I was going to write a point-by-point response to this: The New Epistemic Deal | Two-Phase Cosmology But any given point looked pretty heavy on its own, so I decided to launch the threads one at time.
This one is on point 3: “Epistemic structural realism is true.”
First of all, what even is epistemic structural realism? From the Stanford Encyclopedia of Philosophy:
Most of that article is beyond my comprehension, but this sentence jumps out: “These contemporary debates recapitulate the work of some of the greatest philosophers of science of the late nineteenth and early twentieth centuries.” That is, there a debate going on. And it is not a new debate. That tells me that evaluating this entire field as “true” or “false” is far, far above my pay grade! Just check out the reference list on the Stanford Encyclopedia article. Anyone here ready to sort out “true” and “false” in every single one of those references? Not I! Smart people differ on the finer points. I’m perfectly happy just being a spectator for that discussion!
On a cruder level, what might a normal-ish person understand by “true”, “real”, “knowledge”, etc? To me, consciousness is predictive. I do A. I expect B. If B follows A often enough, I starting filing A → B into the “true” category. As long as A → B keeps working out as expected, that proposition counts as “knowledge”. But like Taleb points out, any turkey has iron-clad knowledge that farmers were put into this world specifically to feed turkeys on a daily basis. Great “knowledge”, until the day before Thanksgiving! It turns out there is an unexpected C and that can overrule A → B. Turkeys learn that the hard way.
So yes, I have a sense of reality based on patterns and relationships that have been relatively consistent so far. I’ve also seen enough things come and go, collapse, transform, reboot, etc. not to lean too heavily on any given pattern as being TRUE or REAL in some absolute sense. My deeper sense of a reality that is always going to be there somehow or another falls more into the “faith” bucket than it does into the “knowledge” bucket. It comes from being in the world more than from having opinions about the world.
Epistemic structural realism (ESR) is relatively simple to understand. It says this:
(1) There is a real/objective world beyond the minds of humans.
(2) It is possible for us to have knowledge of its structure.
Ontic structural realism (OSR) makes a much stronger claim: structure is all there is. ESR therefore avoids the problematic claim that structure is all there is, but leaves open the question of what objective reality is “made of”. This is imporant, because it avoids the assumption of naive realism (the idea that objective reality is the same as the material reality we directly observe within consciousness). Objective reality could be made of matter, or information, or “some other kind of consciousness” (objective idealism), or anything at all. It can therefore be an uncollapsed wavefunction (which clearly has structure, and we know what it is).
What ESR rules out is anti-realism: there is no objective reality, or we can’t know anything about such a thing (not even its structure). It is therefore incompatible with scientific instrumentalism (the idea that science is a useful tool, but does not aim at truth or tell us about anything real).
ESR establishes that there is such a thing as truth and reality, in the correspondence sense. Not just “my truth” and “your truth”, but a shared, objective truth. It is minimal and therefore the most defensible version of that idea.
I think the term “absolute” is unhelpful here. What does “absolute truth” mean? If truths refer to structure, is that “absolute” or not? What is a non-absolute truth?
This is a slightly different point. It’s not about realism/anti-realism but about the perils of inductive reasoning. It feels like every time a pattern repeats, we should be more certain that the pattern will alway repeat. But this doesn’t work, for the reason given above (which is more usually associated with Bertrand Russell).
This is very important when considering scientific claims such as the one made by @Justin yesterday, in another thread:
Science has spent the last several hundred years, time after time, proving how insignificant we are in the universe. To relegate all this in the opposite direction, as your theory does, is like hurtling back to the age of the inquisition.
Yes, science has spent 400 years making discoveries that suggest life is not cosmically relevant, and the Earth is “nothing special”. Does it follow that this will always be the case? No.
There are countless other examples of patterns which repeat and repeat and repeat…until they don’t. Therefore inductive reasoning is always suspect. This argument has sometimes been used to back up the stronger claim of “therefore all scientific knowledge must be assumed to be provisional, and might turn out to be wrong”, but here I think the reasoning fails. Are we ever going to find a kind of water that isn’t made of hydrogen and oxygen? Or that Saturn is actually larger than Jupiter? No, which tells us that our certainty does not have to be based on inductive reasoning about repeated observations. Instead it can be based on the generalised claim that science actually works, and that if it isn’t because it is capable of delivering truths about reality, then we have no good explanation for how or why science works. Presumably it is not a miracle.