Does God Matter? Depicting Belief Systems Through Causal Graphs

One of the most interesting things about science or the scientific process is that it can literally be applied to any topic. Going through higher education, you'd find all sorts of claims made by classmates or educators about religion and two of the more interesting ones were:

"The more educated people get, the less religious they are and the less they believe in God or a deity"

OK, sure but then on the other side...

(from a teaching physician) "The Anterior Cruciate Ligament (ACL) is the most amazing structure in the world, it's how I know there is a God"

I have lots of degrees, and also have needed my ACL reconstructed twice... so should I be an atheist? While I can't say anything to the veracity or logic behind either of the above two quotes, what I can say is that directed acyclic graphs (DAGs) are a great way of depicting belief systems about CAUSALITY (theistic or otherwise!) and so I thought it would be interesting to play out common Western beliefs through DAGs for people to think about. While there will be a bit of theology in here, you'll also see how we can layer in Causal Inference into these graphical systems.

The Atheist

Here, we can see that Humans cause their Outcomes, we can also see that "God" arises from both the Humans and their Outcome. An Atheist may say, that God was created by Humans and then that the Outcomes are then attributed to God. In Causal Inference talk, we would call God a 'collider' as it arises from the other two nodes (circles). In this case, both the Causal Inference analyst and the Atheist would say we should "remove God from the analysis to know how much humans cause future outcomes."

Laplace's Demon

Laplace's Demon is a way of illustrating strict causal determinism. In this system, humans have no free will and their entire outcome is determined by an externality (what we've denoted as 'God'). Laplace (1814) believed that if a being had the intelligence to know the position and velocity of every particle in the universe, it could calculate the future of the universe.

The Calvinist

Bet you didn't know that the comic book classic Calvin & Hobbes was named after two philosophers, did you? We've already covered Hobbes in the Atheist section, now we're on to Calvin. Calvinists would posit a causal system where humans have free will (i.e. have an arrow going to outcomes) but that God determines some very meaningful aspects of life's outcomes. In this Causal Inference world, God is called a "confounder" and something we'd need in our analysis to understand how much of an effect the Humans actually have on their Outcomes; without including God, we would be unable to know how strong an effect humans actually have on their outcomes.

"God Helps Those Who Help Themselves"

I'm sure some more knowledgeable philosopher or theologian out there can give me a good name for this but until that time, I'll use a phrase I once heard a preacher say in a sermon "God helps those who help themselves." In this context, Humans can exert causal effects directly on their lives (and the lives of others), but the total effect of any human actions can only be measured by also quantifying any causal effects that God mediates. There's a whole lot of "stuff" that could be said about that arrow going from Humans --> God but that's for others to discuss and we're using "Acyclic Graphs" so we only allow our arrows to go in one direction from a Causal Inference perspective.

God is Dead! Long Live God! (or Aliens!)

Lastly, we could define a causal system where God caused Humans to be there but then exert no direct impact on the resulting outcomes. I suppose this could happen if Aliens came down, dropped a few of us off on Earth and then left us to our own devices. In this case, God's kind of irrelevant on the question how much to Humans cause their future outcomes, but God or Aliens existed at one point if that's meaningful.

In Conclusion...

This is just designed to be a fun way of merging two hotly debated topics: Religion and Causality. I'll let you decide which system you fall into. Did I miss one? Misrepresent one? Fill me in!

(Disclaimer: DAGs represent deterministic causality, and I don't even believe in deterministic causality! I personally adhere to probabilistic causality as a belief system, but deterministic systems are much easier to depict and rattle off in a blog post. Plus, even if I don't think deterministic causality is correct, that doesn't mean it is not useful to think about.)