Skip to main content


epimetheus and procrustes

'I think in a line — but there is the potential of the plane.' This perhaps was what great art was: a momentary apprehension of the plane at a point in the line. — Charles Williams

A theory is a machine for answering a class of questions. — Judea Pearl What philosophers and socialthinkers and literary critics call their "theories" do not often merit the word. Rather, these people lay out models - with all the artificiality, partiality and temporariness that implies, and seldom the precision. A model is a tool for making sense of things, for filtering out things in self-defence, and is not necessarily a general claim about reality, a truth-machine.

(Some of them agree with this low estimate, like the post-modernists. People don't realise that post-modernists are often self-deprecating - some of them even call their works little: Minima Moralia, Minima Memoria, les petits récits.)

These theories are actually language games. O.K. Except that (the best) scien…
Recent posts


I can handle bullshit. However copious, however arbitrary: I am confident in the extent of my stoicism, I could be a Kolmogorov, it would flow over me without touching the heart. I'm capable of not caring about what I do.

Logically, I would be worse off if I couldn't stand it, if I was forced to be Solzhenitsyn. But I can't help admiring people who can't put up with it. I have an extra ability, and many extra opportunities, but they've the crown.

notable heinleinations

winchell (n.) and lippman (n.): Two words for 'journalist' coined by Heinlein. A "winchell" is a hateful, gossipy pundit. Named for this nasty little man. A "lippman" is a brave, muckracking investigator. Named for this good guy. We really really need two words for these two wildly different ideal types. (More recently Nick Davies tried out "churnalism", but it didn't stick.) It is stupid to attack 'journalism' when you are only dissing winchells, but that is the blunt instrument of language for you.
niladic (computer adj.): Of a function, having zero arguments. The zeroth member of the n-adics.
d'Alembertian (maths n.): the Laplace operator in Minkowski space; the divergence of the gradient of a function in a 4D manifold. What a beautiful word.
bim (1950s US n.): woman. I think the register is the same as "chick" or "broad".
offboarding process (HR adj.): The minutiae surrounding leaving a job: deleting accounts, g…

you do you

Out at London Pride, uncharacteristically. Around 1am, as you do, find myself in an intimate discussion with a stranger. Who leaps to what he imagines is my defence:
"What are you into?""Women, mostly.""Oh right." "It's a really limited view, though. I wish I were more open.""What? There's no such thing as limited! You got to accept yourself!""Sure, but - I'm only attracted to like a tenth of the population; it would be objectively better to have a larger pool, and to see all of beauty.""Hey, hey: stop it, you don't need to justify myself. You are you, so you do you. Don't do anyone else.""That's a beautiful thing to say, and there's a lot of people in the world who really need to hear it. Not me though.""It's true for everyone! You'll only hurt yourself by not being yourself.""No, though. Consider: monkeys like bananas and sex; humans like bananas, sex …

odour of sanctity

"36056e1bc7447cc3d66832a10f9e420beeab3226" (2017) by dorianna Besides being free and endless and potentially superhuman, algorithmic art has a hidden benefit:
The artefacts are not the real problem with conceptual art: rather, it's the pretense. The artist, feigning profundity; the critic, helping them wank. Artbots won't claim profundity, nor will they be attributed it; so they might make museums smell less of shite.

"Accidentally generated pattern........" (2015) by Kieran G. Larkin 'Exploration of Related Points Along a Space Filling Curve #2' (2007) by Don Relyea "inception_4c-output-3-9-3" (2007) by Pete Ashton

notable omens and totems

POPS (n.): privately owned public space. A place where secret rules can apply to you. You can tell them by how nice/sterile they are and by the hi-viz men lurking at the edges. ("Public" is obvs a very vague word but it is fair to gloss it here as "unfenced place with seating or grass, on a public thoroughfare, which people gravitate to in idle hours".)
to roget (v.): to disguise plagiarism through systematic abuse of a thesaurus. Good word game: what rogeted phrase produces "sinister buttocks"? (Actually of course, "to attempt to disguise".) The sad thing is is that this is happening in an age with good automatic paraphrasing software.
yellow books (maths n.): The mathematics textbooks put out by Springer Verlag; a watchword for rigour and broken ground and what you will never ever master.
nooch (n.): Nutritional yeast. A forced and chummy coinage by well-meaning vegans.
senectitude (n.): old age, senescence, dotage.
bobby-soxer (1940s n.): Young f…

Been reading, Q2 2017

‘Fall’ (2012) by Victoria Reichelt. I can only really evaluate the truth of a book if it is on a topic I already know about. But books on topics I know about already are not nearly as worth reading as ones on new topics.

So, to maximise nonfiction, I need to know if a book is accurate before I read it. (Or, better, I need to know where exactly it is inaccurate.) This could be got using extremely detailed reviews by extremely involved experts - but they themselves would be wasting their time reading things they already know about. And reviews are generally too brief for this, even in academic journals or dedicated blogs; they gesture at one or two mistakes as a way of casting inductive aspersions on the whole. So what I really need is a wiki for errors and dishonesties, to spread out the Augean doings of journalists and average authors.

I don't read much fiction because I prefer reading people who know things. (That looks like a dreadful slander on novelists, but consider: …