All gymnastics is mental gymnastics

Various different sources have been mentioning that reality isn't real of late, so I want to consolidate them here. Because I have a terrible cold ahh crap this time it's actually COVID raging at the moment, I'm going to try and limit myself to summaries and possibly interesting questions.

Exhibit 1 is a video interview by the inestimable Fraser Cain with a professional philosopher of science. Now I admit to having a sneaky wariness about such people : philosophers are fine, and scientists are fine, but specifically philosophers of science make me... edgy. How can you possible know what's the best scientific methodology without getting your hands dirty by analysing actual data ? How can you have any special claims on better approaches without having been through the full wringer from grant application to peer reviewed publication yourself ?

At the same time, I feel any effort to reconnect philosophy and science, to get scientists to realise that the data doesn't speak for itself, is to be warmly welcomed. It seems to be underappreciated that science improves itself not just by learning more facts but also by figuring out new methodologies, new ways to avoid fooling itself. The development of improved statistical techniques is an ongoing one : to take any particular value as a yes/no binary choice is generally a very silly thing indeed, but quantification is (and should be) a bloody important core of the scientific process. Science, it seems, is grounded in hard facts.

But oddly enough, philosophy doesn't seem to have come up with much in the way of a satisfying definition of knowledge. Apparently it just isn't necessary to be able to do rigorous science, something I find more than a little disconcerting.

The main part of the video takes this to extremes. As mentioned here previously, we can view the world only through our own unwelt, our subjective sensory realities. Compared to us, pigeons see the world in slow motion, while bats and dolphins use sound much like we use vision, while the star-nosed mole does the same with touch, and some fishes using electroreception. So all perception is mental representation. While I'm happy enough that this clearly debunks the foolish notion of materialism, I'm much less happy about where this leaves those essential hard facts.

In the video the argument goes further and says that even such things like space and time are themselves mental representations. They are not, to use that dread word, real, not things external to our minds, but internal to us. Presumably those mental representations themselves must correspond to something that induces them, but what the hell would that be ? I mean, it's one thing to say that our sensory worlds are incomplete, but quite another to say they're wrong, and still worse to say they're not even wrong... and... space and time themselves ? To quote from an earlier source, the mind recoils from such a stupendous conception !

Anyway this particular philosopher (Sam Baron) is working on what the basic nature of reality might be (as one does), in his case suggesting causation. Rather than causal effects being something that occurs within space and time, space and time, he says, arise out of causation itself. This is pleasingly anti-Humean, but... dude, put it out. No, I don't care if it's legal, you've had enough.

[EDIT :  Exhibit 1a is another video that goes into much more detail about this idea that space and time and mental constructs, going into more detail about how the brain responds to our changing location and marking the passage of time in different ways. This is well worth a watch, although I am not sure if this really means the brain is constructing the notions of space and time or just responding to actual physical phenomena. I'm not sure how we would distinguish these possibilities.]

Exhibit 2 is a nice little piece about the opposing medieval viewpoints of realism and nominalism. Nominalists believed that language only describes reality, that our labels for things like dog and cat and (yes) chair are useful descriptions but nothing much more, with chairs not being a fundamental component of reality. In other words there are no things, only stuff : reality exists, but its fundamental substance is unknown and the best we can do is apply labels to things. Realism, on the other hand, feels distinctly Platonic in saying that general cases like "dogs" and "chairs" (rather than any specific dog or chair) really do exist, somehow.

This... is hideous. Only philosophy can make both the case that dogs do and don't exist seem equally absurd.

Or, it's all trivial and stupid. The mind is capable of generalising from many dogs the basic characteristics of dogginess, and from there it constructs a simulacrum of a general dog and it "exists" mentally but not in physical reality.

Let's go with that before my brain escapes.

Exhibit 3 takes us back to hard physics. Although I still have many questions, this video does the best job of explaining Bell's Inequality I've yet come across. If you have two entangled particles and measure one, does your measurement reveal the previously hidden properties of the other, or does it actually in some way set that property ? Einstein, with his notion of "hidden variables", thought it was the former, that quantum uncertainty only confuses our measurements and says nothing at all about what's - ahem - really going on. 

But experiment suggests it's the latter. That there is some "connection" between distant photons wherein measurement of one directly affects the other, without needing to send another photon : it happens instantaneously. You should watch the video for the description of the experiment lest I report it incorrectly, but the upshot is that the Universe isn't locally real, isn't determined solely by local direct contact. And that's weird.

Personally I would have preferred it if the video had been quite a bit longer. I may or may not try and delve deeper on this one, but not right now. I'm not going to say if I agree with it or not.

Exhibit 4 returns us to language and philosophy. It says that, though the concept has been widely abused, "alternative facts" are perfectly possible, that true facts as we understand them do not in fact exist.

Personally I think this is utter bollocks, which is not helped by the various claims therein which are tantamount to stating that anyone who does believe in facts is basically a Nazi. Look, science is always subject to revision. Its very paradigms change as a matter of course. But most scientists accept the validity of facts and data and - FFS ! - they're not Nazis. Yes, I agree, facts are incomplete and subject to biases, that being truly objective is exceedingly difficult (perhaps impossible for individuals, but maybe not for an ensemble). But to wholly reject them without providing an alternative is... dumb. The mind indeed recoils from such a ludicrous suggestion.

Rather than ending with any kind of conclusion, my main question from all this is : what would it mean to do science in a world without factual, objective reality ? What, if any, difference would it make to suggest that we have no kind of true knowledge of reality whatever, that absolutely everything, even our most basic concepts are naught but mental constructs ?

To me, the revisionism inherent in the scientific process isn't at all incompatible with certainty and facts. We can be certain that given X, a model predicts Y. If we can observe the fundamental aspects of a model directly, then we can be certain it is correct. The existence of atoms and the roundness of the Earth, within our own unwelt, will never be disproven, can never be disproven. But most models can't be observed directly in this way, which is why they never achieve such lofty heights of certainty, why we have to explore the theories more indirectly. And sure, we can question the very basic nature of the observations, but this is the remit of philosophy, not science.

... except where things get quantum. At this point I cannot but throw up my hands and say, "look, I don't know." To question this existence of space and time and locality is a thrilling exercise, but exactly how this helps us do science and make sense of the observations, I have no clue. If we're not going to construct explanations constructed in terms of mentally comprehensible observables, then what are we even doing at all ?