A Critique of “The God Hypothesis”

God the Tinker.

To a carpenter, everything looks like a hammer.

Robin Collins, in his article “God, Design and Fine-Tuning,“ presents the observation that our physical universe is perfectly endowed with precisely those properties which allow life as we know it to exist. He then attempts to prove by means of logic, evidence and analogy that this observation itself constitutes strong evidence in support of the hypothesis that such precision shows the hand of God when compared to any hypothesis that specifically excludes any intelligent designer. This paper will first address in general terms the problem presented by attempting rational, scientific arguments for theism in general, relate specific source materials on the nature of true scientific inquiry to Theistic arguments, present a brief summary of Collins’ “God Hypothesis,” evaluate that hypothesis on its merits as a scientific theory, and then attempt to push beyond the scope of Collins’ paper to specifically address why I believe theism to be an explanation of the observed phenomena which is not only fundamentally irrelevant to scientific inquiry, but also epistemologically unwarranted, practically inadvisable, and even morally treacherous.

Proving that the Judeo-Christian God exists is inherently impossible. Proving “he” doesn’t exist is equally fore-doomed to failure. The God defined in the Canon is imbued with characteristics which are fundamentally inaccessible to “his” creations, that is, to us humans. Not only are attributes such as “All-knowing,” “All-seeing,” and “All-Powerful” un-testable, but they are embedded in a paradigm which can explain any and all conceivable phenomenon as “God’s will.” That the existence and nature of God be inaccessible to logic is perhaps the most fundamental property of any Theistic paradigm, because God is defined as the will behind everything that is otherwise inexplicable. God is basically the personification and domestication of The Unknown. Therefore, as it is my intention to prove below, any rational argument which is predicated on the existence of God automatically fails a number of tests for scientific validity.

Karl Popper suggested that science could be distinguished from pseudo-science by virtue of its being falsifiable, or testable. God’s existence cannot be tested or falsified because no act, phenomenon, or event is beyond the ability of an all-powerful, all-knowing entity. This means that no logic-based falsification can be possible of any theistic premise because God’s omnipotence necessarily requires that “he” callable of salting the environment with evidence specifically engineered to produce such a crisis of faith in the scientist. Given God’s ability to deliberately plant red herrings into creation at whatever level required to produce such a crisis of faith, clearly any Theistic argument fails Popper’s test.

Imre Lakatos believes that “fruitfulness” is the basis for a good scientific theory. If an hypothesis successfully predicts new phenomenon which then opens up new lines of research then Lakatos considers it fruitful. No Theistic argument can ever be “fruitful” because any phenomenon ascribed to The Deity flows from the Mind Of God, which is inherently unknowable, and therefore suggests no new line of inquiry, and so is necessarily barren.

Thomas Kuhn allows that normal science can bear some resemblance to pseudo-science, but that when true science enters a state of crisis because observations are made which are not predicted by the theory, or that the theory predicts will not be made, that the basic paradigm naturally enters a state of flux until a new one can be discovered to accommodate the new observations. While theistic paradigms do evolve, it cannot be said that they usually evolve on the basis of observations of basic, natural phenomena. Instead, they evolve in response to changing social and economic realities which render different world-views more or less applicable to people’s lives. Challenges to dogma that come from observation of physical phenomena are met with either repression or retrenchment, seldom with basic paradigm shift within the institution. The prime example is that of Copernicus. His observation that a heliocentric cosmology fit the data better than did a geocentric one violated dogma, and was initially met with repression. Eventually, the Catholic church stopped denying the validity of the heliocentric hypothesis, but they did it by retrenching their original argument of man’s paramount importance in the universe by simply no longer invoking geocentrism in support of it. Theistic arguments in general have no internal support for self-deconstruction the way scientific ones do, because they seek always to defend and maintain the original premise despite any and all contrary evidence. Real paradigm shift in religion is always promulgated from society inwards, never from Theism outwards.

Paul Thagard argues a similar point, that it is the fruitful reformulation or replacement of a theory in the presence of progressive, alternative paradigms that defines it as scientific or pseudo-scientific. He goes further than Kahn by claiming that “ostrichism,” or deliberately ignoring competing paradigms is not only unscientific, but fundamentally non-rational. Theistic arguments tend to ultimately fail the paradigm shift test, and often display marked “ostrachism.” While Collins does actively engage competing paradigms, and while those alternative paradigms do fail to provide any more than a whiff of progressive fruitfulness (in terms of actual predictions,) the God Hypothesis ultimately cannot, for the reasons stated above, ever make any kind of testable predictions whatsoever. For this reason it ultimately fails Thagard’s test as well.

The nuts and bolts of the God Hypothesis can be succinctly summarized as follows:

  • Current theories of Physics and Cosmology have identified a very small number of physical laws and constants that, together, determine the nature of our universe at every observable level.
  • Those laws and constants remain valid over a very wide range of possible values and forms, and those variations, if manifested, would have profound effect on the nature of our universe.
  • The specific conditions that allowed intelligent life to evolve on Earth can only exist in a very, very small subset of the possible configurations for a universe.
  • The phrase “fine tuning” refers to the extraordinarily fortunate (for us) accuracy with which numbers such as the cosmological constant are observed to fall into that small range which permits life to exist.
  • The fact that we find ourselves in a universe with exactly the right configuration despite the astronomical (pun intended) odds against it shows “that the fine-tuning of the cosmos offers us significant reasons for preferring theism over atheism” – atheism being defined as a belief system which includes “the denial of any sort of intelligence behind the existence or structure of the universe.

Collins admits of two classes of challenge from atheist world-views – the Atheistic Single-, and Multi-Universe Hypotheses. They are characterized by the belief that no one created the universe we live in, and differentiated from each other by the number of universes which they believe exist. The primary difference between his characterization of these two alternative paradigms is the allowance that in the multi-universe case, the statistical improbability of our universe’s existence can be mitigated by creation of sufficient universes by an hypothetical “universe generator.”

Much of Collins’ arguments take the form of situations he considers analogous but which are more accessible to the reader’s imagination. For example, on page 121 he claims that a billion fold increase in gravity is allowed by theory, but precludes life because our bones and tissues aren’t nearly strong enough to support beings of our size, nor would be any other material else that we can imagine. Other objections include observations like that the particular chemistries of water and carbon, both required for life as we know it, is highly dependent on the specific values of those few important constants which happen to be found in our universe. These arguments are presented to show that intelligent life could not possibly have evolved under the conditions you would expect to find in most randomly generated universes.

The most obvious problem with this premise is the assumption that “intelligent life” is synonymous with “hydrocarbon-based, oxygen-breathing, water-drinking beings that are about 6 feet tall and have brains that weight 3 lbs.” The only really meaningful definitions of intelligent life that could apply to a discussion about all possible universes must rely on concepts like complexity, self-replication and self-awareness, not gross morphology. Of these, only the first two are susceptible to physical examination, the third being an emergent property not demonstrably tied to any specific physiognomy. While it is true that animals like ourselves couldn’t evolve on an ultra-high gravity world, the scale at which brains are constructed, or more specifically the scale at which the complexity of our brains becomes sufficient to manifest intelligence, is many orders of magnitude greater than the scale of the molecules that are the raw materials of the universe. It is reasonable to assume that intelligent life will require energy and be made of molecules, but is it reasonable to assume it has to be about our size? There is nowhere any evidence that sufficient complexity to manifest intelligence could not develop with very much smaller structures which might be perfectly happy with the billion fold increase in gravity on Collins’ proposed planet. The number of molecules it takes to build a single neuron is staggering to the point of incomprehensibility, on the order of 100 trillion atoms,* but intelligence is built starting with 1 neuron as the basic building block. That means, if we could make neurons using only a few hundred thousand atoms, with near molecular-scale organelles, we could have brains just as complex as ours are now at a billionth our current size. Which is to say, a size at which one of this world’s cold viruses would be just about the right size to pick up and put in our pocket, were we to consider pockets fashionable and should there be any cold viruses handy we might want to put in them. At the atomic scale, gravity is functionally irrelevant, and at the scale of a virus, it is very nearly so. Which means that the possibility of a self-award life-form existing in a universe with that larger gravity constant cannot be ruled out, a priori, simply because we can imagine very tiny life forms which don’t care a fig about high gravity. Conversely, were gravity very weak, it is equally possible to imagine immense life-forms of even planetary proportions.

Nor is the abundance of the elements and molecules required for carbon-based life a sine-qua-non for the emergence of intelligent life. There exist right here on Earth primitive ecosystems based instead on sulfur chemistry, and silicon chemistry is sufficiently similar to that of carbon that it also could not be precluded from possible primordial soup recipes. All of this remains true without even considering the possibility of utterly novel processes which might possibly give rise to something we would recognize as intelligent life.

This self-awareness could theoretically come to a being at either very small or very large scales, with vastly different underlying processes driving it. So, I argue, the mere fact that a billion fold variation in conditions that limit size is possible, or that hydrocarbon chemistry is unlikely does not, as he implies, constitute evidence that the probability of a universe supporting intelligent life is one in a billion, only that a universe that supports intelligent life as WE know it is. Perhaps there were a billion other possibilities for intelligence that simply did not manifest in this one. To subvert Collins’ affinity for analogy: the fact that a snowflake is unique and beautiful and incredibly unlikely to form in exactly that way by accident does not require me to believe that someone designed it that way, as if it were somehow necessary for all snowflakes to look like the only one I’d ever actually seen, instead of simply be one, beautiful example among trillions of how snowflakes can form.

Many of Collins’ analogies focus, as of course they must, on everyday situations where either chance or conscious intent could have created an observable phenomenon. In these analogies we are naturally lead to the obvious conclusion that any sane person would reach, that the vastly more probable explanation for the observed phenomenon is the one to pick. These analogies are, of course, tailored to produce the conclusion that conscious intent is the obviously more probably explanation for most phenomenon. The analogies are leading you to confront the most basic question he tries to answer with the God Hypothesis: How do you explain the fact that the universe we observe is both overwhelming improbable, and the only one that can actually produce an observer to do the observing? For the sake of argument, let us grant the dubious hypothesis that ours in the only basic template for intelligent life; that our universe is the coin that landed on its side, instead of being heads or tails; that only a child of such a universe coin can become self-aware. If it didn’t happen, there’d be nobody around to be smug about being right that they couldn’t exist in such a universe. Nor is there any way to know how many throws of that coin a hypothetical universe-generator might have made, since such a generator is by definition outside of the space-time of our universe. If the coin is tossed an infinite number of times, the chances of it never landing on its side becomes the vanishingly small probability and the self-aware creatures in those multitude of edge-on universes become blandly commonplace, since even the tiniest number multiplied by infinity is still infinity. Moreover, given that this coin-tossing happens outside our universe, it is theoretically impossible for observers within those lucky few that created observers to have any information at all about those universes that failed to be just right. So if the information available to a resident of any one universe tells you nothing about how many universes there have been and will be, that observers existence proves only that it is possible for observers to exist, not that they were intentional. This is the anthropic principle. It is generally considered by physicists to be fairly weak, and unconvincing, but it suffices to logically counter the argument that the fact that the universe supports life is, itself, sufficient to require a creator to solve the fine tuning problem.

Collins’ counter to these argument in the single-universe case is basically that fine-tuning is likely in the God Hypothesis (Allahu Akbar) while it is unlikely in the atheistic ones. Besides the fact that the phrase “fine-tuning” itself prejudices the analysis from the start by using a term that presupposes intentionality, it is in essence, though perhaps not by Collins’ own intention, a clever sophistry which conceals the surface problem of apparent improbability with an impenetrable veil of mystery behind which improbability need no longer be evaluated because everything behind it has already got one, simple, 3 letter designation in common use throughout the world which not only needs no explanation, but brooks none. In simpler terms, it’s cheating to say that our unlikely situation exists because God made it so, because you are not then allowed to ask how likely it is that a being such as God could exist to be able to so neatly circumvent the laws of probability. If you take “his” existence on faith, it is easy, if you don’t, you have simply traded something improbable by our understanding for something utterly beyond our understanding. That trade may be comforting, one may consider it valid and true, even necessary, but there is no reason to consider it logical. Where is the logic in, proceeding from observation and first principles, declaring that the impossible is more likely than the improbable?

Collins counters the multi-universe atheistic hypothesis by examining the theory that posits the potential existence of a process that occurs somewhere outside the bounds of our universe which can result in the creation of a new universe; a “universe generator.” His claim is that the specific characteristics of such a generator display just the same level of improbability as our universe itself, and so simply pushes the creator back a step – God becomes the designer of the generator, which produces the universe.

To counter the atheistic argument that perhaps fine-tuning is a result of an as yet undiscovered higher law, Collins simply leap-frogs that argument and states that the existence of such a convenient law would, of itself constitute one more reason to favor the Theistic model; i.e. that the discovery of physical laws for universe generation which conveniently create universes people can exist in would also prove the existence of a Creator. This conclusion unfortunately throws his baby out with its bathwater. The whole God Hypothesis is predicated on its being the most likely explanation for the fine-tuning problems in physics. But these problems are specifically related to the fact that for the laws we know about to produce a universe capable of sustaining life, certain numbers have to be just exactly right for no apparent reason. Once there is a theory explaining why a constant has to be a certain value, there is no longer a fine-tuning problem, so no more need for an intelligent designer. An excellent example of this situation is the Flatness Problem.** The spacetime geometry of our universe is generally acknowledged to be suspiciously flat, which is good for us. For it to be flat by accident would require very precise initial conditions and represents an aspect of the fine-tuning problem. Theories have been proposed which attempt to provide physical laws which describe processes that result in our flat universe actually being the natural, likely shape of any universe with our laws, and random initial conditions. Cosmic Inflation, for example, describes a period in the early universe where natural processes flatten any initial shape into the one we see today, so if that theory is correct there is no flatness problem, no fine-tuning, no need to invoke the Hand of God. For Collins simply to say that the existence of Cosmic Inflation in turn proves his point would be to abandon the entire premise of his argument – that observation of the fine-tuning problems points to intelligent design – since there is no fine-tuning problem to do the pointing. This ready abandonment of the most fundamental premise of his argument reveals that the invocation of fine-tuning is merely a pretense for covertly asserting that it is simply the miraculous fact of our own existence, not problems in Physics at all, that proves the existence of God. For ‘how,’ this sub-textual argument asks, ‘could the coming into being of anything so marvelous as us, against all odds, possibly be anything but fate?” But not fate, the hand of God the Tinker, patching together the universe just right, so it’s perfect. Moreover, the evidence of that perfection is how perfect we humans are. To my mind, this isn’t logic and empiricism, it’s magical thinking, or at very least, anthropocentric projection.

Here we find the fundamental flaw in the argument that the statistical improbability of us happening by chance somehow indicates that it is unscientific to believe that intelligent life was not a result of a divine plan. Collins makes an analogy about the creation of the universe being a dart thrown blindly at the dartboard of infinite possibility which assumes that only the bull’s eye produces intelligent life. Unfortunately the only fact that is really proven by his argument is that only the bull’s-eye produces Collins himself. Were self-awareness to be achieved in one of those less universes not perfect for us, say by a nebula, or a mega-virus because gravity was much weaker or stronger which rendered those scales the most stable for complex evolution, then the self-aware version of Collins in that universe would at some point look about and marvel that their dart landed precisely in their own subjective bull’s-eye and be equally certain that such luck could only happen intentionally.

And now we come to what I personally believe is the real motivation for answering the fine-tuning problem with the God Hypothesis. God is made in our image, not the other way around. We make things. We tinker. It’s one of the primary things that defines us in our own eyes. And so, in this mechanistic age, God is a mechanic. God tinkered with the Universe ‘till it worked just right (or with the universe generator.) We need God because it’s lonely without him. We have faith in God because it helps us get through the day. Collins invokes him as part of a scientific theory because scientific theory is part of his world, and he needs to find a place for God to live in it. It is telling that God has been pushed right out of our universe by this argument. He is the original Clockmaker God of the Enlightenment, but instead of his design being the infinite variety of forms, it is the specific values of a handful of constants. But the problem with the God Hypothesis remains that it has no substance of any kind. God simply is, he can be invoked at any juncture, but after doing so, you are left with no actual new information. Even string theory, as arcane and abstruse as it is, provides some actual description of nature, so while it might also fail many of the tests of a true scientific theory which are presented above, it is at least in the same class of thought-objects as one. The God Hypothesis, however, is not, because it does not actually even concern the laws of nature, it concerns the super-natural origin of those laws. That’s a perfectly good article of faith, but it’s not a scientific theory or a logical conclusion.

But what about The God Hypothesis as an article of faith? Does it have some intrinsic merit beyond logic or science that warrants it’s adoption? I believe it does not. As an explanation of how we came to exist in a universe so perfectly suited to facilitating that existence, it is, in my opinion, epistemologically unwarranted. I believe this for many reasons, with two being principle among them.

One principle reason is my interpretation of Occam’s Razor – the Principle of Economy. The Hand of God appears to be a simple explanation for the root cause of any phenomenon, when in fact it is the very opposite. In actuality, it is a dis-explanation. It removes the consideration of root causes from the table by invoking something so enormous and incomprehensible that you are excused from not understanding it at all. But believing that a thing happened because God willed it actually requires far more assumptions than simply saying that an apparently unlikely thing happened anyway, despite the odds. To expose these hidden assumptions, take an example Collins uses to show how clearly he sees the hand of God in all “creation.” He uses the analogy of science discovering the fine-tuned constants to the experience of finding a message on the beach. He says: “if you find rocks in piles that look like a message from your brother, it’s far more likely that your brother piled the rocks than that they just piled themselves due to random forces.” While this sounds reasonable at first, the problem is that the whole point of the analogy is to build a bridge across the unknown to get you to the truth on the other side, but he has placed the starting point of that imaginary journey already on the far side of the bridge that was supposed to be creating. You see something unlikely, you interpret it as a deliberate message from your brother, then you consider how unlikely it is that waves could have spelled out the message randomly. The interpretation step, where it’s a message from your brother, has already assumed the God Hypothesis as an axiom, so you can’t prove the hypothesis by proceeding from there. The primary problem with the analogy is that is presupposes my brother’s existence as an article of faith (cogito ergo frater – presumably – est.) I have close to the same certainty of my brothers existence as I have of my own, but with God, I also need to take on faith that it is even possible for a being like God to exist at all before I can proceed to take on faith that he wrote the message I have found. If I found what looked like a message from my brother, but I not only didn’t have a brother, but didn’t believe it was possible for brothers to exists at all (due to some quirk in my universes’ biology, for example,) I might prefer to start looking more closely at those pebbles and imagining what other mean their configuration might have besides a message from my imaginary “brother.”

The other principle reason for discarding the God Hypothesis is that if something looks too good to be true, it probably is. I view the Hand of God argument as a psychological phenomenon, not a logical one. It is so profoundly comforting to believe that we are cared for that it is almost certain that the imagination will strive to fill that need in cases where the environment fails to do so. I.e. we want someone to care for us so badly that we will invent someone to do the job if nobody around us is doing it well enough. Therefore, anything not obviously and demonstrably extant that appears to fill such a basic emotional need appears to me far more likely to be a cultural/psychological construct than an ineffable, insubstantial, metaphysical reality.

Theism displays a strong tendency to lump all unexplained phenomena together and tag them as Acts of God. This label is a direct appeal to an authority beyond question or understanding. While it may be permissible to investigate the physical causes of an Act of God, the label itself sits, for believers, enthroned with as more authority than any theory of Physics could hope to enjoy, while providing no actual information of any kind, since no one can know the mind of God where the root cause of those phenomena lie. God is ineffable and in-apprehensible, so to claim that God did something is essentially the same act as drawing a line on a map and labeling the region on the other side “here there be dragons.” While some would-be explorers may be intrigued and want to find the dragons, most will be discouraged from further investigation. This makes the invocation of The Divine as an explanation of any aspect of the physical universe at best irrelevant to scientific inquiry, and at worst an impediment to it because The Divine is, by nature, immune to such inquiry. Moreover, I believe that invoking the Hand of God is fundamentally antithetical to scientific inquiry because while for some the metaphysical may simply be a parallel aspect of their worldview with the scientific, for others, the invocation of The Divine represents the demarcation of territory that is off limits for human investigation. While morality and ethics definitely should limit acceptable avenues of scientific investigation, explanations such as the God Hypothesis are not moral or ethical. They carry the same weight and have the same power to deter investigation as moral/ethical arguments do, but all they offer in lieu of potentially productive information is emotional reassurance.

I believe that replacing “I don’t know” with “It’s God’s will” is an expression of the very Human desire to be comforted. While this may be a valid emotional response, I believe that it is a specious panacea that is antithetical to, even a half-conscious antidote for active engagement in the world, and as such, should not be applied to areas of active inquiry lest it dull Occam’s edge, or worse, hide danger, harm or even evil behind a complacent veil of inscrutability.


* The math for the scale argument:

Brain has 100 billion neurons, 100,000,000,000 (1×10^11)

2% body weight avg – for 200 lb man, that’s 4 pounds.

4lbs is about 1800 grams, so say a big brain is 2000 grams of pure neurons, a single neuron then weighs

0.5 x 10^(3-11) = .5×10^(-8) = 5×10^(-9) grams

1 gram of carbon contains about 6×10^23 atoms, which means a neuron has on the order of 1×10^14 atoms, or about 100 trillion.


Our universe is very close to flat, which theory predicts could only be true if it started out very close to exactly flat – very unlikely if all possible configurations are equally likely, the first assumption. However, if the theory of cosmic inflation is correct, then no matter what the initial value, the constant will be driven down to the suspiciously small value the cosmology requires to match observed flatness.


Leave a Reply

Your email address will not be published. Required fields are marked *