The Mane Six as Mitford Sisters


[Written as part of Notebook Blog Month.]

I’ve saved the most important topic for last. As far as I can tell, nobody on the internet has tackled the vital question of how My Little Pony characters map to Mitford sisters. So I’m going to fix that.

As a bit of background, the Mitfords were a wildly eccentric English aristocratic family. The novelist Nancy Mitford is probably the most famous of them, but her five sisters were an impressively bizarre mix of communists, fascists, socialites, farmers and Hitler obsessives. (There’s also a brother who nobody cares about.) I’m not particularly well up on Mitford lore, but I am a big fan of Nancy Mitford’s The Pursuit of Love, and I’m fascinated by eccentrics of all kinds, so I know the basics.

My Little Pony also has six main characters. (Plus Spike.) They turn out to match up surprisingly closely with Mitford sisters, right up to the point where they don’t, and then I just have to make it up.

So, first of all, Nancy Mitford is Twilight Sparkle. This one is completely obvious. In The Pursuit of Love, Nancy is the narrator of a fictionalised version of the Mitfords’ lives, with her as the quieter, more studious observer. Done.

Deborah Mitford was a famous socialite known for… being social and stuff. Lots of witty correspondence with other witty socialite types. She also ran a big stately home that was open to the public, which is kind of like organising parties if you squint hard enough. So Pinkie Pie.

Jessica Mitford was the most adventurous and rebellious of the sisters, running away to Spain and then later becoming an activist in the US, where she worked on civil rights campaigns and joined the Communist Party. Also investigated unscrupulous business practices in the funeral home industry for some reason. Clearly has to be Rainbow Dash.

Unity Valkyrie Mitford is the oddest of the lot. She became completely obsessed with Hitler, stalked him around Munich, eventually made her way into his inner circle, shot herself in the head when Britain declared war on Germany, survived the attempt and lasted out almost another ten years before dying of meningitis caused by swelling around the bullet, which was never removed. I feel like this is a job for Rarity, partly because of the rhyming name, partly because she’s the only one capable of pulling off this much drama.

OK, this is the point where the mapping gets a bit trickier. The two remaining sisters are Pamela and Diana. Pamela was the most retiring of the Mitfords, staying out of the public eye, at least in comparison to the others. She was practical-minded, loved animals and the countryside and managed a farm for a while. A pretty good fit for either Applejack or Fluttershy. She did still manage to do some weird Mitford stuff, marrying a bisexual millionaire physicist and then becoming the ‘companion’ of an Italian horsewoman after they divorced.

Diana is the dud Mitford. She was another fascist, and not even a spectacularly bizarre one like Unity. Mainly known for marrying Oswald Mosley, leader of the British Union of Fascists, editing a fascist magazine, and spending some time in prison during the war for being a fascist.

I don’t really want to lumber either of them with Diana, but I need to make a choice, so I’ve introduced an outside tie break. Fluttershy is best pony. So Fluttershy gets Pamela Mitford and poor old Applejack is stuck with Diana Mitford.

I’m sure everyone’s relieved that this major open question has been definitively answered at last.

Bullshitting about bullshit jobs


[Written as part of Notebook Blog Month.]

Today’s topic is bullshit jobs. I’ve done no preparation for this beyond a reread of David Graeber’s essay, On the Phenomenon of Bullshit Jobs: A Work Rant. This piece was enormously popular when it came out in 2013, so presumably there’s a large secondary literature of commentaries and follow-ups. I haven’t read any of it, this is me just bullshitting from first principles. So I’m probably repeating a lot of obvious talking points.

Rereading the article, a couple of things jumped out straight away:

  • A lot of the specifics of his argument are not particularly convincing (I’ll get to this in a minute).
  • It rings deeply true anyway, because we all know deep in our hearts and viscera that so many jobs are full of useless bullshit. It’s not surprising that it was so popular.

So, I’ll briefly go over his thesis. Advances in productivity and automation should have freed up lots of time by now, and we should have got the 15 hour working week that John Maynard Keynes expected. But clearly we haven’t. Graeber rules out the consumer treadmill as an explanation:

The standard line today is that he didn’t figure in the massive increase in consumerism. Given the choice between less hours and more toys and pleasures, we’ve collectively chosen the latter. This presents a nice morality tale, but even a moment’s reflection shows it can’t really be true. Yes, we have witnessed the creation of an endless variety of new jobs and industries since the ’20s, but very few have anything to do with the production and distribution of sushi, iPhones, or fancy sneakers.

Instead, he points the finger at what he describes as whole new classes of jobs:

… rather than allowing a massive reduction of working hours to free the world’s population to pursue their own projects, pleasures, visions, and ideas, we have seen the ballooning of not even so much of the ‘service’ sector as of the administrative sector, up to and including the creation of whole new industries like financial services or telemarketing, or the unprecedented expansion of sectors like corporate law, academic and health administration, human resources, and public relations. And these numbers do not even reflect on all those people whose job is to provide administrative, technical, or security support for these industries, or for that matter the whole host of ancillary industries (dog-washers, all-night pizza delivery) that only exist because everyone else is spending so much of their time working in all the other ones.

These are what I propose to call ‘bullshit jobs’.

It’s as if someone were out there making up pointless jobs just for the sake of keeping us all working.

So, in this view there are a bunch of jobs that are bullshit. Here’s another sample list further down:

A world without teachers or dock-workers would soon be in trouble, and even one without science fiction writers or ska musicians would clearly be a lesser place. It’s not entirely clear how humanity would suffer were all private equity CEOs, lobbyists, PR researchers, actuaries, telemarketers, bailiffs or legal consultants to similarly vanish.

It’s not very satisfying to me to leave this at the level of a big binary list of jobs that are bullshit (telemarketers, corporate law) and jobs that are not (teachers, tube drivers). To my mind the bullshit is much more fractally distributed throughout the whole economy. It’s definitely true that some jobs are much more prone to gathering bullshit than others. But most of the ‘bullshit’ jobs Graeber lists serve some useful functions. I’m not sure what he’s got against actuaries – insurance seems like a reasonable thing to me, and somebody needs to work out how much it should cost. And some level of financial and legal work needs to go on. (Some of these I do find hard to defend at all. I think telemarketing might actually be pure bullshit given a narrow enough definition? Does anyone need to bother other people by phone any more? I mean I really hate phones so I could be biased here, but that does sound like bullshit to me.)

At the other end of the scale, a lot of his ‘non-bullshit’ jobs get mixed up with bullshit too. Teachers are always having to grapple with the latest bullshit government initiative, for example. The bullshit is mixed right through everything.

I want to probe a bit deeper into what factors are upstream of jobs becoming bullshit. There definitely seem to be warning signs for bullshit. For a start, jobs are particularly likely to contain a lot of bullshit if they contain a lot of abstraction layers. For example:

  • selling abstract things (financial derivatives) rather than concrete things (potatoes)
  • managing people who do things, rather than doing things directly
  • contracting out work to a second company, rather than doing it yourself.
  • producing hard-to-measure output (potato marketing board) rather than obvious results (potato farmer)

These aren’t bad things intrinsically, they need to happen to some extent or we’d all be stuck individually bartering potatoes all day. But they provide places for the bullshit to get in.

For the rest of this post I’m going to play with one potential taxonomy of bullshit jobs. This isn’t supposed to be a Grand Unified Theory of Bullshit Jobs, it’s just me playing around. I don’t have time to try multiple taxonomies in a notebook post like this, so if this one turns out to not be very insightful then I just have to deal with that I suppose. Anyway, it’s inspired by Frankfurt’s characterisation of bullshit in his classic On Bullshit:

It is impossible for someone to lie unless he thinks he knows the truth. Producing bullshit requires no such conviction. A person who lies is thereby responding to the truth, and he is to that extent respectful of it. When an honest man speaks, he says only what he believes to be true; and for the liar, it is correspondingly indispensable that he considers his statements to be false. For the bullshitter, however, all these bets are off: he is neither on the side of the true nor on the side of the false. His eye is not on the facts at all, as the eyes of the honest man and of the liar are, except insofar as they may be pertinent to his interest in getting away with what he says. He does not care whether the things he says describe reality correctly. He just picks them out, or makes them up, to suit his purpose.

The main feature of bullshit, as Frankfurt explains it, is that it is indifferent to the truth, rather than outright false. It’s produced as a side effect of some other self-serving process.

This already ties in quite well with the idea of bullshit seeping through abstraction layers. Abstraction layers tend to be places where caring breaks down – it’s just easier to care about potatoes than financial derivatives. In my taxonomy I’m going to explore three kinds of not caring:

  1. Nobody cares about the problem
  2. Nobody cares whether the solution could possibly fix the problem
  3. The incentives push against caring anyway

I’ll go through the eight options. First there’s the Platonic ideal of a non-bullshit job:

  • Care about problem, care about solution, good incentives. Honest artisan hand crafts a beautiful table to sell directly etc etc.

Then we get into the region of somewhat bullshit jobs:

  • Care about problem, care about solution, bad incentives. This is where teaching often ends up, for example (good teachers, I mean). Loves their subject, wants to teach it well, but also has to tick the boxes for ‘learning outcomes’ or w/e.

  • Care about problem, don’t care about solution, good incentives. Caring about the problem but not whether the solution could possibly fix it is a funny one, but something like it happens quite often once a bit of self-deception gets involved, and some bad incentives. Doing it with good incentives is harder. Maybe a widget factory boss gets infatuated with some kind of trendy-but-useless management methodology. They genuinely want to sell more widgets, the market for widgets functions well, but there’s a layer of cargo-cult stupidity in the middle. That’s the best I can do, maybe somebody else can come up with a better example.

  • Don’t care about problem, care about solution, good incentives. This is somebody putting in a solid day of work at a a job they don’t particular care for intrinsically, but with decent working conditions and high standards for what counts as a good job done.

None of these seem like the canonical bullshit job to me, but they are definitely likely to contain bullshit elements. Then we get towards the real bullshit:

  • Care about the problem, don’t care about the solution, bad incentives. This is the self-delusion thing again. Maybe this is a charity employee who genuinely cares about the cause but has some motivated reasoning going on about whether the thing they’re doing could possibly help. If the charity is able to fundraise whether the work is useful or not, you get this situation.

  • Don’t care about the problem, care about the solution, bad incentives. This is the typical academic with a very specific pet hammer, churning out papers that use it in dubious ways.

  • Don’t care about the problem, don’t care about the solution, good incentives. This is just working a job you don’t care about again, but this time without being held back by any kind of standards of professionalism or craftsmanship. Working conditions are still good.

Then finally we get to:

  • Don’t care about the problem, don’t care about the solution, incentives are terrible anyway. You can get here quite easily by switching the remaining yes to a no in the examples above: the academic doesn’t even care about whether the technique is carried out right, the charity employee is completely indifferent to the cause rather than well-meaning but self-deluded, the apathetic worker also has a horrible boss and isn’t paid well. Then you just have pure bullshit.

I’ll try it on one real example, that job I did where I walked around hospitals measuring things:

There were two hospitals being merged together on a new site, and the project management office needed to collect data on how much storage space the new hospital would need for medical supplies. I’m not sure what the best way of doing this would be, but maybe it would involve, I don’t know, some Fermi estimates based on their current storage requirements, plus some efficiencies for the single site. What they actually did was make a giant spreadsheet of every sort of item ordered by the hospital (bandages, prosthetics, tiny orthopaedic screws) and then employ EIGHT OF US to go round the hospitals with tape measures FOR WEEKS tracking down and measuring every individual item on the list, including the tiny orthopaedic screws.

If that doesn’t score highly on the bullshitometer then something is really up.

So… the original problem, ‘how much storage space do we need’, is a good one, and presumably somebody somewhere really cared about the answer. Once this had been filtered through a couple of management layers and subcontractors most of the caring had been lost. I’ll give it a generous half a point.

Definitely nobody cared whether the solution could possibly fix the problem. Zero points.

Incentives were also awful. The work had been contracted out through a couple of layers, and the people who had to find the answer had no particular reason to care beyond coming up with a number that kept the layer above happy. Zero points again.

So I make that 0.5/3. Definitely a bullshit job. So it passes this basic sanity check. (Not a great one, as I’d have had exactly this example at the back of my mind when I came up with the criteria!)

So… was that particulary taxonomy any good? Not particularly, but it did get my thinking through the space of possibilities. I do think that the general strategy of looking for places where not-caring gets in is a good one for spotting bullshit.


[Written as part of Notebook Blog Month.]

Everybody hates neoliberalism, it’s the law. But what is it?

This is probably the topic I’m most ignorant about and ill-prepared-for on the whole list, and I wasn’t going to do it. But it’s good prep for the bullshit jobs post, which was a popular choice, so I’m going to try. I’m going to be trying to articulate my current thoughts, rather than attempting to say anything original. And also I’m not really talking about neoliberalism as a coherent ideology or movement. (I think I’d have to do another speedrun just to have a chance of saying something sensible.) More like “neoliberalism”, scarequoted, as a sort of diffuse cloud of associations that the term brings to mind. Here’s my cloud (very UK-centric):

  • Big amorphous companies with bland generic names like Serco or Interserve, providing an incoherent mix of services to the public sector, with no obvious specialism beyond winning government contracts
  • Public private partnerships
  • Metrics! Lots of metrics!
  • Incuriosity about specifics. E.g. management by pushing to make a number go up, rather than any deep engagement with the particulars of the specific problem
  • Food got really good over this period. I think this actually might be relevant and not just something that happened at the same time
  • Low cost short-haul airlines becoming a big thing (in Europe anyway – don’t really understand how widespread this is)
  • Thinking you’re on a public right of way but actually it’s a private street owned by some shopping centre or w/e. With private security and lots of CCTV
  • Post-industrial harbourside developments with old warehouses converted into a Giraffe and a Slug and Lettuce
  • A caricatured version of Tony Blair’s disembodied head is floating over the top of this whole scene like a barrage balloon. I don’t think this is important but I thought you’d like to know

I’ve had this topic vaguely in mind since I read a blog post by Timothy Burke, a professor of modern history, a while back. The post itself has a standard offhand ‘boo neoliberalism’ side remark, but then when challenged in the comments he backs it up with an excellent, insightful sketch of what he means. (Maybe this post should just have been a copy of this comment, instead of my ramblings.)

I’m sensitive to the complaint that “neoliberalism” is a buzz word that can mean almost everything (usually something the speaker disapproves of).

A full fleshing out is more than I can provide, though. But here’s some sketches of what I have in mind:

1) The Reagan-Thatcher assault on “government” and aligned conceptions of “the public”–these were not merely attempts to produce new efficiencies in government, but a broad, sustained philosophical rejection of the idea that government can be a major way to align values and outcomes, to tackle social problems, to restrain or dampen the power of the market to damage existing communities. “The public” is not the same, but it was an additional target: the notion that citizens have shared or collective responsibilities, that there are resources and domains which should not be owned privately but instead open to and shared by all, etc. That’s led to a conception of citizenship or social identity that is entirely individualized, privatized, self-centered, self-affirming, and which accepts no responsibility to shared truths, facts, or mechanisms of dispute and deliberation.

2) The idea of comprehensively measuring, assessing, quantifying performance in numerous domains; insisting that values which cannot be measured or quantified are of no worth or usefulness; and constantly demanding incremental improvements from all individuals and organizations within these created metrics. This really began to take off in the 1990s and is now widespread through numerous private and public institutions.

3) The simultaneous stripping bare of ordinary people to numerous systems of surveillance, measurement, disclosure, monitoring, maintenance (by both the state and private entities) while building more and more barriers to transparency protecting the powerful and their most important private and public activities. I think especially notable since the late 1990s and the rise of digital culture. A loss of workplace and civil protections for most people (especially through de-unionization) at the same time that the powerful have become increasingly untouchable and unaccountable for a variety of reasons.

4) Nearly unrestrained global mobility for capital coupled with strong restrictions on labor (both in terms of mobility and in terms of protection). Dramatically increased income inequality. Massive “shadow economies” involving illegal or unsanctioned but nevertheless highly structured movements of money, people, and commodities. Really became visible by the early 1990s.

A lot of the features in my association cloud match pretty well: metrics, surveillance, privatisation. Didn’t really pick up much from point 4. I think 2 is the one which interests me most. My read on the metric stuff is that there’s a genuinely useful tool here that really does work within its domain of application but is disastrous when applied widely to everything. The tool goes something like:

  • let go of a need for top-down control
  • fragment the system into lots of little bits, connected over an interface of numbers (money, performance metrics, whatever)
  • try to improve the system by hammering on the little bits in ways such that the numbers go in the direction you want. This could be through market forces, or through metrics-driven performance improvements.

If your problem is amenable to this kind of breakdown, I think it actually works pretty well. This is why I think ‘food got good’ is actually relevant and not a coincidence. It fits this playbook quite nicely:

  • It’s a known problem. People have been selling food for a long time and have some well-tested ideas about how to cook, prep, order supplies, etc. Theres’s innovation on top of that, but it’s not some esoteric new research field.
  • Each individual purchase (of a meal, cake, w/e) is small and low-value. So the domain is naturally fragmented into lots of tiny bits.
  • This also means that lots of people can afford to be customers, increasing the number of tiny bits
  • Fast feedback. People know whether they like a croissant after minutes, not years.
  • Relevant feedback. People just tell you whether they like your croissants, which is the thing you care about. You don’t need to go search for some convoluted proxy measure of whether they like your croissants.
  • Lowish barriers to entry. Not especially capital-intensive to start a cafe or market stall compared with most businesses.
  • Lowish regulations. There’s rules for food safety, but it’s not like building planes or someting.
  • No lock-in for customers. You can go to the donburi stall today and the pie and mash stall tomorrow.
  • All of this means that the interface layer of numbers can be an actual market, rather than some faked-up internal market of metrics to optimise. And it’s a pretty open market that most people can access in some form. People don’t go out and buy trains, but they do go out and buy sandwiches.

There’s another very important, less wonky factor that breaks you out of the dry break-it-into-numbers method I listed above. You ‘get to cheat’ by bringing in emotional energy that ‘comes along for free’. People actually like food! They start cafes because they want to, even when it’s a terrible business idea. They already intrinsically give a shit about the problem, and markets are a thin interface layer over the top rather than most of the thing. This isn’t going to carry over to, say, airport security or detergent manufacturing.

As you get further away from an idealised row of spherical burger vans things get more complicated and ambiguous. Low cost airlines are a good example. These actually did a good job of fragmenting the domain into lots of bits that were lumped together by the older incumbents. And it’s worked pretty well, by bringing down prices to the point where far more people can afford to travel. (Of course there’s also the climate change considerations. If you ignore those it seems like a very obvious Good Thing, once you include them it’s somewhat murkier I suppose.)

The price you pay is that the experience gets subtly degraded at many points by the optimisation, and in aggregate these tend to produce a very unsubtle crappiness. For a start there’s the simple overhead of buying the fragmented bits separately. You have to click through many screens of a clunky web application and decide individually about whether you want food, whether you want to choose your own seat, whether you want priority queuing, etc. All the things you’d just have got as default on the old, expensive package deal. You also have to say no to the annoying ads trying to upsell you on various deals on hotels, car rentals and travel insurance.

Then there are the all the ways the flight itself becomes crappier. It’s at a crap airport a long way from the city you want to get to, with crappy transport links. The flight is a cheap slot at some crappy time of the early morning. The plane is old and crappily fitted out. You’re having a crappy time lugging around the absolute maximum amount of hand luggage possible to avoid the extra hold luggage fee. (You’ve got pretty good at optimising numbers yourself.)

This is often still worth it, but can easily tip into just being plain Too Crappy. I’ve definitely over-optimised flight booking for cheapness and regretted it (normally when my alarm goes off at three in the morning).

Low cost airlines seem basically like a good idea, on balance. But then there are the true disasters, the domains that have none of the natural features that the neoliberal playbook works on. A good example is early-stage, exploratory academic research. I’ve spent too long on this post already. You can fill in the depressing details yourself.

Some rambling thoughts about visual imagery


[Written as part of Notebook Blog Month.]

I’ve got some half-written drafts for topics on the original list which I want to finish soon, but for now I seem to be doing better by going off-list and rambling about whatever’s in my head. Today it’s visual imagery.

I’ve ended up reading a bunch of things vaguely connected with mnemonics in the last couple of weeks. I’m currently very bad at concentrating on books properly, but I’m still reading at a similar rate, so everything is in this weird quarter-read state. Anyway here’s the list of things I’ve started:

  • Moonwalking with Einstein by Joshua Foer. Pop book about learning to compete in memory championships. This is good and an easy read, so there is some chance I’ll actually finish it.
  • Orality and Literacy by Walter Ong. One of the references I followed up. About oral cultures in general but there is stuff on memorisation (e.g. repetitive passages in Homer being designed for easy memorisation when writing it down is not an option)
  • Brienne Yudkowsky’s posts on mnemonics
  • These two interesting posts by AllAmericanBreakfast on Less Wrong this week about experimenting with memory palaces to learn information for a chemistry exam.

Those last two posts are interesting to me because they’re written by someone in the very early stages of fiddling around with this stuff who doesn’t consider themself to naturally have a good visual imagination. I’d put myself in the same category, but probably worse. Actually I’m really confused about what ‘visual imagery’ even is. I have some sort of – stuff? – that has a sort of visual component, maybe mixed in with some spatial/proprioceptive/tactile stuff. Is that what people mean by ‘visual imagery’? I guess so? It’s very transitory and hard to pin down in my case, though, and I don’t feel like I make a lot of use out of it. The idea of using these crappy materials to make something elaborate like a memory palace sounds like a lot of work. But maybe it would work better if I spent more time on it.

The thing that jumped out of the first post for me was this bit:

I close my eyes and allow myself to picture nothing, or whatever random nonsense comes to mind. No attempt to control.

Then I invite the concept of a room into mind. I don’t picture it clearly. There’s a vague sense, though, of imagining a space of some kind. I can vaguely see fleeting shadowy walls. I don’t need to get everything crystal clear, though.

This sounded a lot more fun and approachable to me than crafting a specific memory palace to memorise specific things. I didn’t even get to the point of ‘inviting the concept of a room in’, just allowed any old stuff to come up, and that worked ok for me. I’m not sure how much of this ‘imagery’ was particularly visual, but I did find lots of detailed things floating into my head. It seems to work better if I keep a light touch and only allow some very gentle curiosity-based steering of the scene.

Here’s the one I found really surprising and cool. I was imagining an intricately carved little jade tortoise for some reason, and put some mild curiosity into what its eyes were made of. And I discovered that they were tiny yellow plastic fake gemstones that were weirdly familiar. So I asked where I recognised them from (this was quite heavy-handed questioning that dragged me out of the imagery). And it turns out that they were from a broken fish brooch I had as a kid. I prised all the fake stones off with a knife at some point to use for some project I don’t remember.

I haven’t thought about that brooch in, what, 20 years? But I remember an impressive amount of detail about it! I’ve tried to draw it above. Some details like the fins are a best guess, but the blue, green and yellow stones in diagonal stripes are definitely right. It’s interesting that this memory is still sitting there and can be brought up by the right prompt.

I think I’ll play with this exercise a bit more and see what other rubbish I can dredge up.

Doing things on purpose


[Written as part of Notebook Blog Month.]

Veering off the list again today. This is a quick post and I’m not trying to make a very coherent point, just chuck out some thoughts. I read two things this morning that were about, more or less, ‘doing things’, and now I want to talk about that.

The first was Venkatesh Rao’s latest post on his Art of Gig newsletter, talking about bad questions newbies waste time on to put off actually having to act:

Here are some examples of bad questions.

LLC or S-corp? (or equivalent question in other countries). The right answer is “probably LLC,” but if you don’t trust me, sure, go with S-corp. It’s not too costly to fix this if you get this wrong.

Blogging to attract inbound leads, or proactive email pitches? The obvious answer is the right one: try both, see what works, double down. Cheap effort.

Targeted, researched pitches versus spray-and-pray? Targeted, obviously. But sure, waste your time on spray-and-pray for a while. Maybe you’re one of the exceptions.

These all have similar ‘answers’: make a cursory stab at researching the problem to rule out obvious dumb ideas, then try one of the remaining options and see what happens. Then you’ll have more information:

They’re bad questions because you’re trying to fix an information deficit (which calls for trial and error) by over-analyzing information you do have.

The second was T. Greer at The Scholar’s Stage, talking about cultures that build:

The Americans of 1918 had carved towns, cities, and states out of the wilderness, and had practical experience building the school boards, sheriff departments, and the county, city, and state governments needed to manage them. Also within the realm of lived experience was the expansion of small towns into (unprecedentedly large) metropolises and the invention of the America’s first multi-national conglomerates. The progressive movement had spent the last three decades experimenting with new forms of government and administration at first the state and then the federal level, while American civic society saw a similar explosion in new social organizations. These include some famous names: the NRA, the American Red Cross, the Salvation Army, the American Bar Association, the Sierra Club, 4-H, the VFW, Big Brothers, the NAACP, the Boys Scouts, the PTA, the United Way, the American Legion, and the ACLU. [3] To a large extent we wander in the ruins of the world this generation built.

Greer argues that this kind of local civic action provides an education in doing things on purpose that then bubbles up to the larger scale:

I understand that the self-organizing neighborhood committee that removes a tree that blocks their street does not go on to build the Empire State Building. My argument is slightly different. To consistently create brilliant poets, you need a society awash in mediocre, even tawdry poetry. Brilliant minds will find their way towards poem writing when poem writing and poem reading is the thing that people do.

Now, I’m some nerd who likes to read a lot of text on the internet. I’m not actually very good at doing things. But I’ve slowly been pushing up my ability to just try things quickly, and that’s going well for me. I guess this notebook blog month thing is one example. I think I want some kind of extra writing discipline on the days-to-weeks level, and this is an easy experiment, so I might as well just try it and see what happens. I don’t think it quite fits what I want, but I’ve learned a whole lot more from doing it than I would have done by abstractly thinking about what I might want. Next month I can do something different.

I got a big one-time boost in my ability to do things on purpose from a single week-long physics workshop I went to a few years ago. It took me from ‘vaguely trying to keep learning physics in my spare time’ to ‘actually focussing on a topic and pushing towards doing serious work outside of academia’. Part of that was findng a community of people with similar interests. Part of it was that the workshop was built from the ground up to encourage agency, in a way that matches Greer’s start-from-everyday-decisions pattern. We were in a beautiful mountain cabin in the middle of nowhere in the Austrian Alps (if you were wondering what the picture is about this time, that’s my attempt at drawing the side of the cabin). So all cooking and cleaning had to be done by us. There was a rota for meals, but everything else was on a whoever-notices-it-fixes-it basis, with no distinction between the organisers and the rest of us. If the bin is overflowing or you need more firewood from the barn, sort it out yourself.

This low-level agency flowed up into higher-level decisions about the workshop. Everyone had to give a talk, and anyone could suggest discussion sessions for the afternoons based on questions that came up. Everybody took turns at moderating discussions and taking notes. If somebody had an idea for a good topic while putting the plates away, they could just gather some interested people and try it.

It also flowed out beyond the workshop, into the small society that runs it. I ended up joining this and it works the same way: anybody can suggest an event, get together some organisers and start running it. Or start some other new initiative, like a newsletter. People are already used to making decisions themselves from the workshop, so new members actually do this. We take turns to go on the board that ‘runs the society’ each year, but this is really just dealing with any admin that comes up and organising a general meeting once a year. Everything else seems to… just happen.

There are obvious downsides to this setup as well. It’s hard to scale; if you’re not fussy enough about entry you get taken over by cranks; if you’re too fussy about entry you turn into an old boys’ club that won’t listen to anybody with slightly different ideas. But it’s definitely good at solving the problem of getting people to do interesting stuff on purpose.

The Bristol Bridge Problem


[Written as part of Notebook Blog Month.]

Last year I walked the Bristol Bridge Walk with my brother. This is a 28 mile circuit in the style of Königsberg’s famous bridge problem, where every bridge in Bristol (or more accurately, every footbridge across the Avon) has to be crossed exactly once. It was designed by Thilo Gross, who figured out that Bristol’s bridge problem can be solved, unlike Königsberg’s. At least until they build more bridges and mess it up.

I wrote a Twitter thread on this a while back but thought I’d try an expanded version here. The interesting bit for me is how much of Gross’s work in creating this route was in defining the problem, mapping all the mess of modern Bristol onto a clean mathematical model.

The original Königsberg bridge problem has the following structure, with four land masses and seven bridges:


Euler worked out that it wasn’t possible to cross each bridge exactly once. The solution is simple and short enough that I may as well reproduce it here. He found a clever reframing of the problem, where the four land masses become nodes of a graph, and the seven bridges become edges between them:


Now, ignore the start and end of the walk for a minute. For the rest of the time, every time you enter a land mass by a bridge, you also leave by a bridge. So if every bridge has been crossed exactly once, then for any node of the graph apart from the start and end ones, there must be an even number of edges coming out of it.

Now this can’t possibly work for Königsberg, because all four nodes have three edges coming out of them, and you can only get away with having the two start and end nodes with odd numbers of edges. So the problem is insoluble.

Bristol has a similar overall structure to Königsberg, so the problem maps over quite nicely. There’s one main river, the Avon, with two islands in the middle of it: Spike Island and Redcliffe. Here’s a map:


So there are four nodes. But modern Bristol has a whole lot more bridges. 45 in fact. (Or maybe 47 now. The two new arena bridges were closed when I did the walk. They don’t change anything though, you just cross one and immediately cross back on the other.) Here’s the graph:


A is the north bank, B is the south, C is Spike Island and D is Redcliffe. There are so many bridges between Redcliffe and the two banks that I just wrote the numbers instead of drawing each edge. There are also several internal bridges linking the north bank to itself, for reasons I’ll get to.

(I came up with that graph myself, and it might not exactly match Gross’s. I tried to construct it to match his descriptions, but there might be some edge cases I counted slightly differently to him.)

Ignoring the internal edges linking A to A (which can be removed), all nodes have an even number of edges. A has 20 without the internal edges, B has 18, C has 12 and D has 26. So it’s a soluble problem.

Now, coming up with that graph was a good deal more complicated than just counting bridges. This is the bit I find interesting as an example of problem formation, mapping complicated stuff in the real world onto a clean mathematical model with an unambiguous technical solution. Gross explains all the work he had to do to figure out the scope of the problem, and what should actually count as a bridge:

It required about three solid days of work spread over a fortnight. Two thirds of the time was spent just finding out where all the bridges were, and what was to count as a bridge and what not. Only bridges that are walkable are included and a judgement had to be made about some smaller bridges (not comparable in size to Königsberg’s bridges) on tributaries of the Avon that would have led the resulting walk long distances through some dull areas. Some bridges comprise two separate structures of segregated traffic and are counted as two.

It turns out that Bristol is like Königsberg on the large scale, but when it comes to the details there are a whole bunch of judgement calls. These generally have a reasonable, common-sense resolution, so the resulting solution doesn’t feel arbitrary, but it does mean that solving the problem is not just a quick matter of drawing some nodes and edges. I’ll go through a load of these below.

First up, Gross needed to decide what counts as a bridge. This is purpose-dependent: rail- and car-only bridges don’t count in this solution because the whole point is to be able to walk it.

Next, it’s not exactly true that there are four nodes. If you look closely at the map, there are two little islets off of Spike Island:


The left islet is kind of complicated, and worth looking at on a satellite view:


The two bottom footbridges linking the islet to Spike Island are genuine public bridges over the Avon and should definitely be included. The two top ones are little lock gate thingies. I don’t think they’re reliably walkable, and they probably shouldn’t count. So it’s a reasonable choice to ignore them and count the islet as part of the north bank node for the purposes of this graph.

The right islet is simpler: one bridge on, one bridge off. It kind of doesn’t matter much what you do with it – it could be counted separately (in which case it’s a new node with two edges), or shoved in with either the north bank or Spike Island. Gross went for the second option, preserving the similarity with Königsberg. Then the south bridge gets counted as the bridge between the north bank and Spike Island, and the second one gets counted as an internal bridge linking the north bank to itself. (At least that’s what I did. Not sure exactly how Gross formalised it.)

Also, there are a couple of minor rivers emptying into the Avon. There’s the Trym:


In a sense this cuts the north bank into two: west of the Trym and east of the Trym. Should these be two nodes? And should you have to traverse all the bridges along the Trym as well, right out into the suburbs of Bristol?

Gross goes with no. It’s a small river and is just not a big deal: I don’t think that which side of the Trym you’re on is part of many people’s mental geography in Bristol, in the way that ‘north of the Avon’ (or ‘east of the motorway’) would be. (While walking this we crossed it without noticing, and had to go back to get a photo of the bridge.) Also walking to the outer suburbs and back would be a big hassle and make an already long walk really tedious. So both sides of the Trym are the same node, and the bridges over the Trym that do get crossed as part of this walk are internal bridges linking the north bank to itself.

There’s a similar argument for the Frome:


This one’s easier, actually, because it’s mostly culverted in the centre and doesn’t really divide up Bristol at all. There are a couple of smaller rivers in the south, the Malago (also culverted) and Brislington Brook (tiny) which are even easier to rule out.

So we’re settled on four nodes. Now for the edges. This is mostly more straightforward, but even here there are special cases:


Is a bridge like this two bridges or one bridge? The aerial photo makes them look very much like one bridge, as you can see the whole intersection, but as a pedestrian they feel more like two. It doesn’t matter too much, but it would be annoying if the rule was inconsistent in different places. In this case they all get counted as two.

One last decision. There’s a single ‘outlier bridge’ in Bristol, the motorway bridge at Avonmouth. Here’s a map of the whole route area:


This bridge does have a footbridge, so it ought to count. But it’s way out in the west, miles down from the next bridge, the Clifton Suspension Bridge. It adds maybe a third of the route just on its own. It would be a reasonable decision to leave it out, but also annoying, because it’s such a big important bridge and the walk wouldn’t feel complete. So it’s kept in. This actually works out quite well, because one of the nicest parts of the walk is the path through Leigh Woods on the way back from Avonmouth.

So… is the Bristol Bridge Problem soluble? It is, in the sense that you can make reasonable arguments for dividing Bristol up into nodes and edges in a way that gives you a soluble problem. It isn’t, in the sense that you could probably make different reasonable arguments for dividing it up in a different way that doesn’t. (It’s fairly robust to small changes, but if you dropped the Avonmouth outlier and counted the double bridges as single ones you could maybe get something that isn’t soluble). In the end the arguments for it being soluble were all good enough that I was happy to go along with them for the sake of an interesting walk. It’s soluble enough for my purposes.

I really like this as a demonstration of the work of problem formation. It’s easy to understand without a specific technical background: the mathematical model is a simple graph, and the decisions about how to map Bristol on to it are all easy-to-understand common sense sorts of decisions. But it’s still not a trivial toy problem you’d find in a textbook. Gross had to put days of work into finding a good formalisation, and the decisions are genuinely hard to call without being completely arbitrary. It’s a great example of mapping between a technical problem and the mess of the world.

Having opinions in public


[Written as part of Notebook Blog Month.]

I’m veering off the post ideas list today to talk about something that’s come up during the month. David MacIver’s been on a bit of a campaign recently to get people to try a daily writing practice, and has a good Twitter thread and blog post on how to get started. (His notebook blog is the main inspiration for my current notebook experiment.)

It turns out I have some opinions on this, focussed on a slightly different but complementary area: getting over emotional blocks around writing in public at all. If I had time I might try and rewrite this into a more generally applicable ‘advice’ type format. But I don’t, so I’m just going to ramble about what I did, and hope it’s possible to extract something useful out of that. Probably the main points are the classic start with something low-stakes and gradually ramp up, but maybe the details will turn out to be helpful too.

(I do still have an emotional block around anything that looks too much like ‘advice’ or ‘teaching’, actually. Like it’s claiming more authority than I have? I’m feeling some of that even writing this rambling sort of post – I mean, it’s not like I write some amazingly well-written mega-popular blog. But I do 1. regularly write posts that 2. some people read and get something useful out of, and that’s probably a more helpful reference point for people starting out anyway.)

The mechanics of writing blog posts came pretty easily to me. I know how to write in a fairly easy and unforced conversational voice, and have the result be coherent and entertaining enough that people will bother to read it. I have a decent intuitive grasp of how to explain why I care about a topic, and how to give enough background that people can follow along. I didn’t have to work on any of this deliberately and don’t have much insight into how I do it. I’m sure I could still improve a lot at these things, but it’s not been a focus for me.

Getting into a habit of writing frequently and coming up with topics to write about was hard at first, but over time both of these problems just solved themselves. I didn’t specifically put much thought into either, and David’s advice for getting round these at the start (minimal success criterion of ‘just write one sentence’, random book page to generate easy prompts) looks good to me.

What didn’t come easily for me, at all, was hitting the damn Publish button. Even under a pseudonym. I actually started from a level that seems extremely pathetic in retrospect, where I struggled to write comments on other people’s blogs under a consistent pseudonym. I’d write one comment under some name, and then I’d go back and nerve myself to look at it the next day, and it would look so cringily pointless and stupid that I’d never be able to make myself use that name again.

Come to think of it, I originally had this problem even with writing I didn’t publish. I’d write it, and the next day it would look atrociously bad, so I’d just delete it or bin it. So I have very little surviving writing from before age 25 or so. I’m not too sure how I solved that one, except that I kept going and eventually it looked less bad (combination of my writing improving and me dropping my standards way down from the very stupid ‘everything should be crafted carefully at the word level with no stale metaphors, Politics and the English Language style’ ones I’d picked up). I think this sort of ‘taste gap’ is very common when your ability to recognise good writing gets too far ahead of your ability to produce it, but I just had a really overly strong disgust reaction for some reason.

Solving writing in public was also quite accidental at first. I had a lot of time to waste and spent some of it going down the rabbit hole of something called rationalist-adjacent tumblr. I think it’s still going, but the peak was probably around 2015 or so. It’s hard to describe exactly what it was like, but a short version is ‘a bunch of clever but bored and unproductive people with fantastically distinctive writing styles and aesthetics having endless repetitive stupid arguments about dust specks and cupcakes and trying to befriend anybody who came along to join the argument.’ (Those two links give a better idea.) I never quite became part of the community there, more of a lurker and occasional commenter, but I did start writing. Mainly because it was incredibly low-stakes, and there was an easy ladder of increasing difficulty. First just reading and liking posts, then replying with the odd comment, then writing my own posts on low-controversy topics nobody was going to start a big fuss about. I started exploring my thoughts on mathematical intuition and discovering that I had something to say.

Around this time I started figuring out what was happening, and stretching my abilities deliberately. One big influence was Sarah Constantin’s post A Return To Discussion. This was aimed at people who were fleeing the spotlight of public discussion:

I have noticed personally is that people have gotten intimidated by more formal and public kinds of online conversation. I know quite a few people who used to keep a “real blog” and have become afraid to touch it, preferring instead to chat on social media. It’s a weird kind of locus for perfectionism — nobody ever imagined that blogs were meant to be masterpieces. But I do see people fleeing towards more ephemeral, more stream-of-consciousness types of communication, or communication that involves no words at all (reblogging, image-sharing, etc.) There seems to be a fear of becoming too visible as a distinctive writing voice.

But it also turned out to be useful for someone trying to learn how to do it in the first place. It called out some of my strategies and pushed me towards doing better:

You can preempt embarrassment by declaring that you’re doing something shitty on purpose. That puts you in a position of safety. You move to a space for trashy, casual, unedited talk, and you signal clearly that you don’t want to be taken seriously, in order to avoid looking pretentious and being deflated by criticism. I think that a lot of online mannerisms, like using all-lowercase punctuation, or using really self-deprecating language, or deeply nested meta-levels of meme irony, are ways of saying “I’m cool because I’m not putting myself out there where I can be judged. Only pompous idiots are so naive as to think their opinions are actually valuable.”

Very on point for someone with a tumblr called ‘drossbucket’! (That I’m-being-shitty-on-purpose thing did get me writing, though, even if it’s limiting in the long run. If you need it to start writing at all, I’d say go with it.)

That post also gets into the moral dimension of being able to have opinions in public, in terms of accountability and open debate. I’m not going to get into that, partly because it would take too long and partly because I don’t feel like I’m there yet. Still too conflict-averse to stick my neck out on most things that matter.

Still, I’ve kept working on it, increasing the difficulty level in small steps. I got this wordpress blog and moved over there, starting off by writing the same sort of short posts I wrote on tumblr. Then I started upping the quality standards and writing longer pieces that I’d obviously put time and thought into. No hiding behind the shitty-on-purpose label if it turned out bad. I got a Twitter account and started talking to people there. Then I attached the blog to my real name, but didn’t advertise that much. Then I added my real name and face to Twitter as well.

I was helped a lot by ending up in a fantastic corner of Twitter where people are really thoughtful and encourage each other with their writing projects, rather than trying to knock each other down and look clever all the time. Same for blog comments. I’d never have managed it otherwise. I’m not at the point where I could venture into some fraught culture war quagmire, or open up about anything deeply personal and upsetting. (I’m not even sure that I want to be able to do those things!) But by now there’s a pretty wide scope of topics where I’m able to blab out some thoughts with my name attached. And I want to keep pushing the scope of what I can do. Either directly on this blog, or by spinning up new alts to explore something different and merging them in.

So, yeah, I guess my advice did boil down to ‘start small and ramp up’ in the end. I hope the details are interesting to read though. Good luck to anyone trying this!

Thin technical terminology


[Written as part of Notebook Blog Month.]

Short one today on a weird aesthetic preference I have. It’s not particularly important but I’m curious whether anybody else has this, or whether it even makes sense to other people.

This relates back to the two types of symbol I was talking about a few posts back in Roses and traffic lights. Symbols as centres of clusters of rich associative meaning (‘rose symbolism in poetry’), vs symbols that are deliberately kept free of secondary associations so that they can stand for one ambiguous technical meaning (‘red traffic light means stop’).

Anyway the thing I really viscerally dislike is when I’m trying to learn a new technical topic, and a word that has a rich set of associations in everyday life is suddenly being repurposed for the ‘traffic light’ type of thin technical meaning. It’s just… bleurgh… and slows me down when learning certain things.

I’ll give an example. One place I’ve noticed this strongly is reading about cryptography and networking protocols. The technical vocabulary of ‘keys’ and ‘certificates’ and ‘sockets’ takes very ordinary words that have vivid sensory associations, and abstracts out one very specific core property. A cryptographic key isn’t metallic or shiny or toothed like a ‘proper’ key, but it does have the core property of having to be exactly correct, or it won’t open the thing you want to get at. I agree that this is the abstract heart of being a key, but somehow the word is still unpleasant without its cloud of visual and tactile connotations.

(There’s a connection to the Bell Jar stuff I quoted in the cognitive decoupling elite post – it’s similar to Esther’s dislike of the physics class.)

I know that if I stuck with the topic it would become better again. I’d start developing associations between the technical words, and then they’d start to have this rich meaning again. If I do get to this other side then technical topics suddenly become REALLY INTERESTING, and I can get obsessive. But I have to put in a lot of activation energy to get over this initial phase, and a lot of the time I just can’t be bothered.

I have the feeling that some people do have the ability to rapidly get interested in stuff just based on thin meanings, like mathematical definitions or rules for puzzles… but I don’t really know. Maybe they are just quicker at converting to thick meaning?

Seeing like a stationery aisle


[Written as part of Notebook Blog Month.]

Back when this was a thing people did, I was waiting in the queue at the post office. There’s an aisle full of stationery where you queue up, and the phrase ‘all stationery is ambiguity reduction’ popped into my head. It’s not quite true – probably the more boring statement ‘a lot of stationery is ambiguity reduction’ is closer to the mark. But I haven’t really thought it through. This will be my first attempt.

Why do I even care? I can think of a couple of reasons. Part of it is that in the last year or two I started to get a deep appreciation for the fact that we engineer the world to conform better to rational systems. I’d undervalued this before, and I still need to read more on this subject – Sorting Things Out and Seeing Like A State seem like obvious choices – but today is going to be more just freeform thinking about specific examples.

The other reason is that a while back I was watching Brian Cantwell Smith’s talk on ‘The philosophy of computation: meaning, mechanism, mystery’, and he had something interesting to say about digitality:

Digitality… is an extraordinarily interesting metaphysical notion, poorly theorised, interestingly enough, there’s very little philosophy of digitality. What I think is powerful, that Haugeland [I think, audio is hard to hear] recognised, is that implementing something digitally is a fabulous way of providing insurance against the mess of the world below.

Stationery provides a lot of examples of ‘proto-digitality’, in a sense that should be made clearer by the examples below. They’re not as binary and noise-free as electronic bits, but the purpose of a lot of types of stationery is to carve the world into X and not-X in ways that are relevant for the current task. This is another kind of ‘insurance against the mess of the world below’. That’s why paper and envelopes and filing cabinets and staples are so important to any kind of pre-electronic bureaucracy.

I think this is all going to become a lot clearer if I stop waffling and start going through specific types of stationery like I’ve just wandered in from Mars and am trying to understand what they do for the first time. The rest of this post is just a list of them, in whichever order they come into my head.

Envelopes. This is a very simple kind of proto-digitality. Everything inside the envelope is X, and everything outside is not-X. The envelope provides a way to transport what we would normally consider to be several things, like individual sheets of paper, as one single thing.

Wrapping paper. This is similar to the envelope, but also needs…

Sellotape. This seals up the edges of the packages to complete the boundary between X and not-X. It can also attach two separate things together to become one thing (like sticking a worksheet into an exercise book, so that the worksheet becomes ‘part of the exercise book’).

Hole punches. You can punch holes into a bunch of individual sheets of paper you consider to be X to then add them to a…

Folder. Similar to an envelope again.

Staples. Also similar function to an envelope, but a minimalist one that just does the categorisation and doesn’t protect the item in transit.

Paper. The workhorse of bureaucracy. Already come up in several of the examples above. I think it’s more raw material for dividing things into categories, rather than category division itself? Rather than dividing up a pile of objects, which may be large and messy and intractable, you divide up pieces of paper that represent them in some way. These are repeatable and stackable in a way that the objects aren’t.

Pens and pencils. These have a lot of uses, not all of which are necessarily ambiguity reduction. Shoving all of writing into that category is probably a stretch. But consider a simple example where you mark some pieces of paper with a triangle, and some with a circle. (Or any other distinguishing marks you like.) Then you’ve created two categories.

Scissors. Sometimes used to cut one thing to make it into two things, or to cut round a border between ‘stuff you want’ and ‘stuff you don’t want’. But there are other uses, like snipping the edge off a pack of frozen peas or cutting out a Christmas snowflake, that don’t fit so obviously.

Eraser. Tool to repair deviations from intended categorisation. E.g. you meant to write ‘eraser’, but you wrote ‘ersaer’, which isn’t a thing, so you rub out the offending mistake and try again.

Pencil sharpener. I feel like this is just a tool for fixing another tool, rather than anything to do with ambiguity reduction itself.

Ruler. Creates straight lines… to separate things into rows and boxes, or just because you want to draw straight lines I guess.

Post-It note. Attaches metadata to an object, with the added feature that you can remove it later, unlike marking with a pen directly.

Highlighter. Distinguishes some areas of a page as ‘important’ by visually separating them from the rest.

OK that’ll do for now. Feel free to argue with any of these in the comments, or add more examples! I literally just typed them out and then hit Publish, so there will be a lot of subtleties that I missed out.

A heap of broken images


[Written as part of Notebook Blog Month.]

I’ve had these notes sitting around for ages now for some kind of pretentious essay on… The Waste Land and ‘oddly satisfying’ videos? Who even knows. I’m never going to write it up properly, so I may as well have a stab at it as part of this notebook project.

So, the original inspiration is David Chapman’s page on the atomised mode. This is mostly going to be an elaboration (or rip-off) of some of the ideas there. Particularly this bit:

In our present, atomized mode of meaningness, cultures, societies, and selves cannot hold together. They shatter into tiny jagged shards. We shake the broken bits together, in senseless kaleidoscopic, hypnotic reconfigurations, with no context or coherence.

Those ‘tiny jagged shards’ reminded me of something else, the ‘heap of broken images’ in The Waste Land:

What are the roots that clutch, what branches grow
Out of this stony rubbish? Son of man,
You cannot say, or guess, for you know only
A heap of broken images, where the sun beats,
And the dead tree gives no shelter, the cricket no relief,
And the dry stone no sound of water.

This isn’t just a random similarity. The Waste Land is about the same process of the fragmentation into shards of meaning, but from an earlier phase of the process. (In as much as it’s about anything, I mean. I realise it’s a poem, not a lecture, and reading it in just the reductive way I’m going to in this post would be a bad idea.) Eliot is writing in 1922, against the backdrop of systems of meaning all in flames. In the third section of the poem, The Fire Sermon, he uses the Thames and its centuries of accreted histories to explore this unravelling. He compares mythic imagery of the past:

Elizabeth and Leicester
Beating oars
The stern was formed
A gilded shell
Red and gold
The brisk swell
Rippled both shores
Southwest wind
Carried down stream
The peal of bells

to random bits of incoherent junk in modernity:

The river’s tent is broken: the last fingers of leaf
Clutch and sink into the wet bank. The wind
Crosses the brown land, unheard. The nymphs are departed.
Sweet Thames, run softly, till I end my song.
The river bears no empty bottles, sandwich papers,
Silk handkerchiefs, cardboard boxes, cigarette ends
Or other testimony of summer nights. The nymphs are departed.

Only fragments of meaning remain, and the poem is made up from a disconnected collage of these. “Sweet Thames, run softly” is from Spenser’s Prothalamion. Possibly more echoes in there I don’t recognise.

(Not going to track down references because notebook blog, but my memory is that the coherence of meaning in the Elizabethan age is a recurring theme in Eliot’s critical essays as well. E.g. vividness and immediacy of imagery in Shakespeare and then the Metaphysical poets. Something like a ‘phoenix age’ in John Michael Greer’s retelling of Vico, to chuck in another reference.)

Anyway, it’s now nearly a century later, and we’ve hit new levels of incoherence. David Chapman’s example is Gangnam Style, which was current when he started writing the page:

Gangnam Style has been watched 2.9 billion times on YouTube. Even counting repeat views, it’s probably well-known to most young people on the planet. Its genre is, in fact, K-pop; but may be the only K-pop song most Westerners have ever heard.

Genre — which defined many subcultures — has disintegrated. Atomization seemed at first like subculturalism taken to an extreme, but it is a qualitatively new mode. K-pop may be a subculture in Korea, but in America it’s just YouTube. It’s normal for a Top 40 hit to mash up country-style pedal steel guitar with bubble-gum-pop vocals, hip-hop rapping, EDM bass, and black metal blast beats. “Authenticity”—the aesthetic ideal of subculturalism—is impossible because there are no standards to be authentic to.

I think atomisation goes one level deeper again, though. Gangnam Style lasts for whole minutes, after all, and is recognisably a song. It has a relatively stable theme, a recurring chorus, and similar images crop up throughout the music video. I’m interested in stuff that’s fragmented nonsense right down to the five-second level.

My favourite examples are from this wonderful article by John Mahoney in The Awl on chumboxes. Chumboxes are those ‘related article’ boxes full of terrible clickbait at the bottom of news sites, supplied by the likes of Taboola and Outbrain. The analogy is to buckets of chum: chopped up bits of fish chucked in the water to attract larger fish. Chumbox chum is designed to appeal at this same shark-brain level:

Like everything else on the internet, traffic flowing through chumboxes must be tracked in order for everyone to be paid. Each box in the grid’s performance can be tracked both individually and in context of its neighbors. This allows them to be highly optimized; some chum is clearly better than others. As a byproduct of this optimization, an aesthetic has arisen. An effective chumbox clearly plays on reflex and the subconscious. The chumbox aesthetic broadcasts our most basic, libidinal, electrical desires back at us. And gets us to click.

Mahoney clicked through chumbox after chumbox to get to the absolute worst chum possible, and analysed common themes. Here’s a couple of examples:

Top left: Sexy Thing and Localized Rule. We won’t dwell on the efficacy of a Sexy Thing in advertising. But do note this Sexy Thing, enhanced with a chummy sprinkle of sinister context (crime? Young women in handcuffs?). Here the Sexy Thing is combined with a more digital-age enhancement, the Localized Rule. Scouring a visitor’s IP for its geographic location, anxiety can be created by informing you of a brand new reason to find yourself handcuffed in the back of a squad car in your neighborhood.

Upper Middle: Oozing Food/Egg. A trend without an immediately recognizable psychological precedent? Oozing eggs are extremely common, and are possibly deployed under similar principles as Disgusting Invertebrates or Globular Masses Presented as Weird Food. Or perhaps the resemblance to an oozing pustular sore brings us back into the familiar realm of the Skin Thing?

The only aim of chum is to hit one-note reactions of disgust, body horror, horniness, nosiness, fear and greed, so that you click through. There’s no need to make sense at all. This isn’t confined to chum, of course. Other examples include:

  • those videos aimed at toddlers of opening Kinder eggs etc
  • ‘Oddly satisfying’ videos
  • ASMR
  • Some bits of TV Tropes would fit. Some of it is longer, more coherent narrative patterns, but some is just recognisable fragments of plot that appear on screen for a few seconds
  • Argument theatre. Short clips of the moment where ‘X DESTROYS liberal snowflake’, or whatever the 2020 version of this 2018 pattern is
  • Almost anything that’s bad on purpose to make you click

Surely this is the end of the line? I feel like we must be right out in the estuary of this process… can we even get much more fragmented than this? (Probably we can, and I just can’t see it.) This looks like the famous lines from the end of The Fire Sermon:

On Margate Sands.
I can connect
Nothing with nothing.
The broken fingernails of dirty hands.

Broken fingernails of dirty hands would make a pretty good chumbox image, actually.

Maybe this post looks like I think this fragmentation is purely a bad thing. Chumboxes are pretty gross, and there there’s all the Waste Land comparisons. Eliot’s response was essentially reactionary, looking backwards towards an Anglo-Catholic traditionalism that he hoped could hold the line against incoherence. His friend and Waste Land editor Ezra Pound later turned to outright fascism.

Actually, I think atomisation is a lot more interesting and positive than that. Again, I’m ripping off Chapman’s take here:

This may sound like a problem. Overall, my description of the atomized mode may sound like a panicked condemnation. However, there is much to like about atomization, and—I will suggest—it provides vital resources for constructing the next, fluid mode.

Now I have some examples of my own to think about, this is making much more sense to me. Maybe they can help to illustrate what these ‘vital resources’ are?

There’s enormous power down at the chumbox level of meaning. All the fragments have been liberated from any top-down need to make conceptual sense on a timescale greater than ten seconds, and appeal directly at the visceral level. Everything is very vivid and raw.

Still, to do anything useful with these fragments we need to build with them. We need to figure out how to recombine them skilfully from the bottom up, while keeping the energy. The kaleidoscope is maybe not the ideal metaphor here; it makes sense of the heap of broken images by imposing a rigid geometric scheme. We need something more local, provisional and reconfigurable. I don’t really know what, or I’d be writing that post instead of this one! But ‘build upwards from the fragments’ at least points towards the right direction.