Science of the gaps
Science leads to naturalism,1 which then leads to atheism—a well-trodden path many atheists have walked. But does science really support atheism?
God of the gaps
Atheists often use science to argue that God does not exist because He is no longer required. God was a convenient idea that answered any problem and could never be disproven. In times past, God was needed for the things we couldn’t explain—He was God of the gaps in our knowledge. Now science is closing gaps in our knowledge, and as those gaps disappear, so does God.
Proponents of this argument complain that ‘God did it’ is an unscientific and unreasonable explanation for observations that we make. Theirs is a strong argument against superstitious beliefs in God—i.e. using the supernatural to explain the unknown. When the supernatural is used merely to plug gaps, it will of course disappear when the gaps disappear. We no longer need Thor to explain thunder and lightning, because discovering electricity provided a natural explanation. We don’t need Poseidon either, because we now know the wind and moon cause waves and tides.
It’s not surprising that science leads to naturalism. The scientific community uses repeatable experiments to assess the behaviour of the universe and then tries to describe it with ‘rules’. This requires the a priori assumptions that there are rules, and these are fixed.
But since a supernatural event is defined by its violation of these rules, science is by definition only equipped to investigate what is natural. It cannot disprove the supernatural, since that would require proving naturalism (i.e. that the rules always apply), and how can it test something it has to assume in order to operate in the first place?
Take one example of a biblical miracle—the iron axe-head caused to float on water (2 Kings 6:5–6). By analyzing the density of iron today, can science prove this miracle never happened? Of course not. Science cannot assess whether God did or didn’t choose to override the rules of nature at that time just because He doesn’t do so every time. Science is simply not equipped to pass judgment on such situations. The supernatural (‘above nature’) power has stepped into the universe with its higher authority; it’s one-off and unrepeatable and thus not subject to experimental test.
So firstly note that ‘God of the gaps’ is not a proof. Of course ‘God did it’ as a hypothesis is unscientific—inherently so. However, it is not really intended to be scientific. Consequently ‘God of the gaps’ is not an argument, but an observation—and a complaint of unfairness: ‘How can we use science to argue against theism when it’s inherently not disprovable?’
Science and theism once worked hand-in-hand because they understood one another’s relative capabilities and domains. Science works with the natural, where the ‘rules’ apply. However, these won’t apply at the point of initial creation, when the rules were presumably written; or during the supernatural events of Creation Week, or any instance of divine intervention by miracle thereafter.
Secondly note that while superstition may be eliminated by good science, overcoming theism will take a stronger foe.
In the ‘God of the gaps’ argument, God is replaced by new knowledge of how things work. But to a theist, God is not only needed for the things we don’t understand, but also for the things that we do understand. Theism answers deeper questions than ‘How does it work?’—Theism addresses ‘Why does it work?’ and ‘Where did it come from?’ In other words, God is not merely God of the gaps, but God of the lot. Understanding electricity may eliminate the need for Thor, but not for Yahweh.
An understanding of mechanism does not provide an explanation of its origin! Just because I know the law of gravity doesn’t mean the law wrote itself. Science and theism once worked together because of this understanding also. Some of the finest minds in the history of science, groundbreaking pioneers in their fields, were convinced biblical theists; they considered the fact that nature follows rules (the first a priori assumption of science) as evidence that it has an author; someone who ‘wrote’ the rules.2
The ‘first cause’
In a sense theism does fill a gap—a very, very big gap, however—not so much a knowledge gap but a philosophical one. It’s called the ‘first cause’. To explain, let’s hit ‘rewind’ on the universe and see where we end up.
It’s the nature and definition of time that each moment is caused by the moment before it. Cause and effect bind everything in the cosmos. As we rewind the universe, we see that everything we know has a cause, and that cause also had a cause, and so on; every cause in this timeline is also an effect; it has itself been caused by something else.
Contemplating this more deeply, we realize time and causality are integrally linked; any event occurring in time cannot occur without cause. So at some stage, we end up with something that has had either no beginning or no cause. There is actually only a handful of potential solutions to this:
—Perhaps the universe itself had no beginning and has always been here. For a long time this was the prevailing theory, however, compelling evidence (e.g. the second law of thermodynamics) indicates it’s not the case.
—Perhaps the rules that govern the universe had no beginning, and then some naturalistic process caused the universe to develop within this pre-existing framework. Maybe, as has been suggested, there was a quantum fluctuation that suddenly went nuts and blew up into our universe. In this case, the past is like the left side of an exponential curve—it gets closer and closer to nothing, but it never actually was nothing. So in this hypothesis the universe has no beginning, but has not always been in its current state.
—Maybe there is a bigger thing than our universe that had no beginning, some big automatic universe-maker which spits out universes like ours (and universes unlike ours) all the time. This is one version of the multiverse idea.
—Or maybe, there was a beginning to all nature. And that beginning was caused by something outside of time and hence unbound by causality: God, who is not an effect because He had no beginning and is not bound by time. He created our universe, wrote the laws of nature, and at the least set it on its current trajectory (and possibly even maintains it on its current trajectory every moment, and even gets involved with everything that happens in it). God the universe-maker: like a multiverse in some ways, but really intelligent, so He only needed one attempt to get the design right.
Naturalism and the First Cause
Despite many variations of argument, the atheism-theism debate reduces to some simple alternatives at its core. Was there a first cause, or was there an infinite regression of causes going back forever? Did time itself have a beginning, or was our beginning somewhere in the middle of an infinite quantity of time? And if there was a first cause, does it have a mind or not?
This then is the true role of science in the debate: that, knowing the nature of time, both sides are forced to look at the known properties of our current reality as revealed through science, and then present their narrative of its origins. The narratives all have one thing in common—they begin with something, somewhere, that has no beginning.
If you commit to naturalism, you need to be able to take it all the way back to the ‘first cause’ in order to fill the ‘first cause’ gap. You need to find a narrative that brings life into existence, and before that the universe; and that explains why these things happened based on whatever mindless thing preceded them. You need to be able to describe and predict these events within an infinite time-frame with a single set of rules that are either unchanging or change for a specific mindless reason. Then you need to dismiss the apparent purpose, beauty, and morality of the result as coincidences. Unless the naturalist narrative can do all of that, it is unable to fill the same gap that theism does. Given that, it’s amazing that most scientists feel theism lacks evidence and has the higher burden of proof.
Naturalism’s problems—the big-ticket items
Science does play a role in the theism/atheism debate. But far from disproving the supernatural and hence ‘proving’ atheism, scientific evidence has been creating increasingly large obstacles to the naturalistic narrative at every step.
Biological evolution is the flagship of the naturalistic narrative—the great tale of origins that says how all creatures came to exist without any supernatural intervention. But a big-picture look at the theory as a whole shows its road has not been easy—in fact increasingly difficult, as modern science progressively reveals life’s complexity.
Evolution makes some basic predictions: we should see ‘transitional forms’ in the history of living things, and this history should conform to a branching pattern that reduces back to a few simple common ancestors. But these first two basic predictions of evolution are the first two problems it has. The fossil record has not provided us with hosts of transitional forms, but rather with the sudden appearance of complete creatures. If this were not so, ‘punctuated equilibrium’ would never have been proposed. The flora and fauna on this planet do not arrange nicely into branching diagrams of heredity. Phenomena like ‘evolutionary convergence’ are entirely unexpected situations where life refuses to establish a branching diagram—good examples are the armadillo and pangolin. Some creatures have such an unusual mixture of features that evolutionists cannot agree what branch they belong on—a great example is the aye-aye (a nocturnal primate—maybe—of Madagascar).
The selection process underpinning evolution is competition between creatures, resulting in survival of the fit and elimination of the unfit. However, all known ecosystems utilize diversity in a well-balanced network that is robust to withstand variations in environment—teamwork more than competition!
DNA’s discovery triggered a complete reformulation of evolutionary theory. Until then, the basic mechanism required a somewhat ‘mystical’ modification of living things. It then became specific: random copying errors in DNA code. However, evolution then became a victim of mathematics. What is the likelihood that such a code, able to provide the blueprint for a person, could accidentally come into being? (Selection is no use until such reproductive machinery exists.) At what rate can any useful error in it be selected by means of some resource competition, and then dispersed throughout a population until it exclusively dominates? Can that rate out-pace the staggering rate at which errors erode the data into meaninglessness?
A handful of basic predictions made in order to operate in a range where evolution would maybe, just maybe, be possible have all been disproven. For the mathematics of neo-Darwinism to work, DNA needed to be mostly junk; but virtually none of it is. In fact, DNA is so not junk that it can be read in multiple ways—much like a crossword where every row, column and diagonal makes sense.
It gets worse. Even if we allow that random changes to DNA do result in improvements that are selected, we still have barely a fraction of the naturalistic narrative needed to explain today’s diversity of life-forms. We still don’t have a mechanism to account for the big structural changes that must have taken place: evolving from asexual to sexual, from single-cell to multi-cell, and the big one—going from non-life to a single cell that has DNA, has proteins capable of reading the DNA and using it to make new proteins including an exact copy of itself! That last one’s a real kicker.
For each of these gaps, there are forms of answers. They usually start, ‘We don’t know exactly how, but somehow [thing] happened, and then … ’. The holes in the story are getting bigger, not smaller, and it’s not just biological evolution that has this problem.
Consider the formation of stars, galaxies and our solar system. They supposedly formed from gas and dust clouds clumping together, rather than dispersing. But no model can successfully simulate this—and definitely not in a way that puts the sun rotating at 7° from the planets’ orbits in our star system. For almost every solar system body, the magnetic field strength is a surprise. Mercury shouldn’t have a magnetic field (but it does); surely Venus and Mars should have one like ours (but they don’t); Jupiter’s shouldn’t be so strong; Saturn’s shouldn’t be so symmetrical; and Uranus’ and Neptune’s shouldn’t be so asymmetrical. The geological behaviour is frequently unexpected, too (volcanism on bodies too small to retain their heat for billions of years—Io, Pluto, Charon, and more).3
Essentially, the preferred naturalistic models for the development of our solar system cannot account for any of its major features. Yet. But they will—just give science time, we’re told, and a solution will eventually come.
What about the universe? The currently favoured evolutionary cosmology is the Lambda Cold Dark Matter (ΛCDM) model. Apparently it’s the best solution and the one we should all believe, subject to a few kinks being ironed out sometime in the future (e.g. the ‘Axis of Evil’—yes, that’s literally what scientists are calling it). However, the ΛCDM model is starting to look like software that has a new hot-fix each time a bug emerges:
— ‘The mass of the universe isn’t adding up—so let’s postulate some type of invisible mass, called dark matter (DM).’
— ‘The universe is (supposedly) accelerating and we don’t know whence the energy—so let’s propose a type of energy that just does this, called dark energy (DE).’
The very definition of DM and DE is ‘something that explains what we didn’t expect to see, without affecting other things we can actually see’.
Now generating an idea that explains your observations is great. However, that’s a bit different to defining something as the thing that would explain your observations. Put simply it’s like saying: ‘I add two and two, and get six—this doesn’t mean I need to look for something wrong in my assumptions, it just means that I need to add ‘dark numbers’, which always magically equal the difference between the two sides of my equation.’4
Does it hold water?
Consider an analogy—you’re building a dam against a rising river. For a dam of proper thickness, you can only make enough bricks per day to dam one metre, but the water is rising two metres a day. So you make the dam thinner. Now of course it’s weak and won’t last, but you plan on making it thicker later when the water stops rising. This is a risk. Effectively, you are building with future-bricks; repeatedly assigning a place for bricks you haven’t made yet, trusting you will be able to make them later. Now this works for some situations, but it has limits. Just how much water can you hold with future-bricks?
I have discussed evolution with friends and heard this objection—‘But how do you know they won’t solve that problem in the future?’ My response: ‘Why should I think that they will solve these problems in the future?’ So far, as we saw, evolution has made almost exclusively false predictions. There are so many future-bricks, one wonders where the real bricks are.
The secret is, atheism is having difficulty with its gaps. They are really big gaps. The new ‘God did it’ is: ‘Science will do it’.
Just because ‘God did it’ is an inherently convenient hypothesis, because by definition it can never be disproved, does that mean we can never have evidence for it? A reason to believe it? Consider another analogy: a father walks into his shed and finds paint splattered over the wall. He immediately blames his son Thomas for playing with his paints. His more generous wife asks him to consider whether an earthquake, or the vibrations from a passing truck, could have shaken a paint tin off the cupboard. Both explanations—including the ‘unscientific’ option, to blame Thomas—involve faith. However, both parents may legitimately seek supporting evidence for their hypotheses. In one sense, neither can prove scientifically that Thomas didn’t do it. Thomas’s behaviour is not subject to natural ‘rules’ like vibrations; He has a mind, and could even do it in a way that would imitate an earthquake if he wished.
Whilst science usually accepts proving falsehood as stronger than proving truth, this is different. The father can more easily find evidence that Thomas did do it—for example, he might find a note on the table saying: ‘Sorry, Mum and Dad, for spilling your paint on the wall. Love, Thomas.’ Similarly, if the mother does some modelling and calculates an earthquake scenario that fits the evidence, this would support her ‘naturalistic’ hypothesis. Consider two scenarios:
—Scenario 1: Mother does some modelling and comes up with a possible but improbable scenario that involves a very, very special earthquake which got the paint tin to jump from the back of the cupboard over the (untouched) front row of paint tins and the lid to come off in a way that left screw-driver-shaped dents around the rim. Mum’s modelling is really evidence for Father’s hypothesis. ‘Thomas did it’ is looking likely.
—Scenario 2: Mother tries to model it but fails. Despite this, she says, ‘We don’t yet know precisely how, but somehow, probably by means of an earthquake, the paint was thrown from the back row in the cupboard, its lid was removed leaving screw-driver-shaped dents, and it was splashed over the wall. And the note left on the kitchen table was most likely a fabrication that Father wrote because he always wants to blame Thomas’. This is going a step further. This is what I call ‘science of the gaps’.
Consensus vs reality
The global consensus of support for evolution and naturalism is astonishingly strong. Can they all be wrong? It would not be the first time in history that the entire scientific community believed something that was false for ideological reasons. Just look up Aristotelian mechanics and revolutionary scientists like Galileo. Is there something we need to learn from that history, where ideology disrupted the scientific process?
In Galileo’s day, the science was not all wrong. The evidence was consistent—balls still fell back to the earth when thrown. The interpretation was wrong. Today I believe we are in the same situation—I don’t think the scientific discoveries about fossils and cells and DNA are false. I don’t have some conspiracy theory about hidden evidence and falsified reports. But I do think the interpretation is wrong. Every new piece of evidence is being crammed into the existing theory. Hypotheses are stretched to fit around evidence in ways that are not comfortable for them.
Perhaps what the scientific community lacks—like what was lacking before Galileo—is a willingness to disprove what we believe. Without falsification, science cannot truly function. A genuine look at the evidence suggests evolution has been sufficiently falsified already—none of the ‘real’ bricks are left. But who in science/academia has the courage to say so? Rather than science supporting naturalism, the ideology of naturalism is inhibiting science.
In truth, it is easy to look at this universe, and say, ‘God did it’. I hear nothing sufficiently compelling to suggest I should abandon my belief in the Bible’s account of creation. All I hear involves ‘future-bricks’—claims that if we try hard enough, we’ll work it out eventually.
We don’t need to wait for science to ‘work it out’, or atheism to plug all its huge gaps. What reason do we have to think that future-bricks will fit the slots left for them? We have something more compelling, a ‘note from Thomas’ that says, ‘I did this’—aka the Bible.
Theism, as a hypothesis, is not gap-filling. Belief in God is both evidence-based and philosophically necessary. The universe as good as has His signature on it. If you claim science as your reason for not believing that, just check—is it real science, or science of the gaps?
References and notes
- The belief that all effects arise only from natural causes, that there is no supernatural. Return to text.
- See Sarfati, J., The biblical roots of modern science, September 2009; creation.com/roots. Return to text.
- For references and further reading on the topics in this paragraph see creation.com/solar-exoplanet-qa, or/and use the search engine on creation.com. Return to text.
- A more apt analogy, though less accessible for most, would be that it’s like fitting a 64th order polynomial to 65 data points and saying you know the relationship behind them. Return to text.