[FRIAM] projection propaganda

[Santafe]

> On Jul 17, 2025, at 23:17, Prof David West <profwest at fastmail.fm> wrote:
> 
> Metaphor became a research topic for me when I was doing my Masters thesis in comp sci (AI) and was prompted by the reciprocal metaphor: "the brain is a computer / a computer is a brain." This metaphor dominated the conversation across multiple fields including computing, psychology, and cognitive science while attempting to intrude on disciplines like cultural anthropology. The metaphor was used as if it was a lexical term, and computer scientists like Pylyshyn insisted that it was. Sure, the substrate was different, silicon and meat, but that was irrelevant.

The example DaveW makes above is -- if my understanding of it roughly aligns with the one he intends -- just the same sense I would use for several cases that I think are problematic in Origin of Life, and what I assume that Quine had in mind, as a continuation of the whole-system dependence of interpretation that Neurath already understood, that Carnap came around to, and that Quine and Carnap, in their “debates”, didn’t really disagree on, but said in different enough ways that people seemed to think for a long time they disagreed.

An instance of the OoL problem construction that to me seems much like the one above is the way “RNA-first” versions of the “RNA World” picture are adopted with the “motivation” or “ “ justification” “ that if true, they would “realize Darwinian Evolution in populations of single replicating RNA molecules”.  

All the levels in the Glen/DaveW exchange seem to be present here.  There is an a-semantic layer of formal representations, and while most people who say “Darwinian evolution” have no interest in being careful enough to tell you _just_ which operational meaning of the term they intend, there are such a-semantic formal unpacks, and other people can articulate them, and you can choose which one you want to work with.

For the “evolution” case, as with the brain/computer case, there are ways you can try to map tokens in the a-semantic model (for which, I agree with Glen, it is neutral on anything beyond what it itself constitutes as a system of tokens and manipulation rules) onto aspects of the molecule system or physical-brain-and-lived experience system.  With various affirmations and violations, those mappings can be sort of held to in either case.  There are some event-sequences or object-instantiations that, if painted with the formal markings, will carry them along for a while within the transformation rules that the formal system would stipulate.

But in the RNA-Darwin case, it is very much _not_ because of how well the whole thing hangs together that the proponents adopt it.  There is a whole forest of taking-things-as-given conventions for applying the Darwinian formal system to the lifecycles and contexts and proliferation of organisms, which are central to how it is used where it is used appropriately in population biology.  In a better science (than anything we currently have), the reason it is possible for any of those kinds of objects to exist, the explanations for why those kinds and not some other kinds, or no kind at all, etc., would all need accounts.  The use of the Darwinian system brackets them off as work for somebody else to do some other time.  The OoLers very much want that aspect of Darwinian practice to be their own MO as well.  They want to assume all the same object-terms, pattern-terms, etc., and just take all of it as given and not explain whether any of it makes sense as a term system for actual objects or actual events that have any reason they should arise in the world.  The difference is that there actually _are_ naturally-occurring organisms in our biosphere.  So even if we don’t understand them, it is probably safe enough (or as safe as anything else) to refer to them and then do some other work with the referents.  The RNA-firsters have no comparable observed phenomena to refer to.  They may be able to engineer molecular systems that will instantiate some patterns, but the argument that those systems are _in any sense_ appropriate “models” of the natural planetary settings they want to project them onto doesn’t look like it has any good basis, and in any case they are not the kinds of people who feel like they should have to justify any of their preferences with arguments.  So this is where the usage becomes very overtly metaphor.  They want to take a context from somewhere else and project it here, in spite of all the reasons they and everybody else see that it may not port, so that they can use a little of the formal manipulation to complete some steps in a toy system.

The fact that it is a metaphor is not something I would criticize in the OoL case.  Most things used there are metaphors of more or less this kind, in some aspects or degrees.  But the fact that it isn’t a very good metaphor, and that they have no interest in dealing properly with its many problems, is something I will criticize them for, because it doesn’t look like a good way to use metaphors (necessary) on the “fringe” of science, as Quine puts it, to get to things that could take on their own real meanings someday as proper terms within its empiricist fabric.

Eric