remember arguments

When I was younger---in college or in grad school---I was sometimes conflicted about whether I should prioritize trying to get to correct conclusions, or conclusions that I knew how to justify. I think for some reason I thought that those were in tension.

If you had to choose, it seems like you should prefer correctness. Justifications are often motivated reasoning, something you rationalize after the fact. People come up with all kinds of justifications for things that somehow manage to beautifully align with their tribal affiliations, so just having a justification doesn't imply that the belief is rationally well-founded.

In many cases, people decide what is true based on pre-logical cognitive intuitions, and then go look for ways to justify it. In such cases you can save a lot of time, and a lot of cognitive resources, by just not doing the bullshit justification step. In life we have to make a lot more decisions than we can reasonably be expected to come up with individual justifications for. I don't need to be able to, like, respond to every argument that's out there about why I shouldn't vote for the candidates I'm gonna vote for. Because I know that a lot of those arguments are bullshit. You know I know that there's a landscape of generating arguments for positions that you already believe in. This is what a lot of political organizations do. And it's just not a good use of my time to engage with all of those arguments, everything in that landscape.

So that was the position I held. I related to that also to the idea of sufficient statistics. Like the great thing about distributions that have sufficient statistics, is that you can throw away most of your data. You don't have to do computation, and store some giant data set. You just summarize it as a sufficient statistic, a sufficient statistic can be something like a mean and a variance. Yeah, I'm rambling here obviously the statistic is whatever it is for the distribution, but you could imagine summarizing your belief about a question by a location and and uncertainty. On a scale of zero to one, roughly what probability do you assign to this being true. And then you need maybe an uncertainty over that probability, so that you know when new evidence comes in, you have some idea of how how sharply to update.

• Yeah, so there's probably in this propositional case there's a beta binomial model going on.

• The thing that I've realized now is that there's another angle to this. Justifications are not just things that you bring up for bullshit reasons to contradict somebody else's just bullshit justifications. Justifications are one of the few ways that getting a truth on a question that you're not intimately familiar with. If I go if I go into a new area, and I read the arguments for different positions and one of the arguments feels stronger and responds better to the other arguments, then I'm gonna tend to believe in that argument. And my belief in some ways will be more objective and more rational than the belief of people who might have been in that area for a long time, following their intuition, and never stepped back to question the larger story.

• So justifications are a way of coming to truth and really crucially, they are a landmark you leave for your future self. If I think there are things that I thought that I think strongly now, but I forgotten that justifications. I know that at one point, I thought strongly, and I had good reasons that I never articulated, maybe because they were based on concepts that I didn't know how to articulate, that's a whole 'nother discussion. But for whatever reason I now have a lot of opinions that I strongly believe are right, because I know that I used to believe they were right. And that I had collected up a lot of arguments and evidence that had pushed me into that position. There was a long accumulation of evidence. I have the explaining away property that I know that I was thinking to myself about forgetting arguments for things, and so the fact that I thought forgot the arguments for many positions doesn't mean that I never had them, or that they weren't good. But now I'm in the position where I often can't change my mind. I feel stuck in a certain mental place because I have beliefs that I can no longer fully test against new arguments that come in.

• Because I don't remember the arguments for a lot of things that I believe. I can't defend them to other people. And yet, I know that I can't change them, because these are things that I really believe in.

• And that also leaves me stuck. Because if I did want to change my mind about something that I believed, many years ago; statistically, some of those things probably are wrong. They are mistakes. But I now don't trust myself that I've come up with sufficient weight of evidence. By getting rid of the justification. I've really gotten rid of my ability to judge the weight of evidence. It turns out that a number, really wasn't enough. Like a beta distribution over belief really wasn't enough.

• So I guess there are two conclusions: there's a takeaway for younger people, and there's a takeaway for me now. I originally thought about this in terms of the takeaway for younger people so I'll say that first.

• The takeaway is that justifying your beliefs is really a landmark for your future self. You'll only be able to preserve beliefs over time if you leave justifications of them, because your future self will essentially be a stranger. You have to convince your future self. You can't just decide to believe something and then forget why.

• And this does connect with further discussion about, for example leadership. And if you're a leader, then you have to convince, not just yourself, but other people. And there may come a time when you believe in something and the best way for you to act on that belief is to be a leader. If there's a project that you really believe in, then the best way to make that project happen, is to find people to help do it. So, you also have to convince other future people. But if nothing else, convince your future self. Okay.

• So what are the takeaways for me now? I think it's helpful that I'm writing a Roam notebook. And that I'm leaving these recordings, because they are a way of articulating and justifying some of my thinking. And recording those justifications for my future self.

• So I think that's useful. I need to distinguish between arguments as political theater and arguments as a profound path to truth. It's a fine line.

• But the complicated thing is that they're both arguments that are profound paths to truth that I need to listen to. And to respect arguments and applaud political domain. I don't need to listen to.

• And I do have a sense, often, of whether an argument is being made logically, rationally, disinterestedly, or if it's being made as an attempt to further an agenda.

• Now being an adult is complicated because I had now often actually have to make arguments to further agendas. I try to make sure that their arguments that I actually believe. But I know that without the agenda I would be making different arguments.

• So I do understand that arguments are speech acts, in a sense, and working in a world where you deal with them as speech acts feels different than being in a pure world where we can think about them only as statements of truth. Even as even as speech acts of course arguments do still have the force of pushing a truth. That's why they're powerful acts.

• I should remember that even though logic is very flawed, logic can be built on flawed concepts, models are always wrong, therefore, arguments are always wrong. Despite all of this I believe that arguments are the are the best we have, in many cases.

• I want to be a person who can stick up for my beliefs, who can defend my beliefs with speech acts if needed.

• I want to be able to be responsive to argument. I want to change my mind when argument demands it, when I noticed that I'm wrong.

• Sorry, I, I lost my train of thought. I do think there's a rich area here. I think this connects deeply to my feelings about academia. Part of my disillusionment with academia, has been how little bit is what I would see as an objective search for truth, and how much of it is roughly political. I never wanted to engage in the political parts. Maybe through the same part of me that doesn't want to engage in protest. Maybe it would be good for me to engage in protest. Okay, I'm gonna stop now. I think I'm rambling, and I probably don't have any more coherent points to make.

• Transcribed by https://otter.ai