Atheism - An Epistemologically Tenable Position? | Overwhelming complexity

AI translated using Gemini 3 Flash

This text is definitely nothing new and is probably also less well-sourced. I hope you find it interesting nonetheless.
As a counter-thesis, one could read Richard Dawkins, for example, whom I personally find to be self-dogmatic. I should also add that I have not read any primary sources for most of the things mentioned here so far.

Some definitions for the purpose of this post:

Atheism: The conviction that a divine being of any form does not exist, or that it is wrong to assume such an existence.

Distinguished from this is Agnosticism, which makes no statement about whether a divine being exists or not, thus moving neither toward atheism nor toward any form of faith or spirituality.

The definition below for “empirical” used here ultimately goes back to Karl Popper, who postulated the possibility of falsifiability as a criterion for a scientific theory.

For the “classical” empirical method to work, repeatability and thus control over the relevant factors of an experiment are essential. Conversely, this means that a divine being that does not want to “play along” with the experiment is not detectable. No verifiable (falsifiable) hypotheses can be set up; the question of a divine being thus lies outside the scope of science1.

This results in infinitely many hypothetical worlds that are indistinguishable for us through experiments. Even more serious: even without the possibility of any “divine being,” there are infinitely many models2 that cannot be distinguished by our experiments so far, and beyond that, numerous interpretations of these models3 (where theory and interpretation are not sharply separated).

Conflict with Occam’s Razor and Untestable Gods

To select from a multitude of non-falsifiable models one that is considered the “most probable”4, additional assumptions are needed. A very popular principle here is Occam’s Razor. It refers to the rather abstract concept of choosing the “simpler” theory – typically the one that requires fewer or less pronounced assumptions. This principle can, of course, lead to theories being excluded that are closer to the underlying “real” truth5.

I would even argue that the assumption of Occam’s Razor together with the assumption of a divine being that does not intervene repeatably is necessarily inconsistent6:

  1. Model 1 (with God) cannot be empirically falsified, while Model 2 (identical, but without God) would be falsifiable – according to our assumption about the divine being.
  2. If we assume Occam’s Razor, we would choose Model 2 because it contains fewer assumptions; however, this is in direct conflict with our assumption of a divine being.

This doesn’t make Occam’s Razor a bad assumption at all; it is, for example, very useful for deciding in which direction to continue research or which content should be disseminated in science communication7. It is important to note that Occam’s Razor itself is only an assumption that does not necessarily lead to the underlying truth.

The problem is that Occam’s Razor is often cited in debates to “refute” God. In doing so, Occam’s Razor is just as carelessly a belief (or an assumption) as the assumption of a God, which atheists often dismiss as mere assumption. There is no deeper reason for it other than practicality and simplification – just like faith, which can lead to comfort and community even without the existence of God.

Conclusion

Fewer assumptions and thus more general validity are achieved if one adopts neither Occam’s Razor nor God (or belief) and instead takes an agnostic stance. For concrete belief systems that have falsifiable hypotheses (e.g., the effectiveness of talismans or homeopathy), the picture is obviously different. At the same time, it does not exclude occurrences of divine appearance, as these are inherently non-repeatable and thus also do not represent a testable hypothesis.

Update Jan 22: I received feedback that it is not sufficiently documented that (many) atheists make the restriction in their world view of excluding a God in any form. To provide this, I’d like to point to, for example, Atheism: A Very Short Introduction, which explains that most atheists are naturalists, “meaning that the world is to be understood as a purely natural phenomenon.” - Wikipedia

Footnotes

  1. One (or several) potentially existing divine beings apparently do not participate in experiments (e.g., a conversation) because they cannot or we haven’t found the right experiment yet. ↩

  2. For example, all super-symmetries and string theories discussed in modern physics. ↩

  3. In quantum physics, for example, the Copenhagen interpretation versus the many-worlds theory; previously equivalent, experimentally indistinguishable interpretations like the Dirac sea or the luminiferous aether. ↩

  4. For the purpose of prioritizing research or deciding what is taught and communicated. ↩

  5. The simplest example is Flatland8: A society of beings that can only perceive a two-dimensional world, but in reality live in a three-dimensional one. The book is a recommended satire that illustrates the limitations of our imagination when we try to grasp higher dimensions. ↩

  6. Mathematical term for “contradictory” (inkonsistent). ↩

  7. Noting that we are not sure is a rather boring explanation for people who have a basic interest in the current state of research but not the time and capacity to look at a hundred different models. ↩

  8. Also mentioned on the Influential page with media that impressed me. ↩