Is there a God? How would we know?
Among atheists today there is a sizable subset who think that if God is real, he ought to be detectable through science. I can see the appeal in thinking that, since science tells us so much about the world. Even better, it has ways to adjudicate factual disputes, especially when it’s possible to employ very careful measurement and control of variables.
Ironically, those are exactly the factors that make science a poor way to detect the reality of God. If you wonder about God, it only makes sense to seek him on his terms. Or you might put it this way: if you want to know whether there is a God, you ought to ask the question in a way that you could tell the answer if the answer were yes. There are so many things science is good for, but for this task it’s not up to the job. Its competence is broad, but it’s in the kind of things that won’t discover God, if there is a God.
Scientists often make an informal distinction between “hard” and “soft” sciences, with specialists in the “hard” sciences often expressing doubt that the “soft” sciences are science at all. The two groups are distinguished by how finely they can measure the behavior of their subject matter, which happens to run almost exactly parallel with how much personal freedom their subject matter can express. People are harder to measure than rats, which are harder to measure than chemicals in test tubes.
If God is truly personal—and especially if his personal freedom exceeds that of humans—his activity is likely to be very difficult to measure with any precision.
True experimental science involves controls. In the classic format, an experiment involves two or more samples, specimen sets, etc. matched in every way possible, with one of them being subjected to some experimental manipulation or intervention, and the other not receiving that treatment. If the two groups’ outcomes after treatment/non-treatment are significantly different, researchers generally find it safe to conclude that the treatment was the cause of the difference. (There are complexities galore on top of that, but that’s the basic picture.) The great virtue of experimental research is its ability to isolate and control variables.
If God is truly sovereign over the world, it’s unlikely that he would subject himself to being controlled like a lab rat or a chemical in a test tube—or even a particle in a high-energy collider, named after him. It’s unlikely that his effects as creator could be cleanly isolated from his work in creation.
There are other approaches to research, of course. In the social sciences, correlational research is more common than experimental research. In one simple form of correlational research, persons are measured on a pair of variables, for example “happiness” and salary. If the measurements reveal that salaries tend to be higher among happy people, and lower among less happy people, then researchers conclude that (a) greater happiness tends to cause higher salaries, or (b) higher salaries tends to cause greater happiness, or (c) neither of the above, or (d) both of the above, or (e) nothing definite at all. Option (e) is by far the most commonly selected option, because except in very special circumstances, correlation does not show causation.
(By the way, my example here is fictional. I’m not aware of any reliable research showing that happiness varies uniformly [monotonically] with salary.]
Correlational research can’t generally lead to conclusions about causation because it’s not well controlled, as experimental research is. Its single greatest weakness is that on its own, it’s insensitive to the presence of other related variables. My first research methods prof told our class about the finding that ice cream sales was highly correlated with crime rates in St. Louis. Does ice cream cause crime, or does crime make people hungry for ice cream? In this case the likely answer is neither, but outdoor temperatures could influence both.
Where correlational research is most able to show causation is where relationships are very obvious, clear, and simple, with a minimum of potential hidden variables. When things get complex, though, science is very limited—unless it happens to be in some field where experimental methods can be used. Look where you find more controversy, and where you find less: Complicated (mostly experimental) physics can reach considerably more definite, reliable, non-controversial conclusions than complicated (mostly correlational) psychology can.
If God is truly sovereign over the entire created order, it’s exceedingly likely that he could manipulate variables beyond number, acting in such complicated ways that no non-controversial conclusions could ever be derived about him through correlational methods.
The God Science Could Detect
Notice that so far I have said nothing about God that depends on the truth of any religious belief. I’ve only made “if” statements that ought to seem reasonable to reasonable readers. And I suppose theoretically those “if” statements about God might be premature. It might be that there is a God science could detect. If so, that God would have to be fairly impersonal, regular, predictable, with various aspects of his action able to be isolated, like forces of nature are, for example.
Is it any wonder, then, that people who look to science for their every answer tend not to find God, but find something god-like about the forces of nature?
A God Science Could Never Detect
Suppose, though, the question is whether there is a God whose freedom is limited only by consistency with his own nature, and who is the sovereign creator of the natural world. If that’s the question, then science is probably, by its very nature, the wrong way to ask. It’s highly unlikely the maker of mankind would subject himself to experimental manipulation. It’s highly unlikely the omniscient, omnipotent God would simplify his actions in the world to a few, simple, measurable variables.
To Seek God On His Own Terms
If, then, there is a sovereign, omnipotent, omniscient creator God ruling over all the universe, it’s unlikely on the face of it that such a God would submit himself to the level of control, manipulation, and isolation of variables that science employs—especially science in its most powerful, experimental mode. It’s unlikely that he would play our game our way. It’s far more likely he would set his own agenda, and (if he wanted to be known) reveal himself in his own way.
So it seems to me that without any other religious guidance or information, a rational person who was wondering about the reality of a powerful, sovereign, wise, personal God would avoid asking whether science could reveal such a God. That person would instead ask, “Is there any hint of any knowledge or tradition in the world where God seems possibly to have revealed himself on his own terms?”
It seems to me that would be the only sensible starting point for one who would seek God, who even wonders about God.
There are of course several such traditions. They disagree on many things. Maybe our sensible searcher would then ask, “Is there anything they all agree on?” At this point I would, finally, point the quest in a particular direction. I would suggest this searcher look at one of the few things all the traditions seem to agree on: in one way or another, they all regard Jesus Christ very highly; they want to claim Christ for themselves.
So I would suggest to this searcher, if Jesus Christ is such an important common figure in all the major traditions, maybe the actual tradition he founded would be the most likely place to begin your search. If you do, you’ll find that he meets the description of what you’re looking for: a God who wants to be discovered — on his terms, though, not ours.
Related: What Would God Look Like If He Came To Earth?