#ironism

[jadagul]

Scott Alexander at Slate Star Codex has been blogging lately about Thomas Kuhn and the idea of paradigm shifts in science. This is a topic near and dear to my heart, so I wanted to take the opportunity to share some of my thoughts and answer some questions that Scott asked in his posts.

New blog post, this time responding to the posts @slatestarscratchpad made recently about Thomas Kuhn and paradigm shifts.  Some of Kuhn’s ideas are really important to me so I wound up having a lot to say; the piece covers traditional blog favorites like modeling, ironism, and arguing about Bayes.

[wojdiz]

🌙🦇

[jadagul]

discoursedrome replied to your post: bambamramfan replied to your post: discoursedrome:...

FWIW I don’t mean capital-U Utilitarian specifically so much as “some beliefs are more useful than others”; my point there is mostly that the /existence/ of superior beliefs is uncontroversial even from a skeptical standpoint, despite that (as you say) we can never agree about which beliefs are superior

I’m pretty sure I disagree with you.  (Would that be sufficient to prove the claim non-non-controversial?)

Like, I believe that some beliefs are superior to others.  In particular, my beliefs are superior to others’.  They are superior according to the standard of being more in line with my beliefs.

My values are (almost?) tautologically superior to other possible sets of values, when evaluated with respect to my values.  But I’m aware that this is tautological, and that I’m not really saying anything there.  And by someone else’s values, their values are superior to mine.  (I sometimes express this as the claim that you can’t say “better” without specifying “better for what?”).

If you specify a value system to judge by, it’s pretty straightforward to figure out which values are better/superior/more correct. (Though not necessarily easily, since factual questions are hard).  If you don’t specify a value system, the question is incoherent.  

[jadagul]

bambamramfan replied to your post: discoursedrome: (cut for length but continuing...

Yeah but that’s not a generalizable tool for “what method produces correct ethical thinking?”

Sure.  But now we’re just exposing the deep flaws in the concept of “correct ethical thinking”.

In order to decide what ethical beliefs are better, we need to know what we think “better” means.  And that means we need an ethical framework.  

There’s no view from nowhere; you can’t judge ethical frameworks without first adopting an ethical framework.  So any time you ask how to do that, you’ve already taken a wrong turn.

You can see this really clearly in @discoursedrome‘s original post, which begins:

At least in the abstract, everyone can probably agree that some beliefs and opinions are more ethical than others, even if only in a utilitarian sense, and that, other factors aside, a person who holds an ethical belief or opinion is a better person than one who holds a less ethical alternative.

And sure, I’ll agree that from a utilitarian viewpoint we can judge ethical beliefs relative to each other.  (From a utilitarian viewpoint, utilitarianism comes out looking pretty good!)  But without (explicitly or implicitly) choosing an ethical standard, you can’t evaluate ethical beliefs relative to each other.

“How do we decide which ethical views are, objectively, better” is a fundamentally flawed question. It doesn’t have an answer.  But “How do we convince other people to agree with us, and make ourselves open to being persuaded, and come to a mutually acceptable basic consensus” does have an answer.  And that’s the answer I gave.

[jadagul]

The important piece to understanding "casual oaths" is to remember that words don't have meanings in themselves; words have the meanings that they create in shared speech communities. If everyone knows your words to mean X, even if the literal content is Y, then your words mean X.

This is why it's not a lie to tell a fairy tale. Or, you know, to engage in self-deprecating humor, or sarcastically say "and obviously no one would ever do anything that silly". These are all denotatively false statements, but they aren't lies because they aren't interpreted to mean their denotation.

If you stand up and take the Hippocratic Oath at the end of a MD graduation, the meaning you create isn't "listen to these specific words and hold me to them". The meaning you create is "I have just graduated from medical school and am participating in the ritual that accompanies that graduation." And subsequently "violating" that "oath" doesn't say anything about your trustworthiness, because no one was interpreting you as making a serious promise to begin with.

(In the limiting case: if you're acting in a play on stage and your character swears an oath, you would never expect that oath to be binding on the actor. Even though the actor just stood up in front of a thousand people and spoke the oath!)

And you ask "But then how can we distinguish real oaths from performative oaths?" And the answer is that there's no shortcut because all communication is contextual. But we should take seriously promises that are offered seriously in a serious context. And we should take performatively promises that are offered performatively in a performative context.

There's a corollary question, which is where this started on the subject of marriage vows. How can you force other people to take your oath seriously, rather than performatively?

And the answer is, you can't. Meaning is created contextually through communities, and if your community chooses to interpret your words as meaning one thing you can't stop them. (Much like you can't use a slur and then explain "but I mean this lovingly!" You might genuinely mean it lovingly; but that doesn't matter if that's not the meaning it creates for your audience).

In your marriage, you can make a serious oath to your spouse, and both of you can intend that seriously, and it will be a serious oath. As binding on you as your conscience and your spouse dictate.

You can make the oath in front of a community that chooses to accept that oath as serious. And then the community can hold you to the oath, and bind you to it, as much as you abide by that community's norms and wish to remain a member of that community.

But you can't take a community that does not wish to accept your oath as a serious one, and force them to. Because, as the tortoise said to Achilles, you can't force people to accept a meaning they choose not to accept.

[jadagul]

Finally read Scott’s Crystallized Metaphysical Essences post. It took me most of the post, but I figured out what was bugging me about it.

Scott presents a ladder of levels we can handle things at:

Explicit models

Emotional reactions

Reified metaphysical essences

Endorsed values based on the essence

and treats these as basically a hierarchy: the top is more sophisticated, and the bottom more abstracted and visceral.

I think his hierarchy makes sense as a causal pathway: the benefits we can see in the explicit model (evolutionarily) lead to emotional reactions, which lead to us reifying our heuristics as a real essence of the thing, and then expressing values about those reified essences.

But I don’t think that as experienced, they stack neatly into a hierarchy the way Scott describes. In particular, I’m strongly in favor of the top and bottom, while opposed to the middle two.

Obviously, if you can build a good, sophisticated, accurate model of something, that’s good for understanding it. And you should use that model if you have it and it works.

And on the other hand, Scott is completely right that our values have to come from somewhere; “wanting people to be happy and healthy” is just as evolutionarily rooted as “wanting green space” or “wanting revenge”. Despite Sam Harris’s philosophical naivete, you can’t remove all possible value judgments from your ethical code.

But metaphysical essences don’t exist. That’s why they have to be reified in the first place. No one ever talks about reifying a chair. And treating metaphysical essences as things that matter leads to a lot of both confused and generally shitty thinking. To my mind, taking metaphysical essences seriously is basically Public Enemy Number 1.

Emotional reactions do exist, obviously. But my experience of the world is of emotional reactions that can mostly be controlled and retrained; I want people to make their emotional reactions subservient to both their endorsed values and their explicit models. (Yes, I know that this is much harder than I’m making it sound. I can dream big).

But this means that I reject Scott’s hierarchy as, well, a hierarchy. I think you can keep the top and the bottom while losing the middle layers, and that’s what I want to push for.

[jadagul]

Two thoughts on religion and atheism:

1) Under most modern Western post-Enlightenment versions of religion, you don't expect to make radically different observations regardless of whether religion is "true" or "false".

You can, if you choose, make fun of the "God can't/doesn't give definitive proof of his own existence, because that would remove room for faith and/or free will" argument. (And if you do it well, I will probably laugh). But that's the argument a lot of sophisticated proponents of religion are making explicitly, and which most people more or less tacitly accept.

And at that point, when you argue about whether religion is true or false, you're not arguing about what observations you expect to make, so much as how to explain them---about what story to tell about your life.

(In fact, the Rational Christianity piece I linked above makes basically the same point. It suggests that there's no such thing as absolute proof that god exists, because any concrete observation will always admit some other explanation.)

And that has two implications. First, these arguments are hard to resolve: "what story do I prefer" is a more personal and subjective question than "will god magically light this altar on fire." Which is why these debates mostly go nowhere.

Second, people speaking from one interpretation tend to be completely incomprehensible to people speaking in the other. If your worldview/explanation includes god, then much of what you see is evidence of god's presence. But if your worldview doesn't include god, then you see no evidence and can't come up with a reason to believe him.

This would be a fantastic example of the theory-ladenness of observation, if it weren't so controversial and didn't impinge on everybody's core philosophical commitments.

2) There's a deeper reason that religion always feels kind of ridiculous (and perverse and depraved) to me; but most non-religious people are vulnerable to exactly the same critique.

Fundamentally, religion only makes sense if you want an external source of meaning. God provides order and meaning to the cosmos, and it's important to worship him and abide by his judgments.

But the idea of external meaning is absurd. The outside world can't make things meaningful to me unless and until I decide to care about them, so any attempt to externalize meaning is just kind of goofy. (And since I think abdicating your responsibility to create yourself, and your own meaning and values, is deeply Wrong and condemnable, I also find it perverse).

If every factual claim in your scripture of choice is literally, plainly, factually true, that still doesn't give me any reason at all to care what god thinks. It's not possible to make someone value something, or care about something. Values are always a choice.

(This is somewhat ignoring the "do what god says or else option". But that still doesn't give me a reason to believe things---just a reason to superficially comply. And most religious people don't think that God is really Anthony Fremont).