[The philosophical meat of this article is under the section “Current threads and questions”. Feel free to skip to that.]

This tweet has had a big impact on my life:

When I stumbled upon it, I had spent about a year being deeply entrenched in the philosophical worldviews of critical rationalism, espoused by the physicist David Deutsch (which he had inherited from the philosopher Karl Popper). One of the interesting things about this collection of ideas is that it’s very popular in specific corners the internet (e.g. in some Twitter communities), but not popular enough to be seen as a “mainstream” philosophy. You would really only hear about critical rationalism if you happened to be very interested in philosophy and interested in Karl Popper specifically.¹

One consequence of critical rationalism’s relative obscurity is that it’s hard to come by serious criticisms of it that aren’t based on some misunderstanding of Popper’s and Deutsch’s arguments. Critical rationalism seems to have some kind of all-or-nothing hold over people: either you are completely captured by it (like me), or it doesn’t make any sense to you at all. Jake’s tweet was the first time I saw someone who took Deutsch’s ideas seriously—who appreciated their profundity and wisdom—but at the same time was beginning to see flaws in them. This was a big deal.

Becoming acquainted

I’d actually heard of David Chapman once before Jake’s thread, sometime in February of 2021, probably having stumbled upon metarationality.com during random Twitter browsing. I was definitely interested in rationality, so I was curious what this metarationality thing was about. But my initial experience with it was characteristic of anyone who comes into contact with a dramatically new worldview: I didn’t get what Chapman was saying, or why it was relevant.² I gave up and continued reading more of Deutsch and Popper.

But then I saw Jake’s thread in April and my curiosity was piqued once again. It looked like there was something really important here that I didn’t get the first time around. Granted, even after this second round of interest it took a while for me to actually start reading Chapman’s essays. I’m not sure why I delayed so much: maybe I was so excited for it and wanted to save the fun for an opportune time, or maybe I was mildly dreading having my existing worldview and convictions be shattered. In any case, after listening to a few of Jake’s conversations with Christopher on the Do Explain podcast, I felt like I had a better understanding of the context from which Chapman was approaching these problems (i.e. his problem-situation!), and I felt ready to dig a little deeper.

Current threads and questions

Jake recommended four articles to start:

I’ve taken a brief look at all of them, only having read the first one in depth (more on that one below). If you’re looking for an explanation of Chapman’s ideas, I actually recommend the podcasts I linked above as starters, and the objects piece.

From my initial read I have these two questions about nebulosity:

  • If the world at the atomic level really is objective and definite (which Chapman seems to agree with?), how is it that nebulosity can arise out of this definiteness? I would imagine that definite objects can only combine to create more definite objects.
    • Here’s an attempt at an answer: when the sheer number of atoms becomes so large, the computational problem of describing the behavior of all the atoms becomes so intractably complex that, any description of the system will have to be a “projection” onto some lower-dimensional space, and that projection will necessarily be incomplete, indefinite, nebulous. Maybe?
  • As I understand, nebulosity means (with respect to objects) that the boundaries of objects are context- and purpose-dependent, and hence there is no single “objective truth” about them. There is no perfectly definite proposition describing the cup on my desk. This seems fine, but my question is: isn’t there still some objective statement we could make like, “for this particular context and purpose, the boundaries of the cup could be established between this range and that range of atoms”? And from there, couldn’t we begin to enumerate, for every possible context and purpose, what the definite boundaries of the cup would be?
    • I realize this not something we could do in practice, but isn’t it possible in principle? Is there not some way to establish, with definiteness, a whole class (potentially infinite) of contexts and purposes, each very slightly different from the other, and for each one, establish a different set of (objective, definite) boundaries between objects in the world?

I’m mostly putting these down as reference for my future self; I imagine the answers are somewhere in the vast troves of Chapman’s essays that I haven’t read yet.

Rough Notes on Abstract Reasoning as Emergent from Concrete Activity

I think I understood parts of this article but many parts of it didn’t click. Listing some insights and questions below, in very rough note format.

  • Insight: cognition is embedded (or embodied). Cognition does not occur in some sort of completely isolated, abstracted, brain-in-a-vat context. The mind and the world interpenetrate. Chapman:

    “The world and the mind interpenetrate” is the essential insight of the “E-word movement.” This is what rationalists / cognitivists / representationalists have the hardest time understanding.

    I sometimes have trouble defining what it means to be embodied, but here’s an attempt: An agent is embodied if its interface with the world is a nontrivial component of its function and existence. This runs counter, for example, to David Deutsch’s view of our brains, as expressed in The Fabric of Reality. More on this here.

  • Insight (really a reminder): active, creative, critical thinking is not what we are doing the vast majority of the time. Most of the time we are engaged in routine activity, which does not require abstract thought or reasoning. Chapman:

    An experienced driver does not have to think about driving, he just does it; and he can be doing something else, possibly requiring abstract thought, at the same time. It is only when there is a breakdown in the activity that abstract thought is needed.

    The reason this is important is: I believe Deutsch takes abstract reasoning to be the defining characteristic of humans. Humans are universal explainers, creative agents, he’d say. While I agree that we are unique in our capacity to explain and reason, I’ve begun to question the idea that creative thought is our defining characteristic. I guess definitions are always a matter of convention; but my point is that there’s much more to our everyday experiences than abstract thought, and seeing yourself as a disembodied-creative-explainer-brain-in-a-vat can lead to confusion and suffering.

    Of course, creative thought is still a huge deal. It’s a really powerful tool and we are blessed to be capable of engaging in it.

  • Insight/question: abstract reasoning is not one coherent module, but rather a collection of special cases. Chapman:

    We believe that abstract reasoning is perforated: it is not a coherent module that systematically accounts for all or even a class of mental phenomena. It is not a general-purpose reasoning machine, as it appears to be, but only a patchwork of special cases.

    What?? I think this one definitely runs counter to a lot of Deutschian thinking. I’m curious to read more about this and don’t understand what Chapman means yet at all.

  • Insight: distinguishing indexical-functional representations from abstract representations. Indexical representations: “the big one” or “the cup”. They point to some directly present sensorimotor input. Abstract representations: “The Special K cereal box on the counter.” Pointing to things that may or may not actually be present; pointing to things that have individual identity. Chapman:

    For indexical-functional representations, things that look the same are the same. All you can do is see if there is a bowl there or not. There is no way to represent permanent objects in Piaget’s sense.

    Some more on the differences between indexical-functional representations and abstract ones:

    Abstract representations differ from concrete indexical-functional representations along several dimensions, and intermediates are possible. They are universally quantified and so capture eternal truths, independent of the current situation. They are general purpose: a very broad range of sorts of knowledge can be encoded in a uniform framework. They can involve variables which must be instantiated in use; this requires notions of identity and difference. They are compositional: sentence-like, in that the meaning of the whole depends on the meanings of the parts. Natural language spans the range from indexical-functional to abstract. Utterances such as “the big one!” (requesting a coworker to pass a hammer) encode indexical-functional representations; those such as “Everyone hates the phone company” are abstract.

    A big takeaway in general from Chapman, and especially from Jake’s conversations in Do Explain, is thinking really hard about representation. What does it mean for a piece of information to be about something? To represent something? Deutsch likes to say that, e.g. a telescope, or a single-celled organism has knowledge of its environment, but does it really?

  • Question: how does abstract reasoning actually emerge though? They acknowledge not having an answer to this. Sad.

  • Insight: language and abstract thought. The development of language is deeply intertwined with the development of abstract thought. Infant usage of language begins as purely indexical-functional before developing into abstract representation.

    We suspect that early language acquisition is intimately tied up with the first development of abstract thought. Natural language provides a bridge between indexical-functional and abstract representations. Children’s first utterances are single words: “ball!” “Mummy!” “hungry!” The production of these highly-indexical “observatives” might very well be directly driven by indexical-functional representations. There follows a long two-word stage, producing utterances like “give ball,” “more cookie,” and “bad kitty.” There is no real syntax to these utterances, but there are a dozen or so different kinds of semantic relationships between the two words to be mastered. The compositionality makes even such simple sentences less than fully indexical. Production of such utterances both requires and drives the development of abstract thought.

  • Question: is this thesis obvious? This question came to me as I was finishing the article and reviewing my notes. Like, of course abstract reasoning has to emerge from non-abstract activity and representation? I’m not even sure now what I thought before I started reading the article. But I think it is indeed a new insight, and that before I had this vague idea that somehow, our capacity for reasoning was deeply ingrained in our brains, was some fundamental property of how we operate, rather than emerging as a collection of special cases on top of concrete everyday activity.

    Also, the fact that these ideas now seem obvious could be a credit to Chapman: it’s always the best ideas that seem counterintuitive at first and then later have you wondering, “how could it have been any other way?”

Footnotes

  1. While this relative obscurity is true of critical rationalism as a term and as a unified epistemic view, it’s probably less true of Karl Popper’s ideas in general. He does seem to have had a broad influence with his theories about falsification and the problem of induction, which still hold sway over philosophers to this day.
  2. An example: I wrote in my notes, upon reading Rationality, rationalism, and alternatives, “can’t say I really get what he’s saying. Chapman seems to create an unnecessary distinction between rationality (formal systems for reasonable thought?) and metarationality which is a system for figuring out when to apply rationality?” I didn’t get why we needed this distinction: it all just seemed like plain old rationality to me.