mcgann_letterhead

#62 On Foundations

Dear Jamie,

Congratulations for drafting the final chapter of your thesis! You’ve already managed the whole thing with much more grace than I did.

I’m writing this bit before I read your letter in order to explain what my original assumption was about your policy. I did say you’re dismissing skyhooks because of “human, political concerns (sic :S)”. To explain it in a way that only you and I are likely to understand, that mainly meant “protecting the dynamic aspect of our society”. This is because, as you’ve always pointed out, belief in phony bad skyhooks causes real bad problems. Indeed they do.

I assume this “no skyhooks” policy, whoever it is held by, is the result of an escape as far as possible from supernatural and religious explanations, while desperately holding on to justified true belief as one runs. I know you’re fuming right now, but honestly I see a strain of it. Let me explain how. When one discards theories or concepts as unphysical, one is making an appeal that the thing’s “essence” must be of the right stuff. Given two equally well-respected, functional explanatory theories, one physical and one not, in your mind the physical one is allowed an additional verifiying property. It’s in a sense more confirmed by an observation of its essense than the other. The drive to require everything be physical is the result of historical-scale induction; the most successful theories of the past have been physical thus all future successful theories must be physical.

*reads letter*

I agree Dennett’s anti-skyhook stance is in the minority, but I think it’s a wrong one. The majority you refer to are correct in assigning status of real to some causal, unphysical concepts but it’s mostly a coincidence because they’re assigning reality for what are, in my opinion, the wrong reasons. The majority are explicit or closet verificationists, who claim reality to abstract things just because their verificationism gives them the false power to assign reality to experimentally confirmed claims.

Your rule:

if a theory contains entities that don’t exist and has them playing a causal role, we can razor that shit off.

Is in direct disagreement with Popper, Deutsch and by extension, me. Popper is generally against any inductive assumption like physicality. Deutsch, in Fabric of Reality, because of his focus on explanation, says that if something has a causal role we must accept its existence. This is because our best explanation tells us that the thing has an autonomous existence.* Real is what can’t be razored off without losing insight into the world.

Nozick’s thought experiment is relevant here, but not for the reasons it was made famous, but more for its assumptions. The choice is a false dichotomy, not because of technical or even physical feasibility, but that there is no qualitative change in ignorance between the two situations. The idea that “the real” provides some surety in return for a lack of happiness is false. There’s no such payoff in certainty. The comparison is really: a hedonistic ignorance-laden artificial world vs an indifferent ignorance-laden natural world. We are so almost completely ignorant in reality. Actually, by virtue of the limited nature of artificial things, we would probably know more about the artificial world than the real one. For me the answer to the riddle is that reality has an infinitude of real problems that some virtual reality is too unsophisticated or too happiness-blinkered to reproduce, and that’s why the real world is more meaningful.

After reading your letter I’m ready to stand by my original my assumptions. Might I suggest that your skyhook** hatred comes from the same place as the average Joe who dreads a world without skyhooks? You’re both looking for a foundation. The assumption one gets from an unexamined reality is that a skyhook like God is a foundation, but a little knowledge shows that isn’t true. With a little more knowledge we learn that physical explanations have worked very well, and it’s probably safe to assume we’ve hit foundation there.

But with some knowledge of knowledge, we must come to the conclusion that truth is at least forever out of reach or at worst non-existent. In both cases all we have are our theories, and we can take the best ones at their word, or we can mangle them for having the wrong “essence”. If you modify a “best” theory, you get, by definition, a worse theory.

Skyhooks are no foundation for a conception of reality, but neither is physicality. When it comes to knowledge, there is no foundation, it’s tentative concepts all the way down – if there were a down at all.

Mat

 

* In FoR Deutsch uses Hofstadter’s domino thought experiment to show that a number “being prime” is the best explanation of why his domino computer comes up with the output it does. Thus the nonphysical “primality” causes a domino to fall or not. When we test “primality” in other media (computer progams, proofs on paper, cicada life cycles) we find it “kicks back”, it doesn’t get explained away or reduced into anything else. Things that can’t be explained away deserve our attention, not ignored. But I’m loath to talk about mathematics too much because it raises a whole host of distracting things from our current conversation.

** It’s OK for now to keep using the word “skyhooks” but even that assumes some kind of gravity in the realm of theories that axioms must prop against.

 


Also published on Medium.