Atheists typically assert that atheism is just disbelief in the existence of gods, and when they say this it is usually to insist that little, or even nothing, follows from this definition (or has ever followed from it). On this line of reasoning, Stalin’s monstrous behavior in the 20th century (for example) had no causal relationship to his atheism because—let’s all say it together, shall we?—atheism is just disbelief in the existence of gods.
But, as Aristotle long ago noted, every definition bears within it—if it is to be any good—an explicit or implicit genus (or framing hierarchy) and species (or differentia).
So, with the definition, “Atheism is disbelief in the existence of gods,” what’s the implicit genus? Answer: ideas about gods. And what’s the species or differentia that sets atheism apart from other ideas about gods? Disbelief in the idea that gods exist.
I’m bringing to the surface the underlying genus of the definition to make a point that ought now to be obvious: atheism is an idea and ideas have consequences.
Somewhere along history’s way, a person had an idea—maybe there are no gods at all—and that idea opened a range of new implications and possibilities for thinking and action.
Now, it’s true that disbelief in the existence of gods leads to no certain deductions. A deduction is something that, from true premises, follows with 100% certainty. If we say, in a familiar example, that Socrates is a man, and all men are mortal, then Socrates must be mortal. If the premises are true, it cannot be otherwise.
As an idea, atheism is not like that. It does not have any 100% certain deductive conclusions that can be drawn from it or about any individual who embraces it. (Is there any large idea or belief that does function in this way?)
But atheism is still an idea, and thus, again, like any idea, it has consequences. Those consequences rest, not on deductive certainty—as a conclusion following from true premises in a syllogism—but on probability.
In other words, we can at least argue inductively about the probable consequences of atheist profession (both in history and for the future).
So, even though atheism is a word with a narrow definition from which no (or very few) deductions can be drawn, it’s also an idea that, when accepted, makes some things more probable than others. And, though counterfactuals are dicey to argue in history, it’s also quite obvious that atheism’s effects can be plausibly discerned in history. Thus, a person can reasonably argue that atheism in the 20th century did not make the rise of a person like Stalin certain, but it did make him more probable. (Whether Stalin actually did some of the things that he did because of his atheism is to raise the problem of untangling correlation from causation.)
But this is true in all inductive and historical reasoning. For example, the spread into Gentile lands, 2000 years ago, of belief in a messiah crucified by unbelieving Jews, made subsequent antisemitism throughout history vastly more probable. Christianity is not just belief in Jesus. It has other consequences, and the longer the belief exists the more probable certain consequences are to be realized. In some sense, Christian belief bears some degree of causal relation to the Holocaust even though Jesus’s early followers, and their rhetoric against Judaism, are separated from the Holocaust by two millenia. Likewise, atheism bears some causal relationship to Stalin and his hostility toward institutional religion.
As to arguments about the future, atheism makes eugenics more probable in countries led by atheists (such as China). The reason for this is that atheism tends to manifest, not just as the ground for humanist philosophies, but for Machiavellian and Nietzschean ones as well. Like Stalin, the leaders of China are far more Machiavellian and Nietzschean in their atheism than they are humanist. Their atheism is not of the humanist variety, but it is from the family of atheism. Likewise, the Christianity of most Mississippians is of the fundamentalist variety, and not of the Anglican or Quaker progressive variety. As such, this makes death penalty, anti-abortion, and anti-homosexual laws in Mississippi far more probable than in a state where few fundamentalists exist (like, say, Vermont).
In short, ideas do not exist in vacuums. If they don’t have strictly deductive consequences, they nevertheless have inductive (probabilistic) ones. Atheism, as with any idea that enters history, should be debated beyond the narrow limits of its formal definition. To echo George Santayana, those who do not learn from history are condemned to repeat it.
For a book that I think lays out some of the historical consequences of atheism thus far, I recommend historian Michael Burleigh’s thought-provoking Earthly Powers (HarperCollins 2005).