Pete Wolfendale

@deontologistics

Wandering philosopher. Purveyor of Platonic heresy, Kantian computationalism, and Hegelian minimalism. (he/him/it which speaks)

193 Siguiendo1064 Seguidores1228 Publicaciones

X YouTube Bluesky

¿Es tu perfil?Reclámalo y deontologistics.dirty.social es tuyo.Reclamar perfil

Pete Wolfendale@deontologistics· 5d

This is going to require porting the interaction from the other site over so that it makes sense. I'll start with screenshots and then shift to plain skeets for convenience. ALT text is provided.

5. N: I agree with (a) and (b), but the fact that a certain formalism is better/a generalization of the old one is not enough to justify reading metaphysics off this formalism.

6. N: To use your own example with Euclidean space, if we did not discover a physical phenomenon that needed Riemannian geometry, it would justifiably remain simply a mathematically interesting generalization.

7. N: So what are the facts about metaphysical modality that are captured by the coalgebraic approach but not the set-theoretic model theory?

8. P: Logically speaking that all the standard problems with identity in non-intensional contexts are a result of treating indiscernibility of identicals as valid, and metaphysically speaking that the deep reason for this in the empirical case is the asymmetrical structure of time.

9. P: In essence, the way that information flows through time, either being lost or accruing more, shows us that all the standard ways of managing temporal modality are something like model-theoretic epicycles, trying to make objects in time act as if they were spatialised.

10. P: Lewis's perdurantism is *exactly* this. It leaves him in a position where he has to reconstruct empirical objects from bisimulation over behavioural traces on branching time, but in a way that ignores the causal structure of what generates the traces, i.e., dumping information.

11. P: If you want a simple analogy, look at automata that produce only binary streams. You can individuate them either by quotienting the resulting traces (i.e., streams analogous to possible worlds), or by looking at the abstract causal structure of the automata that produce them.

12. P: Writ large, this is quotienting the complete histories of possible worlds considered as total behavioural traces (counterpart relations as equivalence classes of observable behaviours), versus the internal states of the individual objects within this world and how these compose.

13. P: So, to put this in other terms: Lewis's Humean view of causality is a fairly to appreciate the compositionality of causal structures. It's essentially Spinoza vs. Hume & Leibniz. The modes compose in such a way that the new modes have behaviours irreducible to their parts.

14. P: Also, I would add that this is a Twitter thread, whose weight is somewhat less than even a pre-print. The burden of proof here is not a complete alternative metaphysical theory but a pathway towards one constituted by sufficiently rich leading questions.

Pete Wolfendale@deontologistics· 5d

If you want a partial explanation of the Hegel stuff, it's all about how metaphysical and epistemic modalities interact, and there's a section from my ongoing Twitter essay on Phenomenal Concepts that covers it just about:

Quine famously said that "meaning is what essence becomes when it is divorced from the object of reference and wedded to the word", and this well captures the modus operandi of the anti-metaphysical and modal skeptic tendencies of the tradition prior to the modal revolution.

371. The semantic externalism that emerges in the wake of the revolution inverts this sentiment: "essence is what meaning becomes when it is divorced from the word and wedded to the object of reference", downplaying description and emphasising causal contact between word and world.

372. But we might better see these as synergetic perspectives on the same process: the questions 'What do we *mean* by 'steel'?' and 'What *is* steel?' play different roles in inquiry, one calibrating our conceptual machinery, the other pushing us into the modal wilds beyond.

373. Of course, this difference is precisely what two-dimensionalism aims to capture: modal variation in what picks out the object of reference versus modal variation intrinsic to that object. But it does so by idealisation, obfuscating the concrete dynamics of the process of inquiry.

374. The correct epistemic intermediary between meaning and essence is the notion of *appearance*, and Hegel captures its dynamic character best when he says that appearance is the appearance *of* essence. For example, Newton's conception of gravity is how gravity once appeared to us.

375. There are *phenomena* that Newton's theory still describes adequately, but we're now aware that these are limit-cases, or the way a more fundamental physical structure appears to us in certain contexts. Similarly, the conflict between GR and QM points to some deeper essence.

This effectively extends Sellars’ account of ‘looks’ talk from immediate perceptual contexts to richer theoretically (and mathematically) mediated ones. It’s not merely a statement that once seemed to be a viable option for assertion, but a whole way of conceptualising the world.

377. A statement like ‘It looks like steel deforms continuously, but it’s actually a granular crystalline structure’ can do important work articulating the relationship between different pragmatic contexts and descriptive scales, from the manifest centre to the scientific outer edge.

378. And a statement such as ‘It once looked like space was Euclidean and absolute, but it turned out to be relative and non-Euclidean’ helps us to articulate the way in which the revisionary trajectory of a concept tracks an essence it endeavours to capture.

379. While a statement like ‘It looks like GR correctly describes gravity and like QM correctly describes the other forces, but they can’t both be true’ helps us articulate the distance between concept and essence, and so divine the paths it might take going forward.

380. To be clear, I’m suggesting that there’s a degree of paraconsistency involved in managing the pragmatic and theoretical inconsistencies that inevitably emerge in the course of fitting our concepts to the world. This problematises both possible worlds epistemic scenarios.

381. The need for maximum specificity and global consistency idealises in a way that not only abstracts from, but falsifies the semantic link between epistemic and metaphysical modalities manifest in the pragmatics of inquiry, while rendering phenomenal concepts conveniently static.

Pete Wolfendale@deontologistics· 7d

Time to repost my best logic joke, about the BHK interpretation of intuitionistic logic.

Brouwer, Heyting, and Kolmogorov try to leave a bar.
They've been in there all afternoon, drinking and talking animatedly about the intricacies of constructive mathematics, but in the heat of the discussion they've completely lost track of their tab.

When the time comes to leave Kolmogorov calls for the bartender.
'Barkeep, can we pay our bill, then leave?'
The bartender walks over to their table and replies, 'Well… I don't know, can you?'
Suddenly worried, they all start emptying their wallets onto the table.

The barman looks at the money, and then looks back to Kolmogorov.
'Seems like you've got a problem mate. That's not going to cover it.'
'Could we come back with the rest tomorrow?'
'No can do. Either you pay or you don't, and if you don't, you don't leave.'
'But we can't pay!'
'Then you can't leave.'

Suddenly, Brouwer has an idea.
'Would you take a cheque?'
'How do I know it won't bounce?'
'How could a mathematician as renowned as myself *not* be good for it?'
'That's no real proof though, is it?'

Kolmogorov thinks for a moment, then takes another tack.
'If we didn't pay you, then we must have found some way to pay no one.'
'I suppose so'
'But there's no way to pay something to no one, so we must'nt've not paid you in the first place!'
'That's not really a solution though, is it?'

After a long, deep silence, Heyting turns to the barman.
'Can we keep drinking? I mean, you never know, we might be able to pay later.'
The barman thinks for a moment, then answers, a bit confused, 'I don't see why not?'
'Then we'll have another round please. On me.'

They never leave the bar.

Pete Wolfendale@deontologistics· Jun 22

Pete Wolfendale@deontologistics· Jun 16

Irritated that I forgot to replace 'Truth' with 'Logic' in the final diagram.

Pete Wolfendale@deontologistics· Jun 14

For anyone curious about what's been going on in my course on The Revenge of Reason for The New Centre, here are some of the diagrams from the last session, suitably cleaned up, pointing in the more systematic direction my work is developing in. Aa sketch of my reworking of Platonism.

Pete Wolfendale@deontologistics· Jun 7

I did a little 40K posting recently, explaining my take on the underlying metaphysics/symbology and how this articulates the relationships between the different factions. People seemed to like it, so I decided to draw up the diagram. There's a basic version, and a more tenuous one with subfactions.

A diagram of the relationships between the main factions in Warhammer 40K, articulated by the oppositions between the Natural, the Artificial, and the Unnatural, and the derivative oppositions between Life and Death, Culture and Savagery, and Order and Chaos.

Pete Wolfendale@deontologistics· Jun 5

I know many lovely people from London, and there are many lovely things about visiting the place, but it is also the source of a deep spiritual sickness immiserating the rest of the country, and indeed, much of the rest of the world:

Dan Hind: Postwar British history looks a lot like the administrators of a vast global and maritime empire coming home and casually destroying one element after another of that empire's infrastructure - in the mistaken belief that they understood how to run a medium sized national economy.

Pete Wolfendale: It’s not entirely unlike the regression of Rome from empire to papal city state, losing political authority but retaining centrality in a vast network of influence. Except London did it much faster, the religion is trans-national finance capital, and the indulgences tax evasion.

Pete Wolfendale@deontologistics· Jun 2

Nearly into the 1000s, and all it took was some 40K posting.

Pete Wolfendale@deontologistics· May 19

I very nearly wrote a review of Morton's Humankind when it came out, because it was, and remains, the worst book I have ever read. I read it twice to make sure. I just found this in my notes.

Pete Wolfendale@deontologistics· May 19

how it started: how it’s going:

The cover of Timothy Morton's book, Hyperobjects.

Post from Timothy Morton's blog, titled 'In Which I Begin to Come Out as Precognitive'

Pete Wolfendale@deontologistics· May 19

I also think this means that the one constituency to whom these capacities look universally impressive are 'general purpose management' types:

Giving managers access to LLMs is like giving each of them their own personal management consultancy (derogatory).

Pete Wolfendale@deontologistics· May 19

This isn’t a new observation, but I think there really is a massive referential calibration problem in public discourse about ‘AI’ and its capabilities. There are at least two sides to this.

Despite 3 years of progress, AI is still worse than me at stuff I understand, but unfathomably brilliant outside my area of expertise. Wonder why this is

Pete Wolfendale@deontologistics· Apr 29

The rhetorical weakness of this distinction is compensated for by mixing metaphors, or rather, analogies, in an extremely spurious and misleading manner. This passage is particularly gratuitous:

An analogy helps clarify the point. Connecting a computer to cameras and robotic arms is similar to attaching measuring instruments to a simulation. The simulation now receives real-world data, but the internal variables of the model are still symbolic representations rather than the physical processes themselves. In the same way, a weather model connected to live atmospheric sensors does not become the atmosphere. It simply receives and manipulates data about it.
The same principle applies to embodied Al systems. Sensors and actuators allow the system to interact with the physical world, but they do not automatically transform symbolic representations into intrinsic, experienced semantics. The system may build increasingly detailed maps of its environment, but interacting with the territory does not by itself turn the map into the territory of experience.

Pete Wolfendale@deontologistics· Apr 13

My reaction to this is some variant of the old Thought Slime line that 'most leftists are annoyed by leftists every single day'. This man is simply unable to be irritated by the discursive norms of Continental philosophy in the way that I am. To me, he is like a little baby.

Pete Wolfendale@deontologistics· Mar 30

Out in Ueno with these lovely lads tonight.

Two guys at an izakaya holding copies of The Revenge of Reason.

Pete Wolfendale@deontologistics· Mar 10

It’s cheating, but…

Pete Wolfendale@deontologistics· Feb 23

A reminder for this coming Wednesday, for anyone near Newcastle:

The Revenge of Reason Book Launch - Peter Wolfendale in conversation with Jacob Parkin, 25th of February 2026, 17:00-19:30, Henry Daysh Building, Floor 1 Room 1.02, Newcastle University

Pete Wolfendale@deontologistics· Feb 22

He distinguishes intelligence from consciousness but spends all his time talking about conditions for consciousness (principally qualia/valence) when the ultimate import is personhood. The closest we get to an analysis of the connection between these comes at the very end:

Pete Wolfendale@deontologistics· Feb 8

You're making Angron sad.

Pete Wolfendale@deontologistics· Feb 5

Pete Wolfendale@deontologistics· Dec 29

Pete Wolfendale@deontologistics· Dec 22

The question is which domain of functions we're picking an arbitrary instance from, by sampling its inputs/outputs. The bigger the domain the weaker the result. We might even use Van Rooij's own diagram to illustrate the options here:

Pete Wolfendale@deontologistics· Dec 21

It actually makes almost exactly the same argument I do, though with a great deal more precision. This is the key point:

The proof would work if the reduction were to a version of AI-by-Learning where the distribution
𝒟
is arbitrary rather than being a distribution of situation-human behavior tuples. However, if that were the case, the intractability of AI-by-Learning would only imply that there are functions that are not learnable. That fact is well-known [Val84].

Pete Wolfendale@deontologistics· Nov 27

First time I’ve put them side by side.

The Revenge of Reason and Object-Oriented Philosophy: The Noumenon’s New Clothes

Pete Wolfendale@deontologistics· Nov 20

It's a crime that Tim Kreider's bush era comics aren't available online anymore.

Pete Wolfendale@deontologistics· Nov 19

There’s a positive feedback loop involved here, as the aging demographic retains political power longer than they otherwise would, and they’re overwhelmingly anti-immigration. See this new poll from the UK about under-50s breaking Green, while over 50s break Reform.

Pete Wolfendale@deontologistics· Nov 3

Obligatory pinned post: my new book, The Revenge of Reason, is now available to buy: www.urbanomic.com/book/the-rev...

Me holding a copy of The Revenge of Reason.

Pete Wolfendale@deontologistics· Oct 27

It exists!

Several copies of The Revenge of Reason stacked in a box.

Pete Wolfendale@deontologistics· Sep 26