a sterile universe given naturalism.

Indeed, if we just consider the fine tuning of the initial entropy, we can calculate that, given the fundamental laws of physics—that is, given naturalism—we should expect vastly fewer life-conducive universes than sterile universes by a factor of 1 to 101098. (See n. 11 in Chapter 8.)

since we have a greater reason to expect an exquisitely finely tuned universe capable of hosting life given theism than naturalism, the observation of a finely tuned life-permitting universe provides greater support for theism.

Effects result from causes distinct from themselves.25 Since the effects in question include the beginning of the universe as a whole and the fine tuning of the whole universe from the beginning, the nature of the effects requires a cause beyond the universe—a transcendent cause. Since God, as conceived by both theists and deists, possesses this attribute (as well as intelligence), God could plausibly function as an explanation for the beginning of the material universe and the origin of its fine tuning.

some theistic evolutionists have suggested, as deists also would have to do, that the information necessary to produce the first living cell and other novel forms of life could have been present in the fine tuning of the initial conditions of the universe. I first encountered the front-loaded view of design

to say that the processes that natural laws describe can generate functionally specified informational sequences betrays a confusion of categories. Laws are the wrong kind of entity to generate the informational features of life. To look to the laws of nature to generate information is to search for the improbable and specific where it is least likely to be found: in the domain of the recurring and the general.

category error. Physical laws do not generate or describe complex sequences, whether functionally specified or otherwise; they describe highly regular, repetitive, and periodic patterns of events. This is not to malign the laws of physics and chemistry. It’s simply to accurately state what they do.

whopping 1079,000,000,000 different ways of combining just the proteins in a relatively simple unicellular yeast.

vastly exceeds the

number of elementary particles in the universe (1080) and even the number of events since the big bang (10139).

front-loaded hypothesis borders on the absurd. If even starting with a biologically relevant soup of information-rich macromolecules (not just their constituent parts) does not ensure that life would self-organize, then it follows a fortiori that the much less biologically relevant configurations of elementary particles or matter and energy fields present at the beginning of the universe would not ensure such a self-organizational origin of life either.

the evidence we have about this universe and about DNA suggests that the information necessary to produce the first cell did not reside in the elementary particles or energy fields at the beginning of the universe. Consequently, additional information must have arisen after the big bang.

Shannon’s tenth theorem,23 named for Claude Shannon, the MIT scientist who developed information theory in the late 1940s. This theorem recognizes that as an information-rich signal or message travels across a communication channel (or through a medium of transmission), it will typically experience what information theorists describe as degradation, corruption, or loss of fidelity, a process described by the second law of thermodynamics. Shannon’s tenth theorem states that in order for the information-rich message to arrive from a sender to a receiver without loss, information from outside the system must be added to compensate for the inevitable loss of information that will occur as the result of entropy-producing buffeting from the surrounding environment. In other words, to transmit information with a high degree of fidelity, information from a “correction channel” outside the main channel must be applied to constrain the inevitable noise.

Which exp!ains why mutations in DNA code "naturally" inevitably occur?

given the facts of molecular biology, the axioms of information theory, the laws of thermodynamics, the high-energy state of the early universe, the reality of unpredictable quantum fluctuations, and what we know about the time that elapsed between the origin of the universe and the first life on earth, explanations of the origin of life that deny the need for new information after the beginning of the universe clearly lack scientific plausibility. And since deism denies that God could have or would have acted to add any such necessary information after an original act of creation, deistic and other truly front-loaded design hypotheses cannot account for the origin of the first life.

this does not prove God’s existence, since superior explanatory power does not constitute deductive certainty. It does show, however, that the natural sciences now provide strong epistemic support for the existence of God as conceived by Judeo-Christian and other traditional theists.32