[ 23/May/21 ]
First I need to give the argument for something like cause and effect.
If everything were completely random, if there were no cause and effect of any degree or kind, then there could be no structure over time. Any structure that did emerge would be ephemeral, disappearing as fast as it appeared.
So given that we are here, and we are asking and answering questions, then there must be something that allows some approximation that is sufficiently close to cause and effect at some levels at least in the sorts of contexts that we exist in.
The question now is:
Does reality require that cause and effect be simple and universal at all levels and all contexts; or is there evidence that something else may be present which allows sufficient cause and effect at the levels and contexts that we exist at for those levels of structure to persist?
In this sense, our very existence guarantees that there must be something that very closely approximates cause and effect at the scales we normally deal with (which for the purposes of biology includes the atomic/molecular level on upward, at least in the sorts of contexts that we evolved in and currently exist in).
Are atoms stable everywhere?
We have strong evidence that in the heart of suns, and in the very early universe, conditions were such that atoms were not stable. We don’t exist in those places, we exist here and now (in a relatively cool and stable outer spiral arm of a galaxy, orbiting 93 million miles from the nearest sun – a member of a rather stable class of suns).
When we look reasonably closely at the evolutionary strategic environment that seems very probably to have given rise to our existence; then we can see successive levels of cooperative systems slowly (in successive very narrow sets of contextual constraints) emerging and building to a level that our level of complexity could emerge in a set of contexts. At every level – cooperation was fundamental to both emergence and continued existence, and cooperation is in most contexts very vulnerable to exploitation by cheating strategies – and thus requires an evolving ecosystem of cheat detection and mitigation systems to survive (every level).
Another important thing to note when looking deeply at that strategic context is that the need to respond rapidly in many contexts will instantiate strong sets of biases in neural networks to prefer simple models over more complex and nuanced models. In many contexts time to compute an answer is more important than the accuracy of that answer. The last to move is often the first one eaten.
Thus there is a strong evolutionary pressure to produce brains with a strong preference for simplicity at every level of abstraction.
In many brains, that pressure is likely to preclude even the possibility of considering highly abstract and nuanced structures. Those brains are so heavily biased for simplicity that everything does actually resolve down to simple binaries, like True/False, Right/Wrong, Good/Evil. Getting to nuance beyond such simplicity is often difficult and in some contexts is just not going to happen.
Thus for someone like me, who has spent over 50 years fascinated by behaviour, biochemistry, systems, uncertainty, chaos, quantum chemistry; dealing with fundamental uncertainty is normal, and for me most simplicity like “True/False” in respect of reality, is in the class of “useful simplification, but highly unlikely to be how things actually are”.
For me, the evidence is beyond any remaining shadow of reasonable doubt, that what we experience as classical causality is real and reliable in most contexts at our scale, and it is build on multiple levels of chaos within probability constraints.
Constrained chaotic systems are unpredictable in the specific, but become very predictable when large populations are considered. When thinking in terms of Planck time units, even a single tick of an atom in an atomic clock contains more Planck time units than there have been ticks of an atomic clock in the age of the universe. Such large collections tend to behave in very predictable ways in most contexts, but not all.
Thus we can get the sort of regularity required to build structures like atoms and life, and computers, and yet still have conditions where there can be sufficient difference between classical “cause and effect” for there to be meaningful separation and choice in some contexts. And that is a deeply complex idea that requires years to explore to a useful depth.
So our existence as physical entities demands something that very closely approximates cause and effect; and our existence as moral agents demands something more nuanced that does allow for degrees of independence in some contexts.
Fortunately, we seem to exist in a reality where both things can in fact be true, and there is no simple systemic explanation that makes that obvious to all, it really does require a lot of work, and most people have neither the interest nor the time to do that work, so simple approximations are required (like independent realms – duality – which while they may not be true in any absolute sense, do deliver useful approximations in practice in most contexts).