Even if we accept the definition that babies can be atheists – which is fine for now – that statement would still be literally true.
A baby does not have atheism in the same way that an adult has.
This does not have to be a denial of baby-atheism. It’s a qualitative statement. The quality of baby-atheism is different from the quality of adult-atheism because most adult atheists have consciously considered the issue and decided not to take up theism. This might be because they were taught theism as children and later rejected it, or because they were taught no religion and came to consider the possibility of a god or gods, however briefly, and rejected it because they considered it unlikely.
You can give an adult a survey: “How many gods do you think there are? How sure are you of the answer?” Lots of adult-atheists, most especially those who self identify, would answer zero and say they’re pretty sure but not totally certain.
You can give a baby the same survey, and all the baby can do is drool on it.
A baby does not have atheism in the same way that an adult has. The quality of their atheism is different. This is an important distinction.
Using a narrow definition of atheism, this is true in a funny sort of reverse-True-Scotsman way. It also brushes over yet another difference in quality.
People who self-identify as atheist I’d say are more likely to assign a low probability to the notion of one or more gods existing compared to other groups. People who self-identify as agnostic are more likely to say “I don’t know” when they’re asked the probability of one or more gods existing.
These are interesting qualitative differences. These are important distinctions.
And in fact, the word as it is actually used out there in big wide world quite possibly takes full note of those distinctions in a way that’s very useful for conversation, even if some would like to deny that. Real world usage of “atheism” and “agnosticism” when people are successfully communicating with each other about their beliefs might not conform so cleanly to one’s favorite definition. The word out in the wild, far away from any personal semantic zoo, might just be thrown around most often for the probabilistic denial of the existence of any gods – not necessarily absolute denial, but the belief that the weight of likelihood is against. The word “agnostic” could be more general.
I don’t have a corpus of evidence about this, but I can say I’m much more interested in what people actually mean when they use words for their own precise purposes, rather than what some outsider decides to label them as in order to maintain the purity of a personal dictionary.
This is especially weird when actual real dictionaries disagree.
Agnostic
New Oxford American: a person who claims neither faith nor disbelief in God
Merriam-Webster: a person who does not have a definite belief about whether God exists or not
American Heritage: One who is skeptical about the existence of God but does not profess true atheism.
People use agnostic all the time to avoid professing atheism. I did it myself. When I called myself an agnostic, I did so without any belief in gods and I outright denied that I was an atheist when asked. If you had asked me the chance that gods exist, I would have said “I don’t know”. You can affix whatever label you like on such people, but I’m sure you realize that the meaning of words does not come from some holy book. In a conversation, the word means what people use it to mean. If I’m having a discussion with one of them, I’m not going to use labels contrary to their own usage because they won’t understand it and communication will be impossible. In addition, they will quite likely take offense. I know that I would personally have taken great offense at being labelled an atheist when I identified as an agnostic, and I would guess this is more generally true.