one long token soup

Birdkeeping

Lachlan Kermode — Tue, 11 Aug 2026 00:00:00 GMT

[idea:birdkeeping]postsystems2026-08-11

Birdkeeping

We need new ways of talking about large language models. The artificial intelligentsia at frontier labs have beguiled us (and themselves) into using an exclusively calamitous tone of voice. If we are to take the Silicon Valley oracles at their word, the latest models are effectively sentient and thus pose existential risk, open weights models are as or more dangerous than nuclear or biological weapons, and the best thing we can do for humanity’s and the world’s increasingly unstable climate right now is to concrete over open green space with massive data centre developments.

Even among the more technically astute, something seems to have gone linguistically awry. Machine learning researchers compare their LLM hackery to the life sciences and quantum physics, as if the terabytes (petabytes? exabytes?) of numbers operationalised in unimaginably large data centres bear more than a passing resemblance to the atom bomb or the human body. An agent orchestration system called gastown, a kind of tool that many if not most software developers now use when coding, developed by an industry veteran, likens agential development to the Wild West and rhetorically reduces agents to low cost workers in a factory. (Gastown has more recently been upgraded into gascity, a framework that markets itself through the nauseatingly capitalist byline: “Your software deserves a factory”.) A respected VC investor and market analyst suggests thinking about LLM chatbots as a shortcut to infinite interns.

Let’s take a deep breath, and think sensibly about this language for a moment. If we metaphorically manage LLMs schizophrenically as both our workers and managers, our interns and our bosses, our slaves and our masters, all at once, their miraculous but jagged ability to comprehend and reproduce language is abased and apotheosised, all at once. We end up talking about robot rights and relitigating Pascal’s wager, rather than exploring the real potentiality and potentially ethical possibilities of working with LLMs.

Re-routing the flight path of AI discussion

I can understand why thinking about LLMs simply as machines leaves a little bit wanting. LLMs’ uncanny ability to comprehend written text and execute real virtual tasks on a computer when strapped into a coding harness isn’t something we’re used to thinking about as mechanical. Their actions are often not reproducible, and they can abide by the letter of your law while utterly ignoring its spirit. I recently referred to my ‘agent orchestration workflow’—the going industry jargon to denote a specific configuration of LLM use to assist your digital work—as a toolkit for taming jagged intelligence, because I think there is something animalistic about LLMs. Perhaps this something is that, like animals, we are vain to consider that they are actually ever under our power. As Gerard Wajcman argues, we project a kind of perfect world onto animals of many kinds as an imaginative escape from the travesty of humanity. ¹ It is a fantasy to think that LLMs will obey our commands perfectly once models get better, as language is an imprecise instrument. If LLMs exhibit anything that deserves to be called ‘intelligence’, it is of an animalistic variety: effectively opaque, imperfectly captured by anthropomorphised vocabulary.

With this in mind, I want to throw another metaphor experimentally into the oversized soup of AI takes. What if we are better off thinking about LLMs as birds? Birds can do many things that we humans cannot—deliver messages by air, glide for hours by locking their wings, fly, etc. There are also many things that humans can do that birds cannot. We are not, however, imminently at risk of a bird uprising that will result in humanity’s extinction (to the best of my knowledge).

There are many different species of birds. Albatrosses can fly for longer periods than other species, and hummingbirds can hover in place through rapid wing movements. Almost all species of bird in my native country, Aotearoa New Zealand, are flightless, as given the lack of mammalian predators (before the arrival of humans on the island), rooting around with long beaks in the New Zealand bush was a surer-footed way to find food than searching the skies.

Rather than thinking about LLMs through the mythology of ‘one model to rule them all’ (the ideology that frontier labs are pushing), I suggest that we’re better to think about how to cultivate a thriving and diverse model ecosystem in which many species of models coexist. Massive models may be useful and even environmentally justifiable for certain tasks, such as finding security holes in critical infrastructure. But smaller models are already sufficient for many of the use cases for LLMs such as codebase reconnaissance and harness engineering. It doesn’t make sense to use a Harpy eagle to deliver a small note when a carrier pigeon would do.

As this metaphor already begins to suggest, analogising LLMs to birds also makes the art of taming them appear as an activity that could be artful and personal, rather than necessarily industrial. What if working effectively with LLMs in our day-to-day lives is more like falconry than producing pins in a factory? We can send an LLM off to do certain kinds of work that we trust it to do well, while being cognisant that its operation is always subject to its own obscure and animalistic impulses, at least to some degree, no matter how ‘well trained’ it is.

From beads to birds

LLMs are excellent at implementing directed workstreams when strapped into coding harnesses. The issue is working out how to give them a structured set of directives so that they don’t exploit any unspoken ambiguity that might lurk beneath the surface of these directives. It is also important to track the work they complete so that we can both evaluate their work at reasonable intervals, and roll back or redirect them when we notice something amiss.

Sanguine use of a version control system (VCS) such as git or jj smooths over many of these concerns. So long as LLMs can package their work into comprehensible revision sets (‘revsets’), a human can reasonably both follow along and steer its course with a good amount of granularity. In order to queue up new tasks for LLMs to have a go at, some kind of issue tracking system also goes a long way. The hitch is that neither git nor jj provides an integrated issue tracking system. A VCS lets you register work that has already been done, but it doesn’t have an opinion on how you line up work that needs to be done.

Beneath the capitalist tropology of gastown and gascity (the AI orchestration systems built by Steve Yegge that industrially adopt and scale agents to try to build software), there rests the germ of an important idea in beads, the queueing mechanism for those bigger, more chaotic systems. As Yegge correctly notes, sprawling Markdown plans are a terribly unstructured way to keep an LLM on track: it can execute more complex activity in a reasonable fashion when given a dependency tree of atomically specified units of work.

Having worked with beads almost daily since November 2025, it is hard to imagine developing software with an LLM without something like it. Beyond allowing me to queue up work for agents, it also acts as a local issue tracking system across every codebase that I’m working on. The set of open beads across all of my projects represents all of the work that I have set out to do. Beads has almost fully replaced GitHub issues as my per-project issue tracking solution.

But the divergent philosophy of software development that beads was built to support has started to grate on my everyday use of it. Beads are not intended to be read by humans, and the beads CLI is not intended to be used by humans. I have been using the excellently named abacus, but even through this better lens, the experience of inspecting beads with human eyes constantly reminds me that they were designed to have a jagged intelligibility. Beads are supposed to be parcellated directives for a factory line of interned agents.

Birds are a conceptual alernative to beads. Let us think of birds as agent-ready units of work, analogous to beads, but designed with human legibility as a first-class citizen. Birds could therefore be a reconceptualisation of how to solve many of the same problems of keeping state while working effectively with LLMs in the software development lifecycle.

The art of falconry with birds

Let’s recap. Imagine that LLMs are not interns, workers, or superintelligences, but birds. These computational creatures can convincingly, stochastically parrot human language in the form of text and (to some extent) audio. We can also trust them to execute certain kinds of digital actions, just as we can trust pigeons that have been trained accordingly to deliver messages tied to their leg, or falcons to assist with raptor rehabilitation. I do not offer this metaphor to raise concerns about whether we should let the LLMs fly free, nor do I think of domesticating biological birds as a practice free from serious ethical concerns and questions. What I do believe is that an analogy between LLMs and birds captures the spiky capabilities of the former in 2026 (their ‘jagged intelligence’) better than an analogy between LLMs and humans.

As I have already suggested, if LLMs are birds, coordinating them to assist with the labour of software development is something like falconry. Instead of a cyber Charles Babbage, a silicon superintendent of the industrial information factory, as Yegge’s gascity positions her, I propose that we reframe the effective software developer in the age of LLMs as a master falconer, an algebraic austringer.

Footnotes

^ “We observe them with admiration, with envy, and also in grief, grief at having treated them so badly and resembling them so little. From nature, language keeps us at bay. Man is a denatured animal. We are animals sick with language.” (Wajcman 2009, p.131).

References

Wajcman, Gérard. 2009. “The Animals That Treat Us Badly.” Lacanian Ink 33. https://www.lacan.com/lacinkXXXIII7.html.

Taming jagged intelligence with beads and jj

Lachlan Kermode — Fri, 17 Jul 2026 00:00:00 GMT

[idea:taming-jagged-intelligence-with-beads-and-jj]postsystems2026-07-17

Taming jagged intelligence with beads and jj

There’s no need for a prelude on how agential harnesses such as Claude Code, Codex, and Opencode are changing the political economy of software engineering. The entry barrier to writing code has never been lower. LLM-enabled agential harnesses won’t replace software engineers wholesale; but their existence and reasonable accessibility (at least while frontier labs maintain monthly subscription pricing) dramatically refactor the economic calculus of building, customizing, and maintaining software. Paraphrasing a comment from Steve Klabnik at a conference I organized earlier this year, the hysteria around the imminent demise of software engineering and (sometimes, even) computer science is an ironic phenomenon with a touch of poetic justice. Software engineers have been threatening to automate humans—and every so often following through successfully on the threat—in other industries out of their jobs for decades. In 2026, our LLM overlords have deigned to descend from on high to give the most talented doctors of 10-times (computerized) capitalism a taste of their own medicine.

I’m lucky to be a graduate student in a humanities program in the midst of this 2026 overture to the great age of gentle singularity. Courtesy of the generalized inexistence of career trajectories in the not-explicitly-scientific academy, I was graciously disabused many years ago of the fanciful notion that I should expect a full-time job (let alone one that pays reasonably) to be waiting for me upon graduation. The economy isn’t structured to prioritize the kinds of work we should care about—but I digress. We are here to walk through what agentic engineering can look like, and how I use the terminal as my daily driver for research and development with LLMs.

For context, I am currently writing and maintaining a CLI tool in Rust, a distributed system in O(x)Caml, an interpretability research harness that is mostly Python, and several full stack platforms (Node, React, Bonsai, etc.). I run NixOS on both my desktop and laptop, and have primarily viewed, edited, and written code in a customized version of Neovim since approximately 2019. Despite this relatively unique personal computing configuration—and intractable codependency on vi shortcuts and text-based user interfaces—I believe that the workflows detailed below are nonetheless fairly general. One could set up similar environments on Mac or Windows, with VSCode or other code editors using analogous plugins.

Planning with `beads`

Beads is “a coding agent memory system” that Steve Yegge built and released in October 2025. It has since become the foundation for two relatedly wild software projects, gastown and gascity. Both of these urgently require critical attention from close readers, if only for the conceptual aesthetics of their documentation. I don’t believe in the full-blown ‘no code’ approach these two latter frameworks incline towards, in part because I don’t trust that AI agents will (always) write better code than I can in a shorter amount of time. The heart of my scepticism, however, lies in the suspicion that refusing to look directly at the code as a matter of (quasi-religious) principle bases itself on a misconception of the value of code in the enterprise of software engineering, whether it is small-scale or industrial.

When I have an idea for a software system, I refine my conception of it through observing the ways in which existing programming languages allow me to realize it. Code is one of the most essential mediums of this conversation between me, the machine, and the material social history of software in public or otherwise accessible libraries. Code is the way in which I ‘see’ the functionality and performance tradeoffs that have been made in every piece of software, including the software that I am actively developing or maintaining. Delegating the art of looking at code to an LLM is like building a house purely in 3D simulation, refusing ever to step foot in the physical space. It might be possible to build a house this way; but why would you not visit the construction site when the option is available to you?

What Yegge has got right, I think, is that LLMs can do better when we package the work we give them more thoughtfully. LLMs exhibit a jagged intelligence (putting to one side the obvious pitfalls of surrealizing a machine’s capabilities as ‘intelligence’). If it is right to think about the ability to set an agent to work through an agential harness as analogous to having infinite interns, then it is clear that we shouldn’t just let them loose on every half-baked thought that crosses the mind as we build out software. The beads CLI is one of the best tools that I know for packaging work in bite-sized pieces, so that an agent can chew through them without choking.

I currently use a fork of Yegge’s original codebase called beads_rust (br) and a TUI to visualize beads in a repo called abacus. ¹ Here’s how abacus looks when open in a repo:

Each bead represents one ‘unit’ of work. I create the majority of my beads via a planning process in which I work collaboratively with an agent in order to file features and bug fixes as epics (associated sets of beads), although I also file and refine beads through review as work progresses. My planning process is aided by a global CLAUDE.md detailing codebase-agnostic workflows combining beads and jj. The only general directive I currently give here for beads creation is:

**Bead names:** Keep them as short and simple as possible. Prefer concise 3-4 character identifiers over descriptive hyphenated names. For example, `rwq` is much better than `airborne-splash-rwq`. The bead ID carries the identity; the name is just a local shorthand.

Beads are a useful way of structuring complex work and keeping state as a project progresses in their own right, rather than relying on an agential harness’s in-built memory. If you run out of tokens mid-feature, another agent can always pick up work from where the active bead begins, provided you’ve kept their scope in check. They also shoulder some of the state-keeping burden that agential harnesses do mostly (as far as I can tell) through writing Markdown files to disk. It’s not likely that we’ll get away from Markdown in agential harnesses any time soon, as I suspect that it’s effectively packaged in as part of the reinforcement learning stage of model training. But I’ve found that working first to architect beads epics representing a project’s roadmap seems to free up space for the agential harness to run through tasks more effectively, as it’s common knowledge by now that these models perform more reliably when the scope of the prompt is smaller.

Beads really come into their own, however, when coupled with a version control system like jujutsu or git.

Version control with `jj`

Because LLMs cannot be depended upon to write high quality code, version control (VCS) is an essential part of agential engineering. ² Jujutsu is a lighter touch, more powerful, and back-compatible VCS in comparison to git, which is why I use it almost exclusively as my version control system. ³ I imagine that the processes explained below, however, are mostly reproducible using git as a substitute for jj if you prefer.

Here are the elementary steps that I define in my CLAUDE.md to harness beads and jj together:

**Per-task sequence:**
1. `br update <id> --status in_progress`
2. `jj log` — if empty unnamed commit below working commit, name it: `jj describe -m "..."`
3. `jj new` — fresh working commit
4. Do the work, run tests
5. `br close <id> --reason "Done"` — close BEFORE squash; this writes `.beads/issues.jsonl` into `@`, which gets included in the next squash
6. `jj squash --use-destination-message` then `jj describe -r @- -m "Present tense description"` — using `--use-destination-message` avoids the interactive editor that pops up when both commits already have descriptions
7. `jj log` — verify history shows correct author on each commit; `@` must be empty and unnamed

As an agential harness is working on a beads issue, it is marked in_progress, so that I can easily locate it in my abacus view. Once it thinks it has completed the work, it squashes it into a distinct revset so that it is packaged and ready to review. I use a combination of neovim and tuicr to review the changes, which gives me a rough equivalent of the GitHub PR feature set, i.e. adding comments to independent lines, blocks of code, or on the changes in general, and seeing everything with syntax highlighting—but all locally on my machine. I’ve also configured my editor with a slightly enhanced form of jj.nvim so that I can easily switch between jj commands, real neovim, and tuicr:

This set of customizations to Neovim makes it my one-stop shop for code review. I ferry reviews from tuicr sessions to my agential harness through the beautiful simplicity of copy-paste: I have a shortcut to :clip the full set of comments (a tuicr feature) to clipboard, and then I navigate into my other tmux pane where Claude Code or similar is running. A screencast of the side-by-side setup of tmux, abacus, neovim with jj.nvim and tuicr, and Claude Code is at the top of this post. I refer the detail-oriented reader to my NixOS config, which contains the mechanisms for neovim customization, versions of software, etc. (If you are part of the lilliputian part of the population who doesn’t use NixOS, point an agent at my post and the repo and ask: “how do I get the same Neovim/tuicr/jj.nvim setup in my terminal given I am not on NixOS?”. Your mileage may vary.)

I generally instruct my agents to package revisions in response to review as new jj revsets. The exception to this is when the work they’ve done is wrongheaded in some fundamental way, in which case I just squash their fresh approach into the revset with their first attempt. In projects where I am making related changes across multiple repos, I generally start an agent at the root and then selectively file and work on beads issues in individual repos.

Reflections on context engineering

The power of automatic interns comes with a kind of responsibility, though one more gentle than the ethical responsibility we should have if the interns were real people. While I don’t have deep blue, I do think that the art of software engineering is changing dramatically. I don’t think that the category of software engineers will be automated out of existence. (This thought suffers from a category error regarding why software engineers exist in the first place.) Engineering is about working with meaningful concepts at the boundary of scientific understanding and machines. Computer engineering is about working with concepts that map scientific understandings of how computers and code work to what concrete computers and code actually do. As long as we want computers to do things for us, some form of computer engineering will exist.

My undergrad advisor Arvind Narayanan recently gave an excellent keynote at ICML puncturing some of the inflated notions about the death of computer science. He grounds a set of his claims in an historical analogy between electrification and LLM-ification. In the same way electricity illuminated (pun intended) an entirely new scope of what was possible in physical space, LLMs and their effective deployment in agential harnesses change the way we (should) think about software production and maintenance.

Maybe now more than ever, we need careful, calculated, and critical thinking about how software works and how it is deployed and maintained. We need to be able to adjudicate when software can help us do something more effectively, and this shouldn’t just mean ‘faster’ or ‘at lower cost’. It is ridiculous to think that LLMs will be ‘better’ at tasks ‘in virtually all domains of interest’ (Nick Bostrom’s definition of superintelligence), as this would require us restricting the domain ‘of interest’ to exclude activities like ‘having sex’, ‘enjoying a meal’, or ‘watching sport’. But LLMs do change the state of the art of software development in a meaningful way, and one ignores shifting goalposts in artistic and scientific arenas such as computing only at one’s own risk. As local models become viable, and as frontier labs stick their heads in the sands of the arid inevitability of climate collapse for the sake of competitive advantage and profit, we ought to take political and social action accordingly. Getting a practical sense of how these jagged creatures do and don’t work is par for the course.

Footnotes

^ Yegge’s ‘no code, let the agents at it’ philosophy sits very snug with Zuckerberg’s infamous adage ‘move fast and break things’. A month or two ago I woke up to find the upstream beads had broken compatibility with abacus, and through a bit of research found that the maintainer of abacus recommended using beads_rust. Given that I find it useful to look at code (as explained above), I also generally prefer to use software written in Rust, as this is the language that I use most actively these days and am most comfortable reading.
^ ‘Agential engineering’ and ‘context engineering’ here are terms opposed to their irresponsible younger sibling, vibe coding).
^ Others have given the expanded definitions of ‘lighter touch’ and ‘more powerful’ here.

Writing weeknotes

Lachlan Kermode — Sun, 07 Jun 2026 00:00:00 GMT

[idea:writing-weeknotes]postgeneral2026-06-07

Writing weeknotes

These past few days I have been inspired by Anil Madhavapeddy’s and Jon Sterling’s online presence to start a practice of weeknotes. Jon’s weeknotes offer a complete example of what weeknotes are, and I also recommend his post on intellectual junkyards as an overview of how the practice emerged. As he notes in that post, there was a kind of organic friction with the ‘timeless’ design architecture of Zettelkasten note-taking that pushed him towards weeknotes:

Anil’s research group at Cambridge has a practice of internal blogging and weeknotes: pretty much everyone is writing about their work in blogs that are then syndicated into a Matrix channel. Then, discussion proceeds organically in the chat and in person. These blog posts are not the transient/fleeting notes discussed by Matuschak, because they are not scraps—they are evergreen in the ‘free’ sense of being permanent discourses on a moment in time.

I have experimented variously with the practice of academic journaling, a public-facing grad log (which turned into this very-much-not-daily blog site which consists mainly of book reviews), and time-tracking softwares such as bartib to try to organize the grand structurelessness of time that is the making and breaking of the graduate student (in the R1 American institution). These writing exercises have been valuable, and have helped me to crystallize and record aspects of my thought in ways that would have been otherwise lost to the vagaries of individual memory. Reading about Anil’s group, however, reminded me that I still do crave the self-imposed structure of a regular interval through which I can hold myself accountable, given that no-one else will in the ABD (All But Dissertation) period.¹

What I love about the idea of weeknotes and how Anil’s research group seems to work (though note that I have no first-hand experience of it, I have just been browsing their activities online) is how a set of relevant notes can be federated (using RSS, the great untapped potential/protocol on the web) to form a notes network. This network can then serve as the basis for a richer collaborative undertaking, wherein sets and subsets of individuals can follow each other’s research activities without the cumbersome ceremony of video/in-person catchups. Meetings of course still have their place: but there is much than can be shared, refined, and progressed through a practice of sharing regular writing in advance. As Jon Sterling writes, “[j]ournalling in the most progressive sense is simply blogging with a restricted audience”. Federated research journalling, through weeknotes or otherwise, is exactly the kind of collaboration that the Internet makes possible—yet there are too few of us who take advantage of this information highway out in the open, sharing our activities semi-publically online.

Weeknotes also seem to me exactly the kind of structure that could cohere an ABD graduate student’s final few years, especially one in a humanities program, meaning one with an intellectual commitment to the production of writing. As I have noted before, it has often seemed to me—perhaps counterintuitively—that there is greater intellectual generosity on offer in the computer sciences than the humanities: ‘perhaps’, as given the sharply precarious political economy of the latter in 21^st century University, this may come as no surprise to some.² The point of the Research University, to my mind, is to facilitate the production of knowledge that structurally escapes the imagination of the private sector, and to allow this knowledge to be shared in ways that challenge and subvert the vociferous demands of financial and economic reason. While I do not think that making information ‘free’ for all is a wholly sufficient measure to take against capital’s inexhaustible drive for surplus value, privatizing writing and research for the sake of a symbolic hierarchy (so as not to dilute one’s peer-reviewed writings with off-the-cuff thoughts in paltry blog sites, as the standard academic reasoning seems to go) seems a much more suspect route to take.

It also excites me that many in Anil’s group are building their own homebrew infrastructure to host their notes and research. Anil’s website runs on a system called Bushel; Jon Sterling has an impressive tool called Forester; Patrick Ferris uses Forester with a preprocessor he wrote called Graft; Michael Winston Dales has an OCaml static site generator. My weeknotes will of course be written and published using Rheo, the CLI tool I have been working on for a few months now that is part of the document infrastructure for augmented reading we are working on at the Free Computing Lab.

I intend that these weeknotes will allow me to continue developing Rheo not only as a more fully featured static site generator, but also to add a roam-style package for HTML export to allow backlinks, tagging, and many of the other features that exist in Forester and Bushel. One of my complaints about roam-style systems is that they are hard to adopt/navigate unless one is already quite familiar with the architectural ideas that support them. Thus I want to keep ‘vanilla’ Rheo un-roam-like so that it can still be used as the driver for sites like the Language of Language Machines, Screening the Subject, the Digital Theory Lab website, and the Rheo docsite.

And so, announcing my research weeknotes, started the day of this post—June 7th— which you can find publically available at:

https://weeknotes.ohrg.org

Footnotes

^ And nor should they. The beauty of the small, temporary, contingent, but nonetheless real pockets of academic freedom produced by the University is precisely that the only person telling one what to do is oneself.
^ For LRB subscribers, I recommend Stefan Collini’s latest article for a thoughtful rumination on the state of this nation in Britain.

Ben Tarnoff and Quinn Slobodian: Muskism

Lachlan Kermode — Thu, 09 Apr 2026 00:00:00 GMT

[idea:muskism]postreviews2026-04-09

Ben Tarnoff and Quinn Slobodian: Muskism

After buying a hard copy of this book at the RiffRaff launch with Quinn Slobodian—Ben Tarnoff pulled out briefly before on account of illness—it has sat on my bookshelf for a few weeks while I’ve pottered around fulfilling various obligations to others and myself (a conference, a short presentation about my dissertation, a bikepacking route over the alps from Switzerland to Italy).

Muskism reads like a trade book. At least, it reads like it has been edited by an editor who doesn’t trust readers to hold arguments and episodes in their head from one chapter to the next.

I don’t judge this ‘tradey’ nature a bad thing, necessarily. Maybe many readers don’t, and the repetition certainly does give color to the vignette of Musk as a consciously contradicatory and thus psychoanalytically coherent character who represents a distinctive brand of 21^st century capitalism. Tarnoff and Slobodian cast a thread that loops together the parts Musk played/plays in PayPal, SpaceX, Tesla, X (formerly Twitter), Neuralink, and DOGE (both the Department of Government Efficiency, and the memecoin), hemming them together with alliterative seams such as “fortress futurism”, “attention alchemy”, “becoming bot”, “cybernetic collectives”, and “sovereignty as a service”,. The apparent mistrust of the reader’s memory from chapter to chapter does, however, make one think that a more more concrete theoretical position might have been developed had so much time not been spent looping back on previously established episodes.

The book’s subtitle—“A Guide for the Perplexed”—suggests a critical introduction to the life and times of Elon Musk, and it more or less delivers on this suggestion, to my mind. Muskism pieces together Elon’s erratic activities as milestones that mark a kind of progress towards a unified ideological project despite themselves, despite their apparently groundless individual motivations, and in doing so constructs Musk as iconic of his own memetic syntagm of capitalist caprice, ego, and devastation.

I’d recommend the book as a way to get churning on the empirical chaos of America’s catastrophic tech economy. Musk is certainly an important agent of this chaos, and the book covers his most notable shenanigans. In both the RiffRaff discussion and some others with the authors that are circulating online, a fair bit was made of Musk’s upbringing in apartheid South Africa, as Tarnoff and Slobodian present this in the book’s first chapter as the establishing shot of his psyche-qua-politics. Rearing its ugly head in recent years, Musk’s swing to the far-right in recent years, accentuated by his calls for the ‘repatriation’ of immigrants to their home countries across Europe and the United states, cannot but be compared to the racist and segregationist regime of South Africa’s Nationalist party. Tarnoff and Slobodian’s take here is not vulgar psychoanalysis: it is not that Musk’s childhood in South African apartheid represents a totalizing historical explanation for his xenophobic tendencies today. But this history is not unrelated, and nostalgia—the (mis)remembering and (mis)valuation of past events—weighs heavy on our estimates of the present’s psychic composition. The milieu of a billionaire’s troubled childhood is fertile material for a critical snapshot of what motivates the drive for capital, or lies beyond it.

Writing in Typst | Hacking on Neovim with Claude

Lachlan Kermode — Sat, 01 Nov 2025 00:00:00 GMT

[idea:writing-in-typst]postsystems2025-11-01

Writing in Typst | Hacking on Neovim with Claude

What is a ‘good’ writing system?

I have been incrementally hacking on my writing environment for some time now, since at least 2013 when I started seriously using computers in undergrad. A couple of years ago, I migrated to Orgmode as the best markup syntax for my needs, and wrote a post about how Emacs and Orgmode serviced my writing needs.

Here’s a summary of that post and the core tenets of what I consider an acceptable writing environment, parsed out over the five or so years I’ve been experimenting with one through grad school:

Flexible, powerful and distraction-free. In short, this means that the environment needs to be an extension to a modal editor in the terminal. I started using a modal text editor around 2018, and use a range of ergonomic keyboards in funky ways that make using a mouse undesirable in most cases. (The web browser is the one environment where I still get some mileage out of a mouse. I do a lot with keyboard shortcuts via Vimium, but there are still some contexts where it’s just quicker or more comfortable to use a mouse.) One of the main reasons that I settled on Orgmode rather than, say, Markdown at the time was because of its more standardized bibliographic management.
Non-proprietary and sane markup format. Microsoft Word documents and Google Docs are great for a lot of things, but I refuse to rely on either of them as a primary format for all of the writing I do, as their formats are too hard to parse (to write custom software for) and bound to Microsoft’s and Google’s ecosystems respectively. The ability to run Unix-style commands on a simple markup format from a terminal to search and replace, for example, is an essential. Writing documents in a plain-text markup language also gives me the safety of knowing that, if it really came down to it, I could write my own parser and compilers. My writing archive shouldn’t strictly rely on some company’s infrastructure to host, search, or otherwise make use of the thought it contains. Using such a format also means that cross-platform editing is made simpler and possible. (I run linux mostly, but still regrettably use Android as my phone’s operating system.)
Multi-format export. Most of the world’s documents are still PDF. There’s no getting away from needing to export writing as PDF in many cases– for e-readers like the reMarkable that I use, or for submission to conferences. But we increasingly read writing on a web page of some sort, and so I also need a workflow to export fully functional documents to HTML and CSS, too. Other formats that are interesting if not essential include some kind of presentation file (PowerPoint, or better: just a website that has slideshow-like interactions), Markdown for rich formatting to copy somewhere, and plain text.

I have up until very recently used Orgmode as my markup language of choice, exported them to PDF with latex, and exported them to HTML with pandoc. But I am very attached to the Neovim ecosystem for my code editing and writing, and so it was clunky to open up an Emacs installation (that I barely understood) exclusively to edit Orgmode. So I switched to editing Orgmode in Neovim along with everything else, using plugins and custom functions to get towards the writing experience that I wanted.

This has actually worked surprisingly well, but it has some sharp edges. One of the more significant ones is that any time I want to produce anything more complicated than basic, formatted text with citations and footnotes– for all of which pandoc transformations produce reasonable output in both HTML and PDF– I need to start embedding LaTeX into Orgmode, and deal with the LaTeX toolchain / dependency management in order to compile a PDF. Similarly, if I want to produce an interactive HTML document, I need to embed the source code directly in Orgmode and ensure that the export process handles dependencies and the like appropriately.

Some of this is unavoidable. If I want to run custom Javascript in a website that is well beyond the expressive capacities of a markup language, at some point I just want to be able to write Javascript. But what I found frustrating about my Orgmode / LaTeX / HTML workflow is that there wasn’t any reasonable way to work towards extending the markup language in some ways, unless I was willing to start developing my own bespoke flavor of Orgmode plus plus. I also don’t particularly like wrestling with the LaTeX ecosystem, because– and this is hardly controversial to say– LaTeX has a lot of bloat. What I wanted was a more extensible system which had saner defaults.

Enter: Typst

A few months ago, I started seriously considering typst as a potential replacement for LaTeX. At the very least, I thought, it would be more fun to wrestle with a modern ecosystem when struggling to produce some custom table or figure in my output PDF, as typst has a layout system that uses terms that are a lot more intuitive to me than the black magic of laying out LaTeX documents.

It just so happened, however, that I started to follow typst development more closely at a time when the final touches to the basic foundations of HTML export, such as footnotes and bibliography, were just about to be added to the upstream. So I made a few contributions to spirit it along, and started more serious experimentation using typst as a unified way to produce both PDF and HTML in my writing environment. Pandoc can convert to and from typst, so I originally intended to keep writing documents in Orgmode and then transiently convert them to typst in order to produce PDF and HTML both. But I quickly found that the typst syntax natively accommodates all of the features that I make use of regularly in Orgmode such as citations, footnotes, headings, links and text decoration– and then some.

So why not write my blogs, papers, and documents directly in typst? I considered the critical features of my Neovim / Orgmode writing environment that I didn’t want to abandon:

Shortcuts for markup. The nvim-orgmode plugin makes writing Orgmode in Neovim pleasurable, providing shortcuts to insert a link and basic text decoration while composing.
Citation and link picking. Though I’ve gone without it for a few months for reasons that are immaterial here, I used to have a shortcut to bring up a fuzzy finder for all of my bibliography entries to easily insert a citation. The same fuzzy finder would make it easy to link to local files (in a website, for example, to link to other posts).
Document folding. The ability to fold away all of the text beneath a heading is very useful when navigating larger documents, as it helps me to compartmentalize writing tasks and organize longer documents such as a dissertation chapter.
Export shortcuts. I have customized my Neovim editor so that I can easily export the active Orgmode document (through the pandoc and LaTeX processes described above). Personally, I don’t feel that I need a real-time live preview of the document as I type, as I generally just want to check that it looks reasonable at certain junctures in the writing process, rather than continuously.

The one other feature of Orgmode that I have come to rely on heavily is its TODO functionality. I typically only use this in notes related to projects or tasks more generally, however, and not in documents that are intended for publication such as a paper or blog post.

Enter: Claude Code

At this point in the past of a new writing technology’s prospecting, I would go searching for a Neovim plugin for typst and hope that it provides features that satisfy a majority of these requirements. I’ve spent a fair bit of time tinkering with my init.lua, the entrypoint for customizing Neovim, but I’ve never had the time nor interest to sit down and write a plugin from scratch.

LLMs, of course, are at time of writing taking the coding world by storm. I have started moderately relying on Claude Code when writing some– though certainly not all– kinds of code. As is well-known by now, Claude is especially good at scaffolding hacky scripts or modules from scratch, when no large codebase or domain-specific knowledge needs to be kept in context. A Neovim plugin, I realized this morning, is a pretty ideal domain for LLM-assisted coding. The ‘codebase’ is often just a single configuration file, and the domain-specific knowledge is the Neovim editor itself, a well-documented and expansively customized software for which there are many examples on Reddit.¹

So I fired up Claude Code earlier this afternoon, and– fast-forward an hour or two– I have a fully functional writing environment for Typst that essentially has feature-parity with my Orgmode environment. Moreover, my Neovim config is now much more comprehensibly modularized; and I have a tried-and-tested method for extending it without needing to spend days learning the ins-and-outs of Neovim’s API; and some bugbears in my NixOS config were eliminated while I was at it. (If that last bit means nothing to you, count yourself lucky!)

My new writing environment

I use treesitter for syntax highlighting, and Typst already looks pretty good with it. I get function completion by integrating an LSP for the format, for which I’m using tinymist.

As I noted above, I haven’t had dynamic link or citation insertion for some time. It was one of the features that got lost in my move from writing Orgmode in Emacs to writing it in Neovim. I use telescope.nvim for general search and file-picking when coding in Neovim, and I figured that I could use a customized pop-up to dynamically pick available citations from the relevant BibTeX file, too. After a few minutes of vibe-engineering, I have the following:

When I am writing in Typst, and I want to bring in a reference, I can open a panel. Note that the search is full-text, not just using the reference ID. I also have a shortcut to specify which bib file to use through the #bibliography function in Typst. I can insert links in the same way as citations, both references files relative to the current one (blog posts on the same site), and external links. Both the citation and link insertion work either by highlighting text and annotating it, or to insert new links/citations. I also have a similar shortcut to add footnotes.

This is pretty functional now for generic writing!

Future work

Typst isn’t ideal for producing fully-featured websites currently, as HTML export is experimental. Even when it becomes better supported, the project is– understandably, given its priority supporting PDF– taking a relatively conservative approach to HTML generation. Anything that doesn’t have a robust analog in a PDF document, such as videos and hover panels, will have to be ‘embedded’ in Typst with HTML/CSS/JS, rather than being written in Typst syntax. The current experience isn’t much worse than Orgmode with Pandoc, though, and the Typst roadmap promises that it will become much better in the relatively short-term future.

There is a longstanding issue that I’ve had with links in Orgmode that I haven’t yet tackled with Typst. When I’m writing, I like hyperlinked text to appear as it will in the final document, i.e. without the underlying URL on display. When editing any particular line, though, it’s better that all of the links are ‘expanded’ to their full source syntax (#link("...")[...]) so that it’s feasible to edit the markup without requiring any fancy shortcuts. The effective shortening of lines that occurs when hiding these URLs results in different Neovim line-wrapping requirements, with which the Orgmode plugin I have been using does a bad job, giving ugly linebreaks in documents with long links. This link presentation will likely be the next feature I add to my Neovim Typst plugin.

I’ll add to the capabilities in my Neovim config files, and might eventually release a separate plugin if the features become significant/mature enough.

Footnotes

^ It’s impossible to mention LLM coding at this time without adding some sort of disclaimer that, no, I don’t think AGI is around the corner, and yes, I do expect both programming languages and language writ large to remain ‘a thing’ in the foreseeable future. LLMs are an incredibly powerful tool to write and analyze code and text, but the purpose of code and text– as a medium of symbolic communication amongst social beings– has not been rendered valueless since ChatGPT became publicly available. If anything, the value of adeptly and adroitly handling written language has taken deeper root. For my preliminary thoughts on why we are so keen to imagine that computers will supplant the usefulness of the human, I refer the reader to this talk I gave in 2024.

Mladen Dolar, Rumors

Lachlan Kermode — Sun, 15 Jun 2025 00:00:00 GMT

[idea:mladen-dolar-rumors]postreviews2025-06-15

Mladen Dolar, Rumors

First on the syllabus in Rohit Goel's current BICAR course is Mladen Dolar's 2024 book Rumors. This short book is an extended and adapted version of Dolar's contribution to the 2022 edited volume Objective Fictions: Philosophy, Psychoanalysis, Marxism, and in its first sentence it lays out the topic at hand:

To take rumors as a serious object of theoretical scrutiny and philosophical reflection seems like a contradiction in terms…. [C]an rumors acquire the dignity of a concept?… If one deals with such an unworthy and undignified subject, isn't this a sure way to ruin one's philosophical reputation? (Dolar 2024, 1)

Dolar has become something of a specialist essayist (in what he writes in English, at least). He published the delectable What's in a Name? (which I wrote about here;) in 2014 as an extended essay associated with an art exhibition at the Institute for Contemporary Art in Ljubljana, and there is more than a passing resemblance between this book and Rumors. Both begin by rhetorically turning Western philosophy on its head. The first sentence of What's in a Name? is: "It all began with Plato" (Dolar 2014, 6), and Rumors, after a brief digression through the winning explanation of how quantum physicists understand the eerie materiality of Higgs boson ('the God particle'), takes us to "the Socratic foundations of philosophy" (Dolar 2024, 5) in the distinction between doxa (opinion) and epistēmē (knowledge). Doxa will come to be associated with rumors over the course of the book, although it requires important clarifications. Opinions have a subjective source, someone traceably responsible for their elocution. Rumors, on the other hand, distinguish themselves in that they run on the wind, which is to say autonomously, words that appear through the grapevine without a source. Rumors are a peculiar kind of language that circulate socially without a clear cause. Perhaps this is why, in the part of Vergil's Aeneid that Dolar cites after his Socratic introduction, "they go around like a breeze of air, stemming as if from nowhere" (Dolar 2024, 6). Rumors are a substance that appear materially as an effect in the world, but with an epistemologically (and perhaps ontologically) obscured origin.

So Rumors starts with Socrates, but not directly with his more respectable philosophical ideas (which we might find in The Republic, for example), but with the trial that led to his death. A more serious philosopher might have dismissed this incident as mere historical 'context' that only bears indirectly on what Socrates has to tell us philosophically; but Dolar is no such supercilious simpleton. Following Freud, perhaps– as well as Marx, I can't help but add, another master of suspicion– Dolar's inquiry is motivated by the idea that often the most apparently unimportant aspects of a thing are essential to what is at stake in its essence. Philosophy, Dolar's gesture seems to suggest, cannot be divorced from its historical situation. As Fredric Jameson famously quipped, it behooves philosophy to 'always historicize'.

Crucially, however, philosophy can also not be reduced to a plain and simple historicization of its actors. Whereas the identity-political criticism of Socrates might dismiss him as a White Man philosopher, resultantly disregarding all his ideas as tainted and thus in need of discarding, Dolar's historicizing approach is not quite so crass. The more interesting reading of Socrates in the twenty-first century, Dolar suggests, is not to castigate the man himself but to suggest new ways to read what he stands for. (The real political impact of cancelling dead men, white or otherwise, needs some serious thinking through.) What if Socrates' trial shows us (students of the history of philosophy and the philosophy of history) the 'first' true triumph of doxa over episteme, the opening scene of a history of modernity that can be characterised by the trading of blows between these two sides of the same coin, of speech/language, of the big Other (episteme) and its shadow/symptom (doxa)? The history of Western philosophy does begin with Socrates (we do ourselves worse by denying it), but its beginning there does not necessarily mean that from that day on, only White Men can do philosophy. Dolar's reading suggests something similar to the 'decolonial' reading, i.e. that Socrates inaugurated a particular tradition of knowledge that has egotistically taken centre stage in the history of philosophy. But it responsiblizes Socrates in a very different way, refusing to dismissively cancel him on account of an inexact identity in favor of a more psychoanalytically generous approach which sees him both as a subject capable of his own decisions and as a victim of the times in which he found himself.

Dolar will repeat this gesture throughout the book with a canon of other philosophical and literary greats– Vergil (The Aeneid), Shakespeare (Hamlet, Henry IV, The Comedy of Errors), Mozart (The Marriage of Figaro), Rousseau (Confessions, Rousseau Judge of Jean Jacques), Kafka (/The Trial, The Castle), Cervantes (Novelas Ejemplares), Gogol (Dead Souls), Kierkegaard (the Corsair affair), Heidegger (Sein und Zeit), Balzac (Illusions perdues). The works of each of these figures inflects the inaugural gesture of Socrates' trial. And it is usually the lesser considered works, or the lesser considered aspects of well known works that reveal this inflection. Doxa stalks episteme as its shadow, conceptually inseparable from it. Knowledge does not work if unfounded rumor does not cohere a social setting to aspire to it.

The one great literary and philosophical work that Dolar's book briefly touches on that I want to examine in more detail here, however, is (surprise!;) Karl Marx's Das Kapital. Dolar mentions Marx in "a small digression" (Dolar 2024, 16) in the section 'Rumors and Capital', which appears between the classics (Socrates and Vergil) and Shakespeare. Rumor's tempest recalls Marx's account there, Dolar tells us:

An analogy can be drawn with another mighty entity that expands and multiplies seemingly by mere circulation, namely capital. There, too, we encounter the expansion of something that starts modestly, then increases as if through mere circulation, and is amplified into an overwhelming force, gradually engulfing everything as it goes along in a potentially limitless process. (Dolar 2024, 16–17)

It may seem like a sophistic nitpick, but I think it is important to qualify that the homology here should really be between rumors and value in capitalism. Of course, capital is the name Marx gives to value in motion, and so Dolar is not exactly wrong to say that we should see a relation between rumors and capital. But the analogy here risks suggesting that, just as rumors have haunted our pretensions to a clear-headed construction of sociality since at least Socrates, capital too is a sticky symptom that is co-constitutive with sociality rather than a particular form of social (non-)relation that could theoretically be done away with. Put differently, by drawing an analogy between rumors and capital, one risks projecting the apparent trans-historical role of rumor as an excess or surplus of language that nonetheless coheres the social fabric (a role for which Dolar makes a convincing case) onto capital, making it into an essentially trans-historical entity, too.

If we instead refine the analogy that sees rumors ghastly increase via circulation– ghastly, because it appears to contravene the third law of thermodynamics, entropy, whereby a value decreases as it moves– as homologous with value, then capital's stickiness as a symptom in society doesn't appear as inevitably transhistorical. Capital's pernicious philosophical alchemy is to render a historically specific set of social relations based on property rights and the wage relation as transhistorically inevitable, to transmute essentially changeable social customs into apparently natural laws that misrepresent things 'as they actually are'.

The commodity is Marx's paradigmatic example of this alchemy. In his infamous example of the dancing table (in the marvelous new English translation of volume I):

Of course, the table isn't any less made up of wood for having been worked on, and the wood remains an ordinary sensuous thing. But the moment thetable begins to act as a commodity, it metamorphoses into a sensuous supersensuousthing. It doesn't simply stand before us with its feet on the ground; rather, in its relations with all other commodities, it turns upside down and spins bizarre notions out of its blocky head, aperformance far more fantastic than if it were to start dancing of its own accord. (Marx 2024, 47–48)

Objects come to appear imbued with the qualities of subjects (living humans), and contrariwise human laborers appear as if they are just another input feeding the autonomous subject of capital rather than as social ends in their own right. Commodity fetishism naturalizes the historically specific movement of value in capitalism, and makes the entire set of affairs seem the natural state of things. While Adam Smith was convinced that man was naturally selfish and thus capitalism an inevitable social relation, Marx denaturalized both this assumption and the conceptual system built on that premise in capital in Das Kapital.

Though he makes out that he will return to the connection between rumor and capital, Dolar does so only briefly in the final section of the book, 'The demise of the big Other'. There he takes up Zupančič's idea of 'conspiracy without a conspiracy theory' to bring capitalism back into the narrative as "just such a conspiracy without a conspiracy– a conspiracy that imposes the rule of One that subtends all conflicts and diversity, but is not orchestrated from the center by some mastermind?" (Dolar 2024, 111). (Boštjan Nedoh has evocatively called this idea in Marx's theory of capital as 'a theft without a thief.') Indeed, arguably what makes capital so sticky as a social relation is its ever-inventive disavowal of any responsibility for the immiserations that occur and seem to deepen over time according to it. To transpose this into more Marxian language, what is philosophically remarkable about value in capital as an entity is that it structurally covers over its own cause in the sphere of production, namely the wage relation's appropriation of surplus value from the worker under the prentense of a 'free and fair' relation. Like rumors, capital's existence in the world seems to just appear on its own steam, so to speak, "stemming as if from nowhere" (Dolar 2024, 6), as the fetishism at work in its machinations expropriates agency from human subjects and plants the appearance of that power in commodity objects. The process of its circulation covers over its real cause of production (labour) in the wage relation.

Here lies the real tenet of the homology between rumors and value in capitalism, I think. Both exist only and because of a lost cause, in a subject whose ownership of some thing has been disavowed, and so the phenomenon (rumors or value in capitalism) appears as a thing in itself. The subject-object relation is distorted, and the truth of the (human) subject as the object's unconscious cause goes missing.

Bibliography

Dolar, Mladen. 2014. What’s in a Name? Aksioma - Institute for Contemporary Art.

———. 2024. Rumors. John Wiley & Sons.

Marx, Karl. 2024. Capital: Critique of Political Economy. Edited by Paul North. Translated by Paul Reitter. Princeton: Princeton University Press.

Alma Steingart, Axiomatics: Mathematical Thought and High Modernism (introduction)

Lachlan Kermode — Sat, 14 Jun 2025 00:00:00 GMT

[idea:alma-steingart-axiomatics]postreviews2025-06-14

Alma Steingart, Axiomatics: Mathematical Thought and High Modernism (introduction)

I started reading Alma Steingart's 2023 Axiomatics: Mathematical Thought and High Modernism in preparation for a reading group on Math(s), Philosophy, and History that I have been running for almost a year now. (We meet approximately bi-weekly, and wander from text to text based on the group's discussion and shared interest.) Steingart's narrative is deeply relevant to the third stream of my dissertation, the question of mathematics in the 20th century, as it argues that there is an intellectual history of (the rise of) axiomatics in mathematics in the 20th century that corresponds to modernist cultures of art, linguistics, sociology, and other disciplines that became infatuated with mathematical rigor in the ascent of "the postwar abstractionist theory of knowledge" (Steingart 2023, 23). Concretely, Steingart promises (in the introduction) to provide an account of how topology and category theory can be understood as mathematical cultures that represent what she calls "high modernist mathematics" (Steingart 2023, 20), how the utility of mathematics in various disciplines took on a more 'abstract' character after the war in projects such as game theory, and how the abstract character of axiomatic mathematics bears a relation to the "aesthetic autonomy" (Steingart 2023, 22) of artistic formalism and structuralist anthropology, among other disciplines that were coming to be characterized by a flirtation with mathematization, during the Cold War era.

The book's introduction frames this effort in intellectual history through a smattering of terms that constellate the 'bigger picture' which theoretically coheres each individual chapter's analysis of more specific historical actors and moments. It took me a second reading to begin to feel like they were adequately disambiguated, but I now feel that they give a helpful coherence to Steingart's project. The important terms clarified in the book's introduction, which I will develop in pairs in this b/log, are as follows:

Abstraction
Modernism
High Modernism
Axiomatics
Mathematics
Mathematization

Abstraction and Modernism

Abstraction is the principal term that grounds Steingart's inquiry, and thus in the introduction she accordingly gives us two, mutually informing accounts of it: "one following abstraction's history as a mode of thought and the other as a marker of modernism" (Steingart 2023, 6).

The history of abstraction as a mode of thought (pp.6-10) is summarised above. Historically speaking (for Steingart), abstraction originates in a quandary regarding language's apprehension of the physical world. How do apparently universalizing words such as 'color' emerge from the world's particularity? Thought, it seems, must have to take something away from the world, to abstract something from it, in order to produce such concepts.

The pivotal figure in this sketch of a condensed history is George Berkeley. Berkeley is "a staunch idealist" (Steingart 2023, 7); yet he is also the first in Steingart's condensed history to strongly rebuke abstraction. For Berkeley, "reality consists exclusively of minds and their ideas" (Downing 2021); yet Steingart pens Berkeley as notable in the history of abstraction precisely because he refuses to allow a conceptual space of ideas that is not firmly rooted in the phenomenological world of ideas (Steingart 2023, 7).

With Berkeley, the nature of the difference between idealism and materialism is revealed clearly as a contradiction, Steingart seems to suggest. She glosses over the philosophical specificity of Berkeley's claim by lumping him in with the empiricists Locke and Hume: "What Locke, Berkeley, and Hume made clear is that the problem of abstraction as an operative mode of thought touches directly on the foundations of geometry, the signification function of language, and the relation between perception and reason" (Steingart 2023, 8). Fair enough here, perhaps, as Steingart has flagged up front that her concern is not with the nuances of philosophical positions on the matter of abstraction, so much as it is with painting a historical canvas which provides context for what she means when she says 'abstraction' in the rest of the book. The lineage presented does make clear what is at stake in abstraction (notwithstanding its suspiciously British contrivance): namely, the sticky and difficult relationship between mind and matter, subject and object, idealism and materialism. Though this is a strange progression from Aristotle to Marx in some respects, it sets the stage for the way in which abstract mathematics dances around the politicality of its own position, an argument that Steingart will make with respect to concrete actors in the 20th century.

The 'other side' of abstraction's definition, Steingart has told us, is its apparent entanglement with modernism (p.10). There doesn't seem to be any notable distinction between modernity and modernism for Steingart, and this slippage allows Steingart to settle on Fredric Jameson's account in his 2013 book A Singular Modernity as the backstop for the claim that modernism and abstraction are two sides of the same coin. I haven't yet read this book of Jameson's, and so I can't be sure of the following points. But I would hazard a guess that modernity for Jameson is structurally coupled with the problematic of the mangling of subject and object that Marx argues is a symptom ('fetish') of capital's installment as a social relation, and thus bemoans a philosophical history reaching back at least to Descartes if not further; whereas modernism refers instead to an aesthetic category that is historically specific to the 20th century (whence Jameson's infamous critique of postmodernism). Modernism is not unrelated to modernity, which is why Steingart's account that "the association of abstraction and modernism is not a continuation of the long philosophical and political tradition I just outlined, but instead represents a clean break from it" (Steingart 2023, 10) strikes me as a bit suspicious.

Steingart wants to segment (or perhaps we should say 'abstract') 20th-century abstraction from its pre-20th-century forebears by recognizing the newer phenomenon as a "severing of abstraction from the phenomenological world" , a turn away from the "etymological root" of abstraction in its reference to something of the world (Steingart 2023, 11). But isn't something of this operation already at work in the various idealisms that contest the empiricism of Hume, Locke, and Boethius in the 18th and 19th centuries? We could offer the names Descartes, Kant, and Hegel, for example, as a counter-history of Steingart's abstraction, a lineage which also culminates in an important sense in the first master of suspicion, Marx. All three of these thinkers were (it doesn't seem controversial to claim) concerned with the status of thought on its own terms, thought 'for itself' that was ideally purified of phenomenological reference. This German lineage as a counterpoint to the British one Steingart gives at the least proves that the "crisis of referentiality" (Steingart 2023, 12) of the 20th century is not the first such crisis that abstraction suffered in philosophy. The curious triad of "linguistics and scientific philosophy" (Steingart 2023, 12) that is then given to evidence abstraction's reversal of meaning in the 20th century– Ferdinand de Saussure, Albert Einstein, and Alfred North Whitehead– are brought into the historical-philosophical fray almost as if there is no precedent for the problem of the object, the subject, and the sphere of signification that necessarily traverses the two in language. It might be that 'modernism' names a distinct inflection of the constitutive problematic of modernity in the 20th century, but it is not clear to me that this inflection's distinction rests neatly on its reversal of abstraction's etymological root in the phenomenological world.

High Modernism and Axiomatics

High modernism for Steingart is above all associated with David Hilbert and the crisis in foundations of mathematics. In other words, high modernism emerges as a dialect of modernism through the proliferation of the axiomatic method (associated with Hilbert) in mathematics and other disciplines. The Bourbaki group is given as "the most extreme example of mathematical high modernism" (Steingart 2023, 17), as they emblematise a particular configuration of the relationship between mathematical theory and empirical reality. High modernist mathematics sees theory as a practice/process that is constituted as an autonomous aesthetic domain, free from the demanding contrivances of the (material) world. The high modernist attitude is fundamentally idealist in the sense that the domain of thought and theory is not borne of patterns or perceptions in the phenomenological world, but autonomously, on its own terms. Mathematical theory for the high modernist is conceived beyond or besides the world, but is then tested by its applications in the world. This phenomenology of theory in mathematical high modernism is what gives the history of axiomatics' proliferation its seemingly contradictory characters, "denoting the rationalization of social life on the one hand and the promotion of an autonomous aesthetic sphere on the other" (Steingart 2023, 18). For Steingart, game theory, communications theory, operations research, and computing (by which I assume she means the burgeoning discipline of Computer Science) all suffer this kind of speculative pragmatics that is both idealistic and utilitarian (Steingart 2023, 19), and can all thus be united in axiomatics as their common rationale and prehistory.

Grouping these histories under the rubric of high modernism is evocative also, as Steingart notes, due to James Scott's use of the term as "a particularly sweeping vision of how the benefits of technical and scientific progress might be applied" (Steingart 2023, 17). High modernism names a technocratic worldview that parades as thoroughly materialist, but which in fact is built on an unconscious idealist kernel that hazards scientific theory as autonomous (and thus unimpeachable) rather than grounded in the world and our flawed phenomenology of it. Perhaps it is this quality, i.e. the repression of an idealist basis for the claim of scientific validity, that better characterizes high modernism's break from the modernisms (and modernities) before it. Berkeley, Hume, and Burke would then be correctly identified as the prime philosophical actors in mathematical high modernism's prehistory, but for reasons to do with the way that they each repress in their own way the necessary politics of abstraction, enshrining some part of it as independent of the world and thus paving the way for its unimpeachability. (Kant would surely also feature prominently in this historical drama.)

Though the distinction between abstraction, modernism, and high modernism took me a second reading to make out clearly in Steingart's introduction, axiomatics appeared early as a clear intellectual historical object (if such a thing can be thought to exist). Axiomatics are the methods in mathematics based on the spectral objectivity of numbers posited at the turn of the twentieth century, newly realized through the questioning of their basis in physical reality.

What are numbers? Within the new regime of objectivity, neither the empiricist nor the idealist answer was acceptable. Numbers were not arrived at by counting, nor were they synthetic a priori. (Steingart 2023, 13)

Cantor goes curiously unmentioned, but we get the other usual suspects in Dedekind, Peano, Frege, and Hilbert. Steingart's argument is that the proliferation of axiomatics within mathematics and their percolation throughout other disciplines in the humanities and the sciences should be considered a high modernism in the history of abstraction.

Mathematics and Mathematization

The final two terms I wanted to disambiguate are these. While mathematics seems to usually refer to the actual disciplinary practice of professional mathematicians (and thus 'mathematical high modernism' is an appropriate description of axiomatics' preponderance at universities in postwar America), mathematization seems to refer instead to the dissemination of paradigms in mathematics to other disciplines, such as sociology, anthropology, and linguistics, as well as how mathematical technique was consciously applied in these other disciplines. To the extent that Steingart's book is a story about mathematization, then, it is in the company of recent books such as Bernard Dionysus Geoghegan's Code: From Information Theory to French Theory in that it accounts for how mathematics became the Hermes of a disciplinary pantheon conceptually divided into the humanities and the sciences. (Along with Computer Science, mathematics is– I think– one of the few undergraduate degrees in America that is offered both as a Bachelor of Arts and as a Bachelor of Science, depending on the granting institution.)

Many now are taking issue with mathematization as a sort of root cause of much that is disagreeable politically about the contemporary world. It seems important to me that we give nuance to the widespread claim that it is the mathematization by Computer Science's dissemination into other disciplines that is the root cause of our moment's political problematic. Throwing out maths and computing tout court is not an option on the table, I would argue, and thus we must reject its easy appearance as the scapegoat of bad epistemology. I intend to proceed with reading the rest of this book because it offers some such nuance in locating axiomatics as a specific culture in mathematics and tracing its social and (perhaps) political effects.

Bibliography

Downing, Lisa. 2021. “George Berkeley.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Fall 2021. Metaphysics Research Lab, Stanford University.

Steingart, Alma. 2023. Axiomatics: Mathematical Thought and High Modernism. University of Chicago Press.

Writing with Rheo

Lachlan Kermode — Mon, 09 Jun 2025 00:00:00 GMT

[idea:writing-with-rheo]postsystems2025-06-09

Writing with Rheo

You’ve no doubt already heard about Rheo, the hot new typesetting and static site engine from the Free Computing Lab (FCL). In the unlikely event that you haven’t, let me clue you in. With Rheo, you can flow a folder of Typst files into various different publishable formats, such as PDF (paged documents), EPUB (reflowable documents), and HTML (websites). It’s developed and maintained by… well, me, with contributions from colleagues at the FCL and beyond.

I’ve written before about what I need in a writing setup, and more recently about how I’ve been experimenting with Neovim and Typst towards a modern implementation of such a setup. I outline in the first post how I’ve hackily realized many aspects of my writing environment with Orgmode, Emacs, Pandoc, and LaTeX over the past few years. I don’t like LaTeX. It’s complicated, labyrinthine, and I don’t understand it. But it has always been necessary to cobble in as part of my writing system, because it is the only reasonable way to produce sophisticated layouts in PDF such as tables, maths, careful image placement, and so on.

The only reasonable way, until Typst. Typst takes the ergonomics of Markdown and combines it with the power of LaTeX, packaging it as a modern programming language with a performant and well-maintained toolchain. (You can read more about why I see Typst as the best foundation on offer for a modern hypertext writing system on the Rheo docs site.) The Typst compiler has excellent support for producing PDFs, and experimental support for producing HTML: but the syntax is designed in a way that it’s not unreasonable to demarcate sections of its syntax and find an appropriate representation in a range of different document formats.

Rheo is my first pass at how this could work. Because EPUB is essentially just HTML in a straitjacket, it wasn’t too difficult to take the HTML that Typst produces and wrap it up nicely as a valid EPUB. There are a few essentials of modern websites such as CSS that Typst isn’t scoped to provide, so Rheo bridges this gap as well. Leveraging this kind of wrapping and a little bit of thinking about how to expose packaging options to the user, Rheo is now basically functional as an open source static site engine for Typst.

In the past couple of days, I’ve ported almost all of my writing projects to Rheo. My most recent projects were already using Typst, and so for these it was as trivial as dropping them into a folder and adding a rheo.toml. A few of the websites I maintain—like this one—were largely written in Orgmode (as per my old/original writing system), so I used pandoc to get Typst from the original source and added a few post-processing steps to the conversion where pandoc wasn’t capable of capturing the syntax.

This means that I now have a writing folder on my machine which contains all of my various blogs, academic papers, my dissertation (in-progress), and job materials. Each document consists chiefly of Typst, and different subfolders contain CSS, Typst templates, and images and other assets as appropriate to the targeted formats. All of the writing in this folder is, in principle, exportable triply as an EPUB, as PDFs (combined or individual), and as webpages.

Besides being consistent, which makes it much easier to copy-and-paste writing between projects among other things, it is also all essentially editable in my text processor of choice, a customized Neovim. It is also highly interpretable by agential LLMs such as Claude Code, a product that has slowly but surely become a key part of my coding workflow. There are many justified concerns about how LLMs will impact writing in the humanities and beyond. But if you wash away the fears AI ideologues are stoking regarding the imminent automation of intelligence, a fear that I believe thoroughly misapprehends how mercurial the notion of ‘intelligence’ really is, LLMs can do amazing things in a writing workflow. In a matter of minutes, for example, I fixed some forty typos in this blog with a prompt to Claude Code as simple as ‘find typos in this blog and fix them’. Some of these typos have been around for years! Furthermore, because Typst encodes formatting in a well-documented and well-structured way, I can let an agent loose on other editorial tasks such as finding incorrect citations, inconsistently formatting book titles, and so on. All of the Typst files in my writing folder are code-like, and so I version-control them using Jujutsu and Github, which means that I can always roll back or discard any reckless changes an LLM agent might suggest.

Rheo is now working smoothly for these personal writing projects. In the coming weeks, I’m excited to attempt to port editions of a major critical journal to Rheo. This will avail the journal’s readership to new formats (webpages and EPUB in addition to the PDF that is already available), and avail the editorial team to LLM-enhanced techniques to correct spelling, formatting, and more. I expect to push a few features and fixes to Rheo through this first real case study in what it offers to the open publishing ecosystem. If you’re interested to follow along or have questions about how to use Rheo in your own projects, please join our Zulip.

Karl Marx: Capital, Volume 1 (chapter 22)

Lachlan Kermode — Fri, 16 May 2025 00:00:00 GMT

[idea:karl-marx-capital-volume-1-chapter-22]postreviews2025-05-16

Karl Marx: Capital, Volume 1 (chapter 22)

Chapter 22 of Capital Volume I– following the chapter structure in the 2024 English translation - builds on the previous chapter on simple reproduction;, which had laid out the impossible position of "economic bondage" (Marx 2024, 532) of the wage laborer in relation to capital. Chapter 22 shifts from outlining the standpoint of the wage laborer to a closer look at the standpoint of the capitalist, detailing how capital works through the reinvestment of surplus value.

When the surplus value produced through the structure of the wage relation is reinvested, it amounts to a process of accumulation: "accumulation is the capitalist process of reproduction on an ever-expanding scale" (Marx 2024, 534). Accumulation is a phenomenon that is produced under the aegis of a free and fair exchange, but which in fact rests in the exploitation of laborers. Though it is the laborer who produces surplus value, it is the capitalist to whom it is owed. This is why Marx calls the capitalist a false "legal representative" (Marx 2024, 535) of workers, as it is through an apparently legitimate process (we might think here of 'democracy') that a figure who represents interests antithetical to the workers' own comes to reside in a seat of power. Unpaid labor, conjured through the reduction of labor to labor-time in the form of a wage ("the invisible magic of the production process" (Marx 2024, 535)), is the common coin of this phony demagoguery.

Accumulation is the name for the reproduction of capital for its own sake. The capitalist is not, according to Marx, a machiavellian villain who selfishly expropriates value from the laborers he employs. (This is one of the core distinctions between expropriation and exploitation for Marx: the latter does not imply strictly overt or conscious violence on the part of the dominating party.) The capitalist is rather a role that is subject to the abstract domination of capital's value forms, just like the wage laborer, though that subjection has a different orientation. The wage laborer, as we saw in chapter 21, is compelled to want more commodities, even at the expense of her own reproduction, an orientation that can be denoted C-M-C'. The capitalist, on the other hand, is compelled to want more money, even at the expense of his own reproduction, an orientation that famously can be denoted M-C-M'. It is these two roles working in tandem that produces accumulation, the movement of value specific to capital, an infinite cycle of M-C-M'-C'-M'. (The formal genesis of this cycle, i.e. whether it is M or C that kicks accumulation off, is the subject of the final chapter in Volume I, The So-Called Original Accumulation.)

The search for surplus (') for its own sake is the driving motor of capital's process, the characterizing feature of its implant in human subjectivity, whether its installation produces a capitalist or consumerist 'affect':

Whether the new capital takes the form of money or the things that make up the means of production and subsistence, its substance is the product of a process that extracts someone else's unpaid labor. Capital has produced capital. (Marx 2024, 536)

Marx's use of the word 'substance' here and elsewhere can be understood to mean something similar to what I propose as a 'characterizing feature' above. It does not refer to a concrete and material thing that can be touched, smelled, or felt that underlies those airy, ideological phenomena that are palpably more conceptual. To render 'substance' in this way is to read Marx against his own method, to assert that there is a 'base' that can be consciously unearthed as a 'thing' that determines the 'superstructure'. Substance rather refers to neither thing nor thought, but the dialectical space where they meet– what Heidegger will later call das Ding. The substance of value in capital, then, needs to be explained by way of the process of its (conceptual and material) movement, from commodity to money to surplus, such that we can see how (a certain conception of) labor is the 'ground' of the abstract domination at work in capital.

The network of notions of commodity, labor, value, and so on is what reveals the notion of bourgeois ownership as one that is turned inside-out. It parades as a transparent principle of 'each to his own', but in fact (in 'substance') rests on the underhanded dispossession of workers of their labor. The law of ownership is, in other words, a law that harbors a contradiction within itself. Though an "exchange of equivalents" is "the original operation", the outcome is constitutively unequal (Marx 2024, 537):

Owning yesterday's unpaid labor now appears as the sole precondition for appropriating today's unpaid living labor on an ever-larger scale. (Marx 2024, 536)

A key point in the first section of this chapter is that the capitalist is not, by definition, a miser. Accumulation can only take place if "yesterday's unpaid labor" is put back into circulation, rather than hoarded by miserly factory owners. This is not to say that those playing the part of the capitalist in accumulation's equation do not spend some of the surplus on themselves, buying commodities that they do not need in pursuit of an ever-elusive total satisfaction. But this drive is not what characterizes him as a capitalist:

insofar as the capitalist is personified capital, his driving motivation is to have more and more exchange-value, not to acquire use-value and experience pleasure. A fanatic when it comes to the valorization of value, he ruthlessly forces human beings to produce for production's sake…. Thus, insofar as the capitalist's behavior is merely a function of the capital that has taken on consciousness and a will in him, he sees his own private consumption as an act of theft committed against the accumulation of his capital. (Marx 2024, 542)

The affect driving the capitalist is a renunciation of private consumption, an obscure and alien selflessness in service of capital's reproduction, not his own.¹ (One cannot help but think here of the numerous reports of investment bankers committing suicide upon hearing about the financial crash in 2008. Agents of capital to the end, life is no longer worth living where accumulation has stalled.) Thus is capital a parasite on modern subjectivity, which since at least the Enlightenment has prided itself on the ability to reason first and foremost for oneself. In capital, this vision of clear-headed reason is compromised. Reason itself is unsound, for conscious abstraction in its own right seems to suffer domination at the hands of the movement of value.

Bibliography

Johnston, Adrian. 2024. Infinite Greed: The Inhuman Selfishness of Capital. New York: Columbia University Press.

Marx, Karl. 2024. Capital: Critique of Political Economy. Edited by Paul North. Translated by Paul Reitter. Princeton: Princeton University Press.

Footnotes

^ Adrian Johnston makes this point persuasively in the preface to his 2024 book Infinite Greed: The Inhuman Selfishness of Capital (Johnston 2024).

Is psychoanalysis a Science?

Lachlan Kermode — Thu, 10 Apr 2025 00:00:00 GMT

[idea:is-psychoanalysis-a-science]postgeneralreviews2025-04-10

Is psychoanalysis a Science?

Psychoanalysis was once described as a ‘calling card’ that was to be left at the door of Science. The idea here, I think, was that psychoanalysis should not become (as Freud envisioned it in his first explorations of it) a ‘science’ in the sense of having an answer for everything. If psychoanalysis were to become a theory that accommodated every possible thing, every possible action, to foreclose all possibility of surprise within its framework, then it would be no better than a worldview (or Weltanschauung, in German). Incidentally, Freud dismissed Marxism as too much of just such a worldview, and therefore according to Freud Marx’s account of the world is not something to which psychoanalysis should aspire.

Instead, psychoanalysis– as a truer kind of science than the Science that Freud criticized–the institutionalized Science of physics, chemistry, applied mathematics, and perhaps biology– should be a technique that Science turns to in an idle moment, when one of its formulae fails to produce an adequate account of some happening in the lab; when the scientist is racking her brains, late at work one night, trying to reduce some recalcitrant experiment to an eloquent result. Psychoanalysis, the calling card analogy suggests, is the method that one might turn to in this confounded moment when the structures of organized knowledge seem exhausted, misguided, or just plain wrong. When something (in society) no longer works, and one needs to find the will to try to make something work again, even as it seems that the world has firm ideas otherwise.

The psychoanalytic clinic (the calling card left at each of our aporetic doors) is an institution with a serious kind of humility. In its best form, it doesn’t offer answers, only spaces for its subjects to ask more fundamental questions. As Mladen Dolar puts it, the analyst in the clinic

utters at the most an "Oh!" here and there… he makes himself an automaton in order to give rise to the dimension of the Other, the real interlocutor of the patient's "monologue," and also in order to produce that strange kind of love, perhaps love in its strictest and purest sense, which is transference love. (Dolar 1991, 9)

The proper analyst doesn’t suppose to cure the subject of her ills and maladies, her maladaptations to a broken society, but rather first and foremost seeks to open a space where the subject can start again, where association will no longer be strictly limited by habit, pathology, or unwanted automatisms. Michael Pollan has described the experience of ingesting psychedelic substances as a fresh snowfall on the mountain of one’s mind, where the grooves that have been worn to the earth can be rediscovered or altered– and perhaps this is a kind of description of the successful analytic session, too. Afterwards, the subject can choose the way she thinks, again. When a certain logic of understanding the world has revealed itself as uncertain, incomplete, or unreliable, the clinic should be a calling card. Thought can start again, the suggestion seems to promise, even if we are scientifically sure that it never really can.

The question for analytic bureaucracy is this: if uncertainty is a given, is there anywhere the analysts should stake a claim to certainty? The psychoanalytic philosophical tradition can be read as a kind of reckoning with this epistemological impasse. How does one build an institution when the surety of its foundations are in question? If this question forms the ground of psychoanalysis as a philosophy and as an aspirationally ‘true’ science (instead of yet another Weltanschauung), then perhaps it is to the clinic that we should turn for an institution that will practice restraint when it comes to the inevitable desire to totalize, to be something for everyone, rather than simply everything for someone.

Bibliography

Dolar, Mladen. 1991. “"I Shall Be with You on Your Wedding-Night": Lacan and the Uncanny.” October 58: 5–23.

Karl Marx: Capital, Volume 1 (chapter 21)

Lachlan Kermode — Fri, 07 Mar 2025 00:00:00 GMT

[idea:karl-marx-capital-volume-1-chapter-21]postreviews2025-03-07

Karl Marx: Capital, Volume 1 (chapter 21)

Up first today, some notes from chapter 21 of Marx’s Capital, Volume I, his chapter on Simple Reproduction. I returned to this text (in its new translation) at a writing retreat in Connecticut over the weekend. After sharing a work-in-progress regarding the outline of a theory of a subject in Marx, it was sagely pointed out to me that, in order to really convince a reader that such an outline exists, I should really show its substance in Marx's own writing. Up until this point, my argument has been constructed around instructive 'secondary sources' such as Moishe Postone's Time, Labor and Social Domination and Alfred Sohn-Rethel's Intellectual and Manual Labour: A Critique of Epistemology, both of which are book-length interpretation of Capital by lifelong scholars of Marx. These two texts will remain key interlocutors in my argument's construction, but I cannot only channel Marx's Geist solely through second-hand accounts of him.

My hesitancy to interpret Marx's text directly comes out of an abiding bias absorbed from the academic atmosphere which suggests that it is 'old news' to retrace interpretative grooves that have already been clearly marked out. I have come to disagree quite strongly with this opinion, as it misses the real value of textual interpretation, which is not primarily to produce new knowledge from undiscovered sources, but to work through language and thought in and against the temper of a new time. But the 'innovation exhortation' nonetheless haunts me, aspiring as I am in part to American academic success, as one is expected in this climate to produce something new and newly useful. It is much easier to argue for the newness in an interpretation of an object not yet interpreted than it is to justify yet another interpretation of well-trodden territory like Marx's Capital. In the latter case, one also has to be sure that one is not offering as 'new' an interpretation that has in fact already been offered in the literature. In Marx's case, there are some two hundred years of interpretation in almost every language imaginable. How can one be sure that yet another interpretation is 'new'?

One can't, in short. It is impossible to read all the relevant literature. It is impossible even to read most or a significant portion of it in the case of Marx-related studies. This is one of the ways in which the vocabulary of scientific discovery falls short in textual interpretation. The value of interpretation is not always surfacing something hitherto unknown to an imaginary readership that has already read everything hitherto published. The fantasy of such an academic production is not only an impossible task, but it also has a fantasmatic audience. No-one is perfectly literate in this exhaustively imagined way! A better objective, I am coming to realize, is to synthesize something for a significant portion of readers that is not yet readily visible. Thus I am turning my attention to re-reading Marx's Capital, as it seems to me that there is much that has been 'lost' to the zeitgeist regarding this text's critical value, even as it is nominally the most common coin in contemporary critical-theoretical projects.

21: Simple Reproduction

(In the older translations, this is chapter 23.)

David Harvey notes that this chapter zooms out from what has preceded it in volume I in terms of analytic perspective:

The analysis so far has been concerned solely with the production of surplus-value as a one-shot event. But things look rather different when examined as a continuous process ongoing over time. (Harvey 2018)

At this stage in Capital, a reader has been made familiar with key concepts such as commodity $𝐶$ , money $𝑀$ , value, labour-time and so on, and the movement of value has been atomistically conceptualised in slices such as $𝑀 - 𝐶$ (a purchase), $𝐶 - 𝑀$ (a sale), $𝑀 - 𝐶 - 𝑀$ (the capitalist equation) and $𝐶 - 𝑀 - 𝐶$ (the consumerist equation). These are all expressions of moments in the entangled processes of the production and circulation of value as analytically captured by Marx. These forms of appearance are critical terms; but they are also "phenomena that conceal the inner working of [capital accumulation's] mechanism" (Marx 2024, 520).

Along with the conditions of production (and circulation) come conditions of re-production. Reproduction is the sphere of activities that are not consciously counted in capital's calculus and as such cannot appear to matter in the context of its logic. Marx's argument in Capital is that capital is a synthetic totality, by which I mean to say that though it is self-consciously totalizing– that is, it appears to itself as fully accountable– in reality it represses a range of moments that are actually the unconscious foundation of its character.

The big elephant in the room here is labour. Capital considers labor as a relatively unremarkable input that contributes to its valorization process in production and circulation. The capitalist starts with some money; he uses it to purchase variable capital (labor-time) and fixed capital (machinery) to kick off the production of some commodities; and when he sells those commodities he skims the surplus value off the top that has naturally been generated by his bringing those pieces of capital together to work in cooperation. Marx's project thus far has been to logically reconstruct and conceptualize the elements of this production process so as to show that there is something problematic about this 'natural' interpretation of the process of accumulation, and that something is the way in which labour gets measured as a contribution to the surplus value returned.

Thus Marx locates the logical beginning of capital's valorization:

The purchase of labor-power for a certain amount of time initiates the production process. The initial moment recurs when the time is up, and a production period of a certain length has come to an end, whether a week, a month, or longer. But the worker isn't paid until after his labor-power has produced its effect, realizing in commodities both its own value and surplus-value…. The labor that the worker performs next week or next month will thus be paid for with the labor he performed the previous week or the previous month. (Marx 2024, 522)

The worker necessarily advances his labor-time as a commodity to the capitalist due to the temporal structure of the wage. This advancement is the logical beginning of the production process, Marx claims, as it is the moment in which capital's particular structure of valorization becomes possible. It is the parceling out of discrete periods in continuous time effected by the structure of a wage that makes valorization possible, as this structure allows some portion of labor-time to subconsciously slip between the contract's positions. The worker appears to be paid fairly for his time at some set rate an hour; yet in the very motion of this apparently free and fair conscience, some labor-time nevertheless slips away from him to the capitalist, unpaid. Thus the unconscious advance of the worker's labor-time to the capitalist that is effected by the structure of the wage is the gap in which surplus value is made to appear.

This apparently fair but structurally imbalanced relationship between the worker and the capitalist is the fundamental false consciousness of capital's logic. It is what produces the gap into which capitalist ideology can pour its misconception of equality into the consciousness of both worker and capitalist, even as the former's condition becomes more and more immiserating, and the latter's more and more lavish. This structural imbalance allows capital to take its place in consciousness and society as 'alien power' that nonetheless also feels natural and fair. Take the following (very difficult) passage:

It is the worker himself, then, who constantly produces objective wealth as capital, an alien power that exploits and rules over him, while the capitalist just as constantly produces labor-power as a subjective, abstract source of wealth that has been separated from its own means of objectification and realization and exists only in the worker's body– in short, the capitalist just as constantly keeps producing the worker as a wage laborer. This constant reproduction or perpetuation of the worker is the sine qua non of capitalist production. (Marx 2024, 525)

Despite appearances, it is the workers themselves who are capital's "most essential means of production" (Marx 2024, 527). The capitalist class cannot produce surplus value through an apparently fair commodity production and circulation process without their mobilization of workers through the structure of a wage. Wages are the 'invisible chains' that reinscribe class distinction in capitalist society, replacing the conspicuous chains that patrolled that border in pre-capitalist society:

The Roman slave was fettered with chains. Invisible ties bind the wage laborer to his owner: he merely seems to be independent. The constant turnover among the worker's individual wage masters and the fictio juris of his contract keep this semblance in place. (Marx 2024, 528)

The wage contract is a fictio juris, a false premise taken as true by capital logic to reach its desired outcome: the accumulation of surplus value. Capital has reinterpreted the master-slave dialectic and transsubstantiated it terms, leading to a situation in which both master (now capitalist) and slave (now worker) have developed a self-consciousness of freedom, yet both remain unconsciously in bondage to the desires of an alien power beyond their control. The capitalist is driven to accumulate, as he occupies the position of $𝑀 - 𝐶 - 𝑀^{'}$ in the movement of value, driven to use his money to produce more than he originally put in, the elusive moving target $'$ that can never be satisfactorily achieved. Even when he has accumulated more money than he knows what to do with, more than he needs to produce a fine quality of life, he is driven towards more money nonetheless. The worker on the other hand is driven differently, but towards a similarly unachievable infinity in the position of $𝐶 - 𝑀 - 𝐶^{'}$ , driven to commoditize herself as labor-power in order to acquire more and more commodities, even at the expense of her (quality of) life.

Thus is humanity made a machinery in service of capital's reproduction rather than its own, caught in:

a self-perpetuating trap, continuously thrusting the one person back into the commodity market as the seller of his labor-power while continuously transforming his product into the other person's means of purchasing…. [I]t also produces and reproduces the capital relation, with the capitalist on one side and the wage laborer on the other. (Marx 2024, 532)

The seemingly benign concept of a commodity (unfolded in the opening chapter Capital) has somehow grown into a monstrous misapprehension of subjectivity, entangled in an external, objective process of accumulation for its own sake. Hegel's subject [Subjekt] has become a substance [Substanz], but not in a way where it can find its way back again to the first term. Subjectivity is instead ensnared as a substance that kicks off and sustains the inorganic motion of the accumulation of value for its own sake, reducing humanity to a shadow of itself in the unthinking conceptualization of labor as nothing more than a replenishable repository of labor-power. Thus is labor the sine qua non of capital.

Harvey, David. 2018. A Companion To Marx’s Capital:The Complete Edition. Verso Books.

Marx, Karl. 2024. Capital: Critique of Political Economy. Edited by Paul North. Translated by Paul Reitter. Princeton: Princeton University Press.

Erich Hörl: Sacred Channels: The Archaic Illusion of Communication (chapters 1 and 2)

Lachlan Kermode — Mon, 10 Feb 2025 00:00:00 GMT

[idea:erich-horl-sacred-channels]postreviews2025-02-10

Erich Hörl: Sacred Channels: The Archaic Illusion of Communication (chapters 1 and 2)

Last week I read the first two chapters of a book recommended to me quite some time ago by a friend: Sacred Channels: The Archaic Illusion of Communication by Erich Hörl. The book was published in English in 2018, but was defended as a dissertation in 2003, and published in German in 2005. Friedrich Kittler was the dissertation's second reader, and we learn from the introduction that the manuscript was also read by Bernhard Siegert and André Gorz, among others. Some pretty heavy media-theoretical hitters in the German-speaking context!

I am opening this book at last to begin working my way through the history and philosophy of mathematics that is relevant to my dissertation. The lineaments of this relevance are, broadly speaking, those that trace the emergence of set theory around the turn of the 20th century, and its culmination in the crisis of foundations orchestrated by the young Kurt Gödel when he published Über formal unentscheidbare Sätze der Principia Mathematica und verwandter Systeme - 'On Formally Undecidable Propositions of Principia Mathematica and Related Systems'. This curious and relatively obscure turn of events in the history of mathematics has been heralded by some, the author of the text under consideration in this b/log included, as the climactic moments in the most fundamental epistemological transition in philosophy and thought at large of the last few centuries; the coming of modernity.

Thinking and so being

What does set theory and formalist mathematics have to do with philosophical modernity? Hörl's claim is that, from as early as the 1830s, a "purely symbolic thinking clearly began to take shape in the sciences" (Hörl 2018, 33) Symbolic thinking is marked in distinction to its precursor, intuitive thinking, which is construed as the reigning and preceding paradigm in mathematical thought writ large.

For Hörl, the "great transformation from the intuitive to the symbolic" (Hörl 2018, 47) is a shift in the relationship between Thinking and Being. I capitalise these two terms here to indicate that they designate presumed universal notions, rather than simply historically contingent activities of thinking and being.

Thinking (with a capital T) is the realm of appearances, the full breadth of what is sensible to the subject. The subject, here, is a human one in the sense that human is the word given to all those entities capable of social life by dint of their capability for language. Language, here, too, is not 'English' or 'Spanish' exclusively, but a more holistic category that gets at the curious phenomenon we often now call communication. Language is not language in the strict linguistic sense of having a grammar, nouns, and verbs, but is rather something more curious and not so easily locatable as a set of articulate rules. Language is the name for the many syntaxes of sociality, the protocol(s) through which subjects cooperate, live together, and are insufficient to be themselves alone. Thinking is the domain of what is cognisable and apparent in and as language, the full sensibility of the world as it can be Thought.

Being, on the other hand, is what the world actually is, what actually exists. It is the tree that falls in the forest without anyone around to hear or think it. It is more durable and persistent than what can be Thought, to some minds, as what exists in the world is not necessarily coincident with what appears to those of us subjects that are thinking within it.

If the distinction between Thinking and Being seems finicky, you are not alone in this complaint. Whether or how Thinking and Being overlap is a core question in philosophy. Ontology is the investigation of what exists, to delimiting the domain of Being. Epistemology is the complementary investigation of what can appear transparently to subjects in the world, of the borders of the domain of Thinking. The articulation of the latter usually (always?) implies the former. Epistemology can really only be defined by nominating whether or not there is anything that exceeds it. The exceptions in Thinking, things which exist but which cannot be Thought, are the stuff of ontology.

Intuition is the key to everything

For Hörl, the ascendance of symbolic reasoning in mathematics in the early 19th century is a feature of a shift in the tectonics of Thinking and Being. Hörl is a media theorist and not an historian, and he does not clearly delimit the spatial parameters of what he sees as the age of symbolic reason. (Is it only or primarily in Europe? Is it as intense in France as it is in Germany? Does it suffuse the intellectual scene in Kenya or Kyoto in an equivalent or noticeable way?) But the general idea is that, before this age, the cutting edge of ontological proposition was secured through an appeal to intuition.

Intuition is a term laden with Kantian suggestion, as it played a complicated role in Kant's philosophical system. The sum of it as relevant to Hörl's argument is as follows. Kant was adamant that there were some aspects of Thinking that we could take to be universal, that is, that we could take to exist not just in particular instantiations, but as general truths for all subjects. The echelon of Thinking that Kant reckons universal is specifically that of Reason, and it is the task of philosophy to shore up the boundaries of Reason so that its factions do not contradict themselves; "to expose the illusions of a reason that forgets its limits, and by sufficiently clarifying our concepts to recall it from its presumptuous speculative pursuits to modest but thorough self-knowledge" (Ewald and Ewald 1999, 1:147). Whatever the specific cartography of this universal Reason's shores, Kant was sure that Thinking could eventually map them. There are some kinds of Thinking, in other words, of which we can be sure. The surety of Reason's forms appear to us through intuition, a capability (of sorts) that allows subjects to make out the shape of structures that undergird the many specific sensualities that appear in the world. Intuition, in other words, is what ensures that the hypothetical play of structures and patterns in logic (Thinking) have something to do with what is actually out there in the world (Being). For Kant there is a necessary overlap between Thinking and Being, albeit probably a 'modest' one, about which 'thorough' thinking is fundamentally able to reason.

As Hörl is working in a media-theoretical paradigm that follows Kittler, he periodises and defines this (Kantian) era in the relationship between Thinking and Being as an "optical-medial paradigm of consciousness" (Hörl 2018, 51). Structures and associations appear in the mind of subject as apparitions of sensual particularities in the world through the lens of intuition, much like the shadows playing on the wall of Plato's cave. Intuition is what secures that there is some reality that can be disambiguated from fantasy in the observance of such shadows. Good, philosophical thinking in Kant's cave is the diligent picking-out of objective patterns, forming a structure in the mind that extracts only what is objectively there in the world, and as such will not be contradicted by any particular play of shadows.

Leibniz leading the blind

The first chapter of Hörl's book wrestles with how one should think about the crisis of intuition and its giving way to a notion of the symbolic that was not secured by the same kind of intuition, or at least not in the same way. He evocatively refers to this new paradigm as blind thinking. The media-theoretical avoidance of any requirement that time be stratified as linear, and thus a period given a specific date to inaugurate it, is clear from the chapter's title; 'Blind thinking around 1900' (my emphasis). Symbolic thinking is blind because Leibniz labeled it as such (cogitatio caece) in his aspiration to designate structures of association that exist in the world without any contingency in Thought:

Leibniz had been working on an autonomous symbolic freed from the fetters of the representation of objects and had philosophically conceptualized the formal use of symbols in algorithmic mathematics in terms of a purely symbolic knowledge…. As the "pioneer of the idea of calculus"– that is of a purely formal manipulation of signs free of all interpretation– he kept to the blindness of using symbols despite all the evidence of immediate, instantaneous viewing…. Because it proceeded nonintuitively and was not founded on the evidence of an insight, Leibniz called symbolic knowledge "blind" (caeca). (Hörl 2018, 58–60)

Leibniz's project of calculus is nominated as a decisive break from Kant's '20/20' vision of philosophical reason. In the symbolic paradigm of Thinking and Being, there was no essential security that structures of Thought had any bearing on entities in the world. One can search for consistency in symbols in thought and in theory (this enterprise has come to be known as pure mathematics), but there was no guarantee that such Thinking had anything to do with Being. Thinking after Leibniz, Hörl contends, is fundamentally detached from Being, as there is no longer much Kantian faith in an intuition that binds them together.

It is interesting to me that Leibniz's calculus becomes an "autonomous symbolic", as I have written elsewhere about the status of capital as an autonomous subject in Marx. How is it that the symbolic is 'autonomous'? The domain of the symbolic is no longer bound to the particularities and peculiarities of subjectivity; it is no longer a domain that exists only 'interiorly' as shadows in the mind. The symbolic now exists for itself, autonomously. We as subjects are simply capable of observing it. No longer shackled as shadows in the cavernous minds of stuck subjects, the symbolic now shares the status of real objects out there in the world. It has been freed, made autonomous, just as Socrates had always hoped might be possible for subjects were we all to become philosophers (such that we might no longer stare only at shadows on the wall).

Kittler in the kitchen

In reading these first few chapters, it seems that Hörl is making use of the Lacanian terms symbolic, imaginary, and the real in a Kittlerian manner. What does it mean to use these terms in a Kittlerian manner, exactly? I admit that I need to return to Kittler's work in order to properly reason through this question, in particular his 1986 book Gramophone, Film, Typewriter (translated to English in 1999), as I last looked at his work as an undergraduate (with very different scholarly eyes). This book is a definitional text in German-American media theory, as it makes out three epistemological paradigms in the 20th century– that of the gramophone, of film, and of the typewriter– arguing that each medium represents the conditions of Thinking in its respective historical specificity. Kittler does this using the Lacanian triumvirate, as is summarized by this excerpt from this (curious) wiki:

The gramophone appeals to the real, functions as an analog representation of the voice, and maintains the voice of the dead (the scream of the dying over radio). Film functions as the imaginary, capturing not light waves but the effect of light waves, and reproducing our illusory doppelganger before our eyes at a speed too fast to capture difference. The typewriter corresponds to the symbolic, a symbolic order now stripped of its inherent relationship to male thought as thousands of women enter into the typing fields.

So:

Gramophone	the Real
Film	the Imaginary
Typewriter	the Symbolic

I don't know of work that directly criticizes Kittler for this schematic, though I imagine that there must be some, as (by my very nascent understanding) it seems strange to figure each of Lacan's terms as more active in a particular historical period, at the very least. There is something suspiciously Foucauldian about the attempt, which is not surprising, given that I believe that Kittler explicitly invokes Foucault (and Derrida) in Gramophone as intellectual inspiration. It figures the terms as a set of equivalent intensities in relation to each other, rather than as aspects of the same topological structure. To be sure, I am no expert in either Lacan or Foucault, and so perhaps there is a way hold both conceptions effectively in mind at the same time. This problem inspires me to read Kittler's Gramophone in the coming days, a trial that I may report on here in one form or another.

The reason I will return to Kittler is that I find Hörl's mobilisation of the Lacanian triumvirate compelling; but I am as yet uncertain as to whether it is helpful to frame the history of Thought in such belletristic terms. For Hörl, the Symbolic (Lacanian or otherwise) is evidently associated with the formal-mathematical symbolic of Boole, of imaginary numbers (both discussed in chapter 2), and of the symbolic 'turn' in the history of mathematics. The Imaginary, in contrast, is affiliated with the intuitionist mapping of mathematical structure to the 'real world', as Hörl argues was the case with geometry before its non-Euclidean awakenings (the invasion of the sanctity of Kantian intuition as mathematics' epistemological foundation by the formal modernity of Leibnizian and Boolean symbolic operationalism). The Symbolic, on this view, is constituted by the paradigm of non-contradiction, whereas the Imaginary is an apparent fortress of philosophical reality constructed on the sense that there is some fundamental channel between what one thinks (or can imagine) and what is actually out there in the world, a channel called intuition.

The ascendancy of the symbolic from 1830 through 1850, and whose monarchy of names such as Russell and Hilbert was overthrown by the popular demagogue of reason that goes by Gödel, seems to be Hörl's Kittlerian periodizing of the slow takeover of modernity. Perhaps Hörl is simply tracing the Kittlerian thesis in Gramophone by way of different historical roots?

Bibliography

Ewald, William Bragg, and William Ewald, eds. 1999. From Kant to Hilbert Volume 1: A Source Book in the Foundations of Mathematics. Vol. 1. OUP Oxford.

Hörl, Erich. 2018. “Sacred Channels: The Archaic Illusion of Communication.” Sacred Channels, 1–342.

Terry Eagleton: Why Marx Was Right

Lachlan Kermode — Wed, 05 Feb 2025 00:00:00 GMT

[idea:terry-eagleton-why-marx-was-right]postreviews2025-02-05

Terry Eagleton: Why Marx Was Right

Over the holiday period (which I count as the entirety of January, as my wedding took place on the 11th January, and I have not seriously returned to work until very recently) I read Why Marx was Right by Terry Eagleton. The book was given to me by my father– over in Italy from New Zealand for the aforementioned wedding– as he had gone out and bought it after I had sent my dissertation prospectus to the family group chat, a document that makes clear that Marx's thought is a very major focus in my current work.

I hadn't heard of the book before, and was delighted to discover what I think is a very accessible introduction to many of the key aspects of Marx's thought that drew and draw me to it. Indeed, I wish that I had known about the book when putting together the syllabus for my class on Marx last semester (discussed in the first section here;), as it guides the reader towards the salient aspects of why Marx remains relevant without assuming that they have already laboured through Capital. Now that I have laboured through it, it has become the gateway drug I propose when accosted with the by-now-seemingly-age-old question, dredged up equally often in non-academic and academic discussions; 'but why study Marx'?

Each chapter in Eagleton's delectable little book takes a misconception about Marx head on, explaining why it is unfounded. The wikipedia page does an excellent job of outlining the lines of argumentation in summative depth, so I'll just bullet them here:

Marx's critique has no bearing on 21st century society, as it deals with 19th century concerns such as factories and food riots.
Marxism in theory is for faux-left academics, as in practice it always becomes ``a monstrously oppressive state".
Marx's notion of history was overly deterministic, positing a 'necessary' transition to socialism/communism that never came about.
Humans are inherently selfish, and so the dream of communism is hopelessly utopian and fundamentally unrealistic.
Marx's thought was too invested in economic reduction, and as such is fundamentally less rich as a tradition of thought than more contemporary, more pluralistic alternatives.
Marx's 'materialism' drains humanity of everything spiritual, crudely envisioning the world as nothing but what there is, and as such always ends up in undesirable states of society when applied, as the history of the 20th century proves.
Marx's conception of class is limited to his time, and can't help us to understand the diffuse hierarchies of social standing in today's world.
Marxist revolution is inherently violent.
Marxist revolution can only end in a Foucauldian all-powerful state, and thus the theory is inseparably married to unfreedom at some fundamental level.
The political Left today is not well-explained or supported by Marx or Marxism, and thus we should prefer other theories.

Needless to say, as the sanguinely simple title of Eagleton's book implies, all of these allegations are unfounded at best; and more likely they are just plain wrong. It is, in fact, astonishing how relevant Marx's thought remains today, given how much political and historical distance there is between us and him. Indeed, we live in wake of an executive order given by a popularly elected president of the U.S. to root out 'Marxist equity'; and I will continue to argue that conceiving clearly of what such equity would look like if it were to appear is some of the most important scholarly work one can do. If philosophy and critical thinking are to remain a valued part of organized society, we should carve out space and time for them at universities, even as the capitalist assault on the enduring value of critical thought intensifies (t)here.

Eagleton's book is not just an accessible introduction, however. Its footnotes are concise yet sharp, pointing the more interested reader towards secondary literature and sections of Marx's ouevre that deserve closer attention. The influence of Fourier on Marx's political philosophy of utopia, for example, is a connection that I intend to follow up on in the coming months, having as it may interesting ramifications in the course of my study on what mathematical contours exist in Capital and Marx's other works. Balibar's book Philosophy of Marx– which I had bulleted on my lists;, but have still not yet read– is also mentioned for Balibar's characterization of Marx as an antiphilosopher; a nomination that accords with my continually reinforced conception of Marx as one of the great masters of suspicion.

There are also some comments in the book that help me to carve out space to concieve of Marx's system as a philosophy (or anti-philosophy, perhaps) of the subject. (No doubt this is the terrain that Balibar will cover in that book, when I finally come around to reading it.) In the chapter on why Marx is not a crude materialist (chapter 6), for example, Eagleton cleverly shows how it is really capitalist 'philosophy' that is culpable of painting everything simply 'as it is', leaving no space for an imagination otherwise. (This being the source of Fred Jameson's / Mark Fisher's quip that 'it is easier to imagine the end of the world than it is the end of capitalism') The unfreedom of imagination otherwise in capitalism, Eagleton argues, is a feature and not a bug for that system's zombified reproduction:

The prejudice that thought is independent of reality is itself shaped by social reality. (Eagleton 2011, 135)

It is not too far to go from this claim to the insight that consumed the working life of Alfred Sohn-Rethel, for which I quote from my prospectus:

Though Sohn-Rethel is best known for coining the phrase 'real abstraction' (principally popularized by way of its mention in Adorno's Negative Dialectics (Adorno 2014)), the radical claim on which his sole philosophical monograph, Intellectual and Manual Labor, rests is this: "Exchange abstraction is not thought, but it has the form of thought" (Sohn-Rethel 1978, 99). The idea that there is a formal relation between the transcendental subject ("thought") and the political economy, i.e. the structure of a set of social relations that cannot necessarily be simply reduced to the sum of conscious operations of its constituents, situates Marx in a mathematical-philosophical lineage with which he is not always directly associated.

Marx's work, as Eagleton recognizes, is not just a sociological, historically-specific treatise on the unjust structure of nascent capitalism in the 19th century; though it is certainly that. It is also ``a theory of how historical animals function" (Eagleton 2011, 158), which is to say, a philosophy of the (human) subject, a serious reckoning with whether or not there is anything philosophical (trans-historical) that can be said about society, or whether we are doomed to never be justified in abstracting from a specific time and a specific place, wherein not only is the map not the territory, but the map is never anything useful or justified at all.

Bibliography

Adorno, Theodor W. 2014. Lectures on Negative Dialectics: Fragments of a Lecture Course 1965/1966. John Wiley & Sons.

Eagleton, Terry. 2011. Why Marx Was Right. Yale University Press.

Sohn-Rethel, Alfred. 1978. Intellectual and Manual Labour: A Critique of Epistemology. Humanities Press.

Using NixOS

Lachlan Kermode — Mon, 27 Jan 2025 00:00:00 GMT

[idea:using-nixos]postsystems2025-01-27

Using NixOS

My NixOS configuration

For about a year now, the operating system that I use on my daily driver has been NixOS. I moved to the distribution from Arch Linux, as I was sold on the idea of having all my dependencies declaratively specified, and thus no longer having to mess with stateful package management (apt-get update, pacman -Syu, etc). I was one or two years into grad school, and off the back of working professionally with Terraform as declarative infrastructure for cloud systems. How great would it be to have something similar to provision all the software on a desktop machine, I reasoned. The rumours that that Nix and NixOS were useful in theory but prohibitively arcane in practice didn't dissuade me. If anything, I thought, I would be able to learn more about Linux through incorporating the ecosystem in my day-to-day life as a grad student¹.

These visions of perfect declarative management haven't completely come to fruition, but they also haven't been completely blown out of the water. It has taken me a year or so of idly hacking away at my configuration when I have wanted to install something new to finally begin to feel as though I have a handle on the language and the ecosystem. I am still not totally sure what the best resources for learning more are. Mileage varies in the official documentation, and there seem to be both a wide range of use cases of Nix, and a wide range of parenthetical softwares.

This range of parenthetical software made NixOS especially difficult for me to adopt, as I was consistently unsure as to whether the complexity that I was trying to scale was inherent to Nix, or whether it was embedded in additional tooling that I had unknowingly chosen to adopt. Let me count (a few of) the ways that I feel my NixOS journey was made difficult.

NixOS

The first fundamental confusion for me was the boundary between Nix, the language and package management ecosystem, and NixOS, the Linux distribution that is configured by way of Nix. It was the idea of NixOS that drew me to the ecosystem– as I wanted declarative management for my operating system software specifically– but I realise now that I probably would have been better to have started off by understanding pure Nix well and good, and only moving to NixOS once I felt competent there. This is because NixOS works by way of Nix (of course), but there are also peculiarities of NixOS that one should not confuse with the peculiarities of Nix. What arguments are passed down to which files in a modularized NixOS configuration become a lot easier to understand if one has a clear mental model of how NixOS builds upon Nix.

Note to past self: build a good mental model of pure Nix before hacking together your first NixOS configuration.

Home Manager

As I built my NixOS configuration by referencing an active installation of Arch Linux, I made the relatively unconscious decision early on that I should install softwares at a user level, rather than at system level. Relatively unconscious, because I am not sure why exactly I made this decision, and had I thought it through a little more carefully, I would have realised that this decision also brought an unnecessary amount of complexity into the picture.

Home manager is a community package² for NixOS that makes it possible to install software in a configuration on a per-user basis. It also exposes a different and higher level of abstraction through which to install software than the 'standard' approach when working in the Nix ecosystem, which is to reference packages in a repository with Nix-built artifacts, the primary one being Nixpkgs. As an example, with Home manager one can install the fish shell as follows:

programs.fish = {
  enable = true;
  interactiveShellInit = ''
    set fish_greeting # disable greeting
  '';
  plugins = [
    { name = "bass"; src = pkgs.fishPlugins.bass.src; }
  ];
};

In addition to the enable option, programs installed thus also provide a custom set of attributes that are configured via Home Manager through which one can customize the installation.

One could also, however, install the fish shell directly from Nixpkgs (in a configuration.nix file) as follows:

environment.systemPackages = [
  fish
];

The benefit of Home manager's approach here is the configuration optionality. In addition to the basic availability of fish on my system (the result of both approaches), I can also declare a string that will become its basic configuration file (interactiveShellInit) and a list of plugins that will be installed alongside the shell itself.

On one level, this is great: all my configuration is in one place. Home manager has an option search where one can check whether a package exists, and what options it provides. In theory, if the documentation of options isn't clear, one can go to the module declaration to try to reverse engineer which option will result in the desired outcome. In practice, though, I've found it cumbersome to do this effectively, especially if I'm new to the software in question and don't have an intimate understanding of its configuration options.

The main issue is that raw NixOS doesn't provide an obvious way to manage dotfiles. Home manager fills that void by providing a level of abstraction on top of Nixpkgs that incorporates a sense of the current user (and hence paths to the user's configuration files); but this additional level of abstraction is constructed in such a way that it is not always clear how to map 'configuring new software X based on its documentation' to 'configuring new software X declaratively in Home manager'. (This is not even to mention the more complicated case in which new software X is not supported by Home manager, in which case one would have to add the new module to Home manager upstream, or work out a workaround.)

In summary, though Home manager is still useful for me due to its dotfile management utilities, it confused me early on, as I didn't always have a clear notion of whether I was installing modules via Nixpkgs or via Home Manager, and whether a configuration option was provided by Home Manager or by NixOS.

Note to past self: When installing new software, work first with a basic installation from Nixpkgs before trying to find an appropriate module in Home manager.

Nix Flakes

Flakes are kind of like a new version for Nix, much like Python 3 is when compared to Python 2. The overall use case is the same, but the tooling is (arguably) better with Flakes, and there are some features (such as pinned versions in a flake.lock) that you fundamentally don't have in vanilla Nix without flakes. I personally adopted flakes when I first started using NixOS for no particularly good reason; just because they seemed like the 'modern' way to work with Nix, a recipe for getting in over one's head with the Nix ecosystem, as has come to be the central theme of this post. As with Home manager, I don't regret the decision to use Flakes exactly, as I do think that they were and are useful for my use case. Like Home manager, however, choosing to adopt them steepened the learning curve in a way that probably could have been avoided had I had a better sense of why they were useful for me.

I think of flakes as essentially a different and more integrated way to write Nix configurations. They're not actually fundamentally necessary for my NixOS configuration, as it's not primarily a 'library' that is intended to be used by others: it's really just a standalone software that serves a purpose for me. Where flakes come in handy for me is when I am writing Nix files in association with particular coding projects, where I want a development environment, a build process, and some set of artifacts that are produced from the project. For more on how NixOS with Flakes works in this context, I refer the reader to the excellent section of the open source book NixOS in Production, ``The big picture".

Note to past self: don't use flakes with your NixOS configuration; learn about them afterwards and apply it to dev environments.

Footnotes

^ A glorious time in one's life where it is reasonable to adopt software simply for the sake of it. I did something similar with the keyboard layout colemak, which is now essentially the only layout in which I can type.
^ I use the word 'package' here in a very general and non-technical way. The Home manager documentation says that it is a ``Nix-powered tool". It is not a module, as modules refer to something specific in NixOS; and it is also not technically a package, as packages are the software bundles on Nixpkgs.

Writing Academic Essays

Lachlan Kermode — Wed, 01 Jan 2025 00:00:00 GMT

[idea:writing-academic-essays]postgeneral2025-01-01

Writing Academic Essays

As an undergraduate student in the US, I feel like I never properly learned how to structure an academic essay. This was no doubt partially due to my being a bad student who never asked the right questions; but I do think that there is something in the structure of an interdisciplinary liberal arts education in the U.S. in the 21st century that regretfully avoids giving students strong guidelines on how to write a coherently argumentative essay. The following are a set of principles that influence how I have come to write essays– which is to say that they might be totally inadequate for many forms of writing, and don't necessarily stand up as guidelines for getting writing accepted anywhere for publication. They are rather my own reflections on what I have come to see as 'good' academic work, i.e. work that I find valuable and meaningful in its own right, regardless of where/how it is published.

Have a thesis statement

This is an almost brazenly obvious point, but it continues to surprise me how much writing I read where I cannot seem to make a clear argument out. A thesis statement is a sentence that can be formatted as an 'I argue that…' phrase. There are good arguments (defensible, well-evidenced, offers some bearing on something important) and bad arguments (unfalsifiable, shoddily evidenced or under-substantiated, apparently correct but unimportant).

Each section and paragraph of the essay should service your argument

If the relationship between a paragraph or section of your essay to your thesis statement is not clear, then you should provide orienting notes so that the reader can follow along. For example, 'In order to show that [argument], it will be instructive first to look at [content in this section] to establish [the basis for a dimension of the argument that I will go on to outline]'. Tidbits or references that are not directly in service of an essay's argument should either go to a footnote ('Though it does not bear directly on my argument here, …'), or should not be included in the essay.

Start with a clear presentation of one concept or object

A thesis statement is often composed of terms or objects that the essay will mobilize in a particular way. It is important to show your reader how you understand these terms, a demonstration that can but doesn't necessarily need to reference theorists or texts that use the term in similar ways to how you understand them. For example, if your argument is that 'cycling shoes are a commodity that epitomize late capitalism', it is normal and perhaps advisable to structure your essay according to a progressive unpacking of the terms in your argument: 'I first define cycling shoes as a variant of shoe that first appeared in the 60s… I then indicate that a commodity is an exchangeable object by way of Marx's theory of the commodity in Capital… and finally I argue that late capitalism started in the 70s and persists until today'. A writing advisor once told me that all good essays in both undergrad and grad school are simply extended detailings of singularly enigmatic terms ('late capitalism'), and I am not sure that they are wrong.

Don't try to predict the future

I don't know if it is the increasing instrumentalization of the humanities by 'the sciences' in service of solving their ethical conundrums or something else in the academic air, but predictive arguments about the future seem on the rise. Arguments of the structure 'I argue that [X] will happen because [Y]' are generally on shaky ground (I would argue). Even if you are convinced that you know how something will play out, it is almost always better to temper the tone of your argument to something a bit more moderate, i.e. 'I argue that [X historical or philosophical concept] reveals something important about the [current situation of Y]'. Admittedly, this tempered format is less falsifiable, as it doesn't presume to know what will happen next in some area of social life. But if you want to write an essay that will withstand the test of time, you should be very careful about what straightforward predictions you make.

Provide the reader with insight they would not obviously get by reading your essay's sources

This is the difference between an essay that simply represents other arguments and one that makes its own. Representing and interpreting other arguments is an important part of academic writing, and you should not shy away from doing the work of explaining how you understand another text or writer in your essays. Often this is the most important work in an essay, i.e. showing the reader how you understand other texts. But your argument should not be a simple addition of the texts together (argument X + argument Y = my argument). You should rather provide the reader with something that they could not obviously get simply by reading your sources; i.e. to show that a dimension of argument X can be better understood by taking into consideration the claims in argument Y.

Don't assume that the reader is familiar with the texts/sources you are discussing

Good writing does not assume that a reader has read the texts and sources under examination carefully. Rather, it shows the reader the important aspects of these texts through careful interpretation in service of an argument. If my argument is that Freud's method of interpreting dreams allows us to read tweets as emanations of the unconscious (an argument of which I would be highly suspicious), you should not assume that your reader has read Freud or has any clear notion of what he means by the unconscious. You should rather present a reading of what you understand Freud to mean when he talks about the unconscious in service of showing its relevance to tweets.

A greater number of citations doesn't always amount to stronger evidence

Citations are an important part of academic writing, as they show that your argument is situated within a body of work and that you are not writing 'off the cuff' with no idea of what has previously been said about your object or subject. But citations should also be employed judiciously. As a general rule of thumb, if you do not go on to offer a direct interpretation of a sentence in terms of its structure or word choice, then it is almost always better to paraphrase the original point (while still providing the appropriate page reference from which you drew it). Paraphrasing is a form of interpretation; it shows how you understand a text or object, and what you see as important about it for the sake of your argument. You should not drop in extra citations for the sake of it, or simply to show that you know about the existence of a text or book (except occasionally in a footnote, if you want to flag that some text exists but is not being addressed or bears some tangential relevance to your argument).

Fall semester 2024 update

Lachlan Kermode — Fri, 27 Dec 2024 00:00:00 GMT

[idea:fall-semester-2024-update]postgeneral2024-12-27

Fall semester 2024 update

The pause in proceedings here has been principally due to a (final) semester of coursework and teaching at Brown University; the other side of a PhD in the U.S. I have now advanced to candidacy, which means that I have completed all requirements for my PhD except the submission of the dissertation itself. (This is also known as being ABD, ``all but dissertation".)

I was in Providence for this stint (September through December 2024), and though the period has certainly been generative in terms of meeting milestones and generating sparks for further research, I didn't manage to make any time to write here throughout. So I return with a brief recap of the contours of my past few months.

Teaching CCeAI

One of the features of the PhD program in Modern Culture and Media at Brown is that students are expected to teach an undergraduate seminar of their own design around their fourth year. It is recommended that this seminar trace a student's intellectual fields, as it is ideally taught just after completing the fields examination;. (In my case, I sat the fields exam halfway through the same semester in which I was teaching my own course, due mainly to logistical issues with convening my advisors.)

I titled my course 'Capitalism and Computers in the Era of AI', acronymically 'CCeAI'. As expected, about half of my students were majors in Computer Science, interested to relieve their technical coursework with some critical reflection on the economy and politics of computing more generally. The syllabus lays out the course's focus:

What is the relation between capital and the computer? To answer this question, this course attends to the particulars of the computer as a mathematical proposition, to its claim to legitimacy as a logic of superior military and political organization, and to its elaboration as something of a natural philosophy in scientific practice. In doing so, we will formulate a conceptual basis to critically assess the glittery promises of current discourse invoking Artificial Intelligence (A.I), which sees near-total automation as inevitable and the threat of sentient Artificial General Intelligence (A.G.I) as nigh. We will establish a critique of the computer as a structure (i.e. as a philosophical arrangement with a specific and persistent imagination) to think critically about popular ideologies associated with A.I., often put forward by those with capital stake in its commercial success.

All in all, I was pleased with the way that the course turned out. Due to unexpected interruptions in the academic calendar coupled with a sense that we needed to spend more time with the material in the early weeks to properly get a grasp on it, we skipped the week titled 'Ideology' (Zizek), as well as the final week titled 'Psychoanalysis?'. I am still myself working out the appropriate way to cross-pollinate Marxian theory and Lacanian psychoanalysis, I realised as we approached each of these weeks, and so it felt an unreasonably far throw in the course's overall argument to sidestep into Zizek's theory of ideology (which I wrote some about here and here;) directly after a tough week attempting to chart Marx's theory of commodity fetishism via Michael Heinrich's (excellent but dense) introduction to it; and while I would estimate the final week's reading as at a slightly more appropriate level for the course, there was a holiday that I had not accounted for throughout the semester (election day).

The texts in this syllabus represent a trajectory of my own thinking regarding the intimate philosophical relation between capital and the computer. The course locates the computer through three supposed inventors of it in the first three weeks: Charles Babbage (1830-70s), Alan Turing (1930-50s), and Norbert Wiener (1920s-60s). Each of these three men is esteemed as a populariser of the computer's concept, and we can glean important characteristics of the imaginaries that proliferate from their writing and writing about them.

The course's core argument is that there is a fundamental relationship between the computer and capital, and so in the weeks following an elaboration of the computer, we looked to firm up a sense of capital's contours and characteristics, particularly the aspects of its theory that have been explicitly employed in critical discussion about cybernetics and the computer. Each of the weeks whose titles end with a question mark are designed to provide texts through which students can observe and practise critical thinking with respect to the computer, capital, and cybernetics, consisting of contemporary attempts to theorise a relationship between these terms, rather than of introductions to a concept as the earlier weeks' texts do. The course culminates in a ten page paper on a topic of students' choosing.

Through teaching the course I have certainly learned that there is a real difference between the way that one reads as an undergraduate, and how one reads as a graduate student. I think I knew this already, but in some sense I have resisted the idea, as I sought to flout the hierarchical distinction as an undergrad, and of course don't believe in any fundamental demarcation. (There are certainly particular undergrads who are smarter / better readers than certain grad students.) But in general, undergrads in US institutions are beset with circumstances that make it difficult to imagine that there is value in reading texts closely, repetitiously. This is not something to hold against undergrads, I don't think, but rather against the regimes of value that flood in and out of (neo)liberal arts institutions. Five courses a semester is both far too much and far too little. One doesn't have to work particularly hard to keep head above water, especially at Brown where grades don't hold much weight; but there is also too much noise in the days and weeks if one were to want to seriously study something particular in the oceany midst of weekly obligations.

The Internet and Datacenter Operating Systems

I decided to take two graduate seminars in computer science to wrap up my masters degree in the department, and I learned a lot about research in the discipline through the experience. The experience of reading four papers a week from operating systems and Internet research was instructive in the sense that I realised how different the work of keeping up with computer science is from keeping up with continental theory. The journals are different, of course. But so is the kind of effort that reading each paper involves, and the time one must set aside in the day to engage with it. I like the difference in the experience between the two disciplines in which I have decided to make an effort to become literate; though it is certainly true that I feel more equipped at this stage to parse papers in continental theory than I do those in computer science.

The work involved in each (systems) paper in computer science is also enormous. It takes at least a year to prototype a system, to run experiments, etc, and it is often a multi-person effort. The scale of the research is still relatively small and individualistic compared to other scientific disciplines, but it is a notably different (and more social) experience of research than reading and writing philosophical papers.

The main thread that emerged from these two courses for me was a burgeoning familiarity with vector databases, which I had decided to try to get a handle on early in the semester so as to be able to code something substantially systems-oriented for my final projects in the courses. The system I have now started working on is currently called OAK, and while I am not yet sure that it offers something distinctively innovative or original in vector databases research, it is at least a very good way for me to practically get a handle on the current state of vector indexes, and to think about what kinds of opportunities there might be for turning research into practice in this domain (i.e. coding up research papers as new components of OAK).

German Idealism and Canguilhem

The other two courses that I either took or audited this semester were respectively about Kant with political theorist Alex Gourevitch, and about Georges Canguilhem with my advisor and chair, Joan Copjec. The first of these will likely see some thematically related posts in the coming weeks, as I am now starting work on a paper about Kant and Marx to submit as the final paper for that course, so I won't discuss it at length here. The second course with Joan, which I audited, was revelatory even though I was not able to complete much of the reading for it in advance. The basic premise was that Canguilhem's ideas bear some relation to Lacan's, and so we sat each week listening to Joan hypothesize connections between the two thinkers.

While it perhaps doesn't come directly on the back of this seminar, and is rather an idea that I have been stewing for some time, I intend to read Joan's work more closely in the new year, and to record the way in which I understand it via video, rather than here in writing. I am not exactly sure how this will work at this stage, so I will not speculate further. But I have resolutely decided that I ought to get deeply familiar with her body of work, for though it is far-reaching, it is relatively short in terms of actual page counts. (A credit to the incredible effort that Joan puts into each paper, not dissimilar to the year-long efforts of systems papers I note above.)

Cycling typing

Lachlan Kermode — Wed, 13 Nov 2024 00:00:00 GMT

[idea:cycling-typing]postgeneral2024-11-13

Cycling typing

Last summer I got into cycling. I've long been a standing desk-er, and going out for long zone 2 rides and wondering what to think about made me wonder: could I work at a desk while I cycle?

I have a stationary bike trainer, so I first tried the obvious thing and just slid it slightly under my standing desk. This made certain kinds of work doable– watching videos and reading articles– but it makes for awkward and uncomfortable keyboard handling. Even reaching up to hit the 'J' key to scroll down on the browser (as I am a vimium user) interrupted the flow of thought and poise. Attempting to engage with content while cycling also made me realise that these small physical interruptions had an outsized impact on my ability to do anything meaningful. Videos and podcasts were doable in this MVP, but it left a lot to be desired.

A perspicacious user on the Acquired Slack noted that my keyboard was from the ZSA ecosystem, and pointed me towards the tripod mount for their Voyager keyboard. These are essentially two magnets that attach to the backs of the sides of the split keyboard, and as such allow you to mount the keyboard to any standard tripod screw. So I bought a couple of basic camera clamps and waited in anxious suspense for the parts to arrive.

After some basic experimentation, I have found that the most comfortable position to type while cycling is with the keyboard angled down at a 45 degree angle, with the orientation pointed slightly inwards (i.e. so that the lines from each side converge at the frontmost point of the front wheel). This allows me to rest my palms on the handlebars and to extend my fingers pretty naturally over the top of them to reach the keys. There is definitely a bit of wrist awkwardness palpable in this position, and I aim to experiment with other angles and positions as I get more comfortable with the set up.

But the basic approach has opened up a range of other work possibilities while slugging away at zone 2. This article, for example, was entirely written, researched, and hyperlinked from my position here on the bike. (The only exception was when I had to stop cycling periodically to reconfigure my webcam to show the full context of my bike– and to take and link a photo below; and to actually upload the video.) I can code with almost the same fluidity of computer navigation as I can when at my standing desk.

Note that I am a 'vim everywhere' user, meaning that I context-switch and browse web pages entirely via the keyboard. I only occasionally have to reach up and use the mouse when stuck in some janky context where I haven't yet figured out how to enable vim shortcuts– as is the case in switching the OBS scene in the recording above.

Using two ReMarkables

Lachlan Kermode — Wed, 06 Nov 2024 00:00:00 GMT

[idea:using-two-remarkables]postgeneral2024-11-06

Using two ReMarkables

I'm a PhD student currently, which means that I spend most of my work life reading, or thinking about what I should read next. The timelines for my reading schedules range from what I need to read before a class or meeting next week, to what I ought to read for a paper or research trajectory that won't have any meaningful check-in until next year.

I bought the reMarkable 1 when it first came out in 2017, and was a pre-order customer of the reMarkable 2 when it dropped in 2020. When reMarkable announced and released the new Paper Pro this September, 2024, I had ordered it within an hour. This is partially due to the undeniable fetish that I have for tech, software and hardware alike, being a tinkerer and early adopter by nature. It is also because I resonate deeply with reMarkable's vision and design ethos, which is to recreate the tactile and organizational experience of paper in an electronic tablet. As I often remark to observers when they ask about my tablet(s), the reMarkable will not replace your iPad, and therein lies its beauty. It is at its best when you are jotting down notes as if it is paper, or reading and annotating a PDF or EPUB in place of having a printed copy in front of you.

ReMarkable's exquisite execution of this design imperative, to keep the thing as 'just' or 'better paper', has consequences in a digital workflow. Switching between a PDF being read/annotated and a notebook in which notes/thoughts are being transcribed is not the smoothest experience. One has to swipe down with two fingers from the top of the device (a somewhat error-prone operation given the e-ink's diminished responsiveness when compared to a touch-screen smart phone), select the notebook to switch to, and then wait a second or so for it to open. This is arguably a software deficiency that could be improved, but the fundamental problem can be articulated in terms of the 'better paper' design ethos: a reMarkable tablet is only one piece of paper at any given time. Switching between a notebook and a book therefore naturally has some overhead.

Given my already admitted fetish for new tech, I therefore use two reMarkable tablets in my current paper- and book-reading setup. The new Paper Pro is the tablet on which I view the paper or book that is being read. The bigger screen makes a huge difference when it comes to double-column papers, and the color display means that I can employ a range of highlighters to give a better map of the document when I return to it. (Yellow is for generally important phrases, blue for definitions of terms or critical concepts, and green for terms or references that I want to look up after reading.)

My reMarkable 2 sits to the right of my Paper Pro, and serves chiefly as a notebook where I write down important points as I read. This notebook is my main 'map' of the paper or book that I return to if I want to understand what I figured as important while I was reading. I employ a similar colorized strategy for notes, where black is my default color, and blue indicates a thought in apostrophe or otherwise adjacent idea. Green is for references or things to look up later, and red is reserved for particularly important ideas or notes to self.

Thoughts about what I should read or do next often occur to me as I read, and so I use a notebook on my reMarkable 2 as an inbox for "todo" items that occur to me throughout the day. In the month or so that I've been using the Paper Pro, I've found that it's not (yet) as responsive as the reMarkable 2, presumably as the team hasn't yet optimized the device's software through bug reports and iteration. So the reMarkable 2 serves as my notebook, but also as the device where I can context-switch to jotting down a TODO.

My one other major nit with the reMarkable is the lack of a fluid way to flick to endnotes or references in a document, especially in PDFs that don't have the relative links embedded. Even where the relative link does exist, meaning that the tablet allows you to click on the endnote or reference and jump directly to the page, the return link disappears after 3 or 4 seconds, which is almost never long enough to read the endnote or scrutinize the bibliography entry. In the event that I go over time, the link back disappears and I need to manually find my way back to the page that I was on.

When the link doesn't exist, the best way to navigate I've found is to swipe up with one finger from the bottom of the screen, and then drag the slider to the approximate point in the document where one understands the relevant endnote to be. Not only is this endnote discovery error-prone and relatively slow; the same issue of returning emerges. The document slider retains a reference to the page that one comes from, but it times out in 3 or 4 seconds just like the backlink for endnotes. Moreover, if one doesn't jump directly to the correct page for the reference (which is almost always the case, given that the slide operation is approximate unless one directly keeps track of page numbers), then one similarly loses the reference to the original page.

Having a second reMarkable gloriously solves some of these problems, as I can use my reMarkable 2 to view the endnotes and references as I read on my Paper Pro. This has the added benefit that, when referencing endnotes, one often wants to retain the last page that one viewed an endnote on, as the next one will follow in sequence.

I am fully aware that having two reMarkables is not a particularly low cost solution to my usability issues. But because of the aforementioned fetish, I had both tablets in any case, and have found that bringing the two of them with me to a cafe greatly enhances my productivity. The experience of reading with one tablet and note-taking with a second has allowed me to conclusively leave my laptop at home in most cases, as I would previously take notes in a text editor on it while reading the text on my reMarkable. Because the full reading/note-taking experience is now all in reMarkable(s), I am much less likely to get distracted and idly open up email or a browser tab, as I disconnect my phone from the Internet when deep concentration is needed.

Svitlana Matviyenko and Sarah Pourciau

Lachlan Kermode — Thu, 02 May 2024 00:00:00 GMT

[idea:svitlana-matviyenko-and-sarah-pourciau]postreviews2024-05-02

Svitlana Matviyenko and Sarah Pourciau

Today's log discusses two projects, Svitlana Matviyenko's and Sarah Pourciau's, as a follow on from the last log;.

Svitlana Matviyenko

I came to Matviyenko's work through an event here at the IWM, which took place this past Friday. She is currently an associate professor at Simon Fraser University (SFU) in the School of Communication, and she was previously (I seem to remember, although I cannot now find the Internet proof of this) an assistant professor in NYU's department of Media, Culture and Communication. Both of these departments/institutes are at the forefront of a particular kind of media theoretical critique in the American academy; though like many American things it has something of a European and global reach. Wendy Chun, who used to be a professor in my department;, received a big grant to kick-start SFU's Digital Democracy Institute circa 2020, and ever since has surely played some kind of instrumental role in bringing together an impressive range of up-and-coming names in critical accounts of computing (Steph Dick, for example– Alberto Toscano is also currently visiting faculty there).

Matviyenko did her dissertation with Nick Dyer-Witheford, a scholar who is best known for his first book, published in 1999, Cyber Marx: Cycles and Circuits of Struggle in High-technology Capitalism (Dyer-Witheford 1999). Matviyenko's dissertation is titled Lacan's Cybernetics (Matviyenko 2015), and attempts to track the relationship between Lacan's presentation of psychoanalysis across his seminars (which took place annually between 1953 and 1981) and cybernetics, an intellectual movement/field of thought whose first wave is typically traced to circa 1948, the year in which Norbert Wiener published Cybernetics: or Control and Communication in the Animal and the Machine (Wiener 2019), and which is generally understood to have had a second (1960s), third (1990s), and sometimes even a fourth wave (2010s?) in the ensuing years.

There is no doubt that Lacan knew about cybernetics: he first treats the subject explicitly in his second seminar (1954-55). Though this seminar has come to be known as his cybernetic seminar, Matviyenko makes a convincing case that there is development of the ideas in seminar II, in at least seminars III, XI, and XX. Matviyenko's contextualization of Lacan's position on cybernetics is extremely useful, even if the specificity of an argument beyond contextualization is at times hard to make out. Her dissertation's general task is to "contextualize Lacan's work in the variety of debates on cybernetics" (Matviyenko 2015, 27), and to situate "the cybernetic episteme" as an influential discourse in and on Lacan's thought. This argument consists of what I find a confusing kind of historicism, i.e. through a set of claims that pose cybernetics as a Foucauldian episteme which "serves as the framing for knowledge upon which psychoanalysis operates" (Matviyenko 2015, 43).

I find this confusing as a methodological stipulation in relation to Lacan's seminars because the figure of the Foucauldian episteme seems to me very much in tension with a 'Lacanian' notion of 'discourse', of tracing cause and effect in the language of Lacan's seminars. For example, Matviyenko's summary of her method in chapter 1 is as follows:

To summarize: the cybernetic episteme organizes the subject of science's relation to knowledge as the mechanism of blackboxing maintained by the institutional power formed within certain conditions of mediality. (Matviyenko 2015, 43)

To start, there are a number of terms in this summary that need elucidation. This sentence comes after a section that present the basic context of terms such as communication, control, and homeostasis in their cybernetic formulations, and in the section preceding this summary the Lacanian notion of the "agency of the letter" (Matviyenko 2015, 41) has been mapped. I understand Matviyenko's claim to be meaning to say something to the effect of: cybernetics' presentation of the world as nothing but traceable cause and effect, i.e. its behaviorist and scientistic basis, is an intellectual basis upon which Lacan's idea of the agency of the letter, as well as his related presentation of the notions of 'repetition automatism' and 'the subject of science', can be developed. Yet Matviyenko confuses this point in the quotation above by introducing an idea of 'blackboxing' (a term not properly developed at this stage, nor later in the dissertation) and by suggesting a hazy relation to 'institutional power'.

In other words, Matviyenko's attempt to chart the relationship between cybernetics and psychoanalysis by way of Foucauldian 'archaeology' results in a mystification of the specificity of her historical claim, preferring instead to lean on Latourian/Foucauldian language of mediality and power in a way that I find obscuring rather than enlightening. Throughout the dissertation, she wants to suggest that there is something Lacan's thought shares with/inherits from cybernetics– an unproblematic claim in its generality, given the aforementioned direct treatment Lacan gives to cybernetics– but I have a hard time making out what exactly the substance and consequence of this relationship is in Matviyenko's view. Though ample context is provided about Lacan's views on automatism and cybernetic notions of the subject, the strongest kind of claim Matviyenko herself makes is so generally media-theoretical that it seems barely to be a claim at all: i.e. "psychoanalysis reflects contemporary conditions of mediality" (Matviyenko 2015, 54), or "[Lacan's] thought seems to be caught in the changing condition of mediality" (Matviyenko 2015, 55).

Whatever one can critique in the historicist flattening of argumentation, though, it is undeniably still very rich on context. The dissertation's second chapter, "The Subject of Science", is particularly useful for me in this regard, as it lays out the connections from Norbert Wiener's thought to Leibniz and Pascal, as well as Dutch mathematician and astronomer Christiaan Huygens (working and writing in the 1650s), and putting these references in relation to Lacan's conception of modern science (via Alexandre Koyre) and automatism (via Descartes). Matviyenko convincingly posits both cybernetics and psychoanalysis as taking up the challenge of "the model of the body as a system-coupled-with-the-environment that transcends Cartesian mind-body dualism" (Matviyenko 2015, 78): that is, the problem of the subject in science, of the material/ideal dialectic, of thought's relationship to the world at large (read: ontology, epistemology, phenomenology). It seems clear to me that psychoanalysis should be seen as a critique of cybernetics (rather than, as Matviyenko seems to see it, as a complementary or comparable system of thought, similar in the sense of similarly deficient and perspectival)

Chapter three, "Freud's circuitry", is similarly rich in its presentation of determinism and thermodynamics as influences on Freud's thought. Lacan's notion of structure is then traced as the sublimation (to use Freudian vocabulary, rather than Matviyenko's) of Freud's reckoning, and the apparent allusions to Claude Shannon's concept of error in Lacanian repetition automatism. The relevance of these 19th century discoveries on psychoanalysis is clear and undeniable, and again my main quibble of the dissertation's method would be in that Matviyenko doesn't do much argumentatively with this context other than provide it. Though perhaps I should not be so quick to judge such an argumentative lack, as the text is certainly immensely valuable as a reference for further study (which is perhaps all one should ask of a dissertation).

A slightly more substantive claim is reached by the end of the text; namely that "both psychoanalysis and cybernetics belong to the epoch of graphocentrism, which began in the nineteenth century and extends to the present moment" (Matviyenko 2015, 193). In making this claim, Matviyenko gestures to the ongoing relevance of cybernetics and psychoanalysis to contemporary ideologies and epistemologies of science and the subject, and I certainly agree on this point. Chapters 5 and 6 offer an intriguing tour of more recent theorizations (Kittler, Latour, Foucault, Didi-Huberman, Florian Cramer, Alex Galloway) in an attempt to link graphocentrism with what is seen as its current instantiation as interpassivity. There are many names and theorists in these two chapters (as there are in the earlier ones too, I suppose), and I again find myself somewhat confused by the blur of terms: a "dialectic relation between the modalities of extension and prosthesis" (Matviyenko 2015, 172), "parasitic passivity" (with reference to Michel Serres), and the ultimate argumentation that today's interpassive subject has "lost meanings and practices of solitude, of sensations of mild joy" (Matviyenko 2015, 190).

I don't necessarily dispute these claims, but I am suspicious of the idea that our 'digital condition' is something distinctly different from some previous historical/social condition, when attention was more integral, and desire not quite so distorted. Or at least, I am yet to be properly convinced by an argument in this vein. Perhaps it is just my dormant computer utopianism, instilled in me by a formal education in computer science, which will not allow me to believe that computers render us necessarily more distracted, alienated, or less capable of freedom than what came before. But I still can't quite stomach this media-theoretical kind of thinking that slaps new terms on new conditions and figures it critical. Something I need to work through, I think.

I had wanted to also briefly touch on Matviyenko's most recent book, co-authored with Nick Dyer-Witheford (Dyer-Witheford and Matviyenko 2019); but it will suffice to say at this point that I was similarly underwhelmed by the theoretical effects of this 'psychoanalytic' and Marxist conception of cyberwar. I am not sure that I feel the stakes of the theorization– not in material-political terms, as it is obviously evident that the term 'cyberwar' denotes something perverted and violent going on at global scale– but in the sense that I am not sure what more I understood about it through the reading of this book. Again: perhaps this is a function of a shallow kind of reading on my part, and it might benefit me to work through it again at a later stage.

Sarah Pourciau

I have not considered Pourciau's project as extensively as Matviyenko's, as I have only read one recent essay of hers in Critical Inquiry titled "On the Digital Ocean" (Pourciau 2022). This essay, in contrast to Matviyenko's writing, I found something of a revelation. It sees something problematic in the discourse on 'digitality' in the comparison of data to the ocean, and marks this problem as one that stretches far beyond the historically 'computational' per se into the reaches of philosophy that feminizes and passivizes matter, posing the mind as a master of the seas:

From the book of Genesis through the twentieth century, the oceans of the past figure the feminine matrices of divine and human creation, the wombs and tombs of the earth, the fertile yet deadly milieu of the sirens and the mermaids. They are figures of endless flux and fecund formlessness, and thus of infinitude, in the oldest sense of what cannot be fenced in or finished. (Pourciau 2022, 234)

The problem of the digital ocean, in other words, for Pourciau is an age-old problem of conceptualizing infinity. A certain computerized conception of the world wants to render it totally calculable, atomizably "composed of distinct digits (0s and 1s), which is to say that it is remainderlessly divisible into discreet bits, which is to say that it is per se accessible to the logic of conceptual definition and analysis" (Pourciau 2022, 235). One ideology of digitalization, in other words, is that it flattens ontology into epistemology. Everything is knowable "as a digital sea of quantum bits" (Pourciau 2022, 235), and thus there is no part of the world that is not conceptually, hypothetically, available to know. ¹

Pourciau goes on to give, in the course of some thirty pages, an enormously erudite review of set theory's impact on the philosophical conception of infinitude, covering Cantor in the first section and Turing/Gödel in the second. The major consequence of this review is to call into question the ontologizing of a 20th-century mathematical conception of infinity: As the essay concludes:

Turing's implicit thesis is that the countable set of computables is big enough–as big as an ocean, even, albeit not an apeiron–and the question we need to ask and answer in response is: Big enough for what? We will have a better chance of making headway on this problem, I think, if we stop calling on the remainder of Woman to do the work for us. (Pourciau 2022, 261)

There is an explicit thread in this essay from these mathematical notions of infinity to Lacanian and post-Lacanian thought ("the remainder of the Woman"), and thus I believe Pourciau's project will be extremely relevant to mine.

Bibliography

“Devs.” 2020. Drama, {{Mystery}}, {{Sci-Fi}}. DNA Films, FX Productions, Scott Rudin Productions.

Dyer-Witheford, Nick. 1999. Cyber-Marx: Cycles and Circuits of Struggle in High-Technology Capitalism. University of Illinois Press.

Dyer-Witheford, Nick, and Svitlana Matviyenko. 2019. Cyberwar and Revolution: Digital Subterfuge in Global Capitalism. University of Minnesota Press.

Matviyenko, Svitlana. 2015. “Lacan’s Cybernetics.” PhD thesis, The University of Western Ontario (Canada).

Pourciau, Sarah. 2022. “On the Digital Ocean.” Critical Inquiry 48 (2): 233–61. https://doi.org/10.1086/717319.

Wiener, Norbert. 2019. Cybernetics or Control and Communication in the Animal and the Machine. MIT press.

Footnotes

^ I recently completed the 2020 miniseries "Devs" (“Devs” 2020) which dramatizes this point by way of a hypothetical quantum computer that can accurately simulate any point in space or time in both the past and the future. The series comes to a close on the point of free will's sanctity, whereby we cannot (the series seems to suggest) solely believe ourselves to be deterministic productions. I would not recommend the series, however: it makes for wince-worthy viewing.

Amanda Holmes and Adrian Johnston

Lachlan Kermode — Wed, 24 Apr 2024 00:00:00 GMT

[idea:amanda-holmes-and-adrian-johnston]postreviews2024-04-24

Amanda Holmes and Adrian Johnston

I have been at the institute (the IWM) for nearly a month now, the first month of three. It is the first time that I have had something close to my own office (I share a room with one other fellow), and the first time that I have had the space and time of mind to attend specifically to the task of my dissertation proper. As a PhD student in Providence, there is a bustle of daily university life - lectures, seminars, meetings - that makes it more difficult to properly think. Perhaps this is because it is the aim of a PhD to find a unique space of thought, and such a pocket (carved out against the majority of opinion, if it is to have any academic value) is necessarily solitary. The PhD project is to design such a space, to educate oneself in how it reflects and refracts existing projects (of existing academics), and to come to some confident enough conclusion in it that one appears salable on the job market.

In the past few weeks I have tracked four contemporary projects that relate to the space I am trying to carve for myself at the intersection of mathematics, Marx, and Lacan;. I'll introduce two of them here (Amanda Holmes' and Adrian Johnston's), and defer two of them (Svitlana Matviyenko's and Sarah Pourciau's) to a later log.

Amanda Holmes

Completing her PhD in philosophy at Villanova in 2020, Holmes was advised by Walter Brogan, Gabriel Rockhill, and Alenka Zupančič. Her dissertation was titled Erotology: The Logos of Eros in Lacan's Return to Freud, and its structure is refreshingly un-American in its singular focus on tracking what Holmes calls 'erotology' in Lacan's thought. Erotology is "the logos of eros, the study of desire, the disciplined examination of those gaps in the universe around which desire gathers…. the discipline of identifying and elaborating the philosophical implications of psychoanalysis" (Holmes 2020, 7). Erotology is the fourth term that psychoanalysis, as an inflection point in and critique of the Western (European) philosophical tradition, introduces to the reigning triumvirate of ontology (the science of being), epistemology (the science of thinking), and phenomenology (the science of feeling). Lacan's return to Freud, a project that takes psychoanalysis seriously as a philosophical invention and intervention, is erotological because it is "a discourse that takes desire as its central focus" (Holmes 2020, 8).

Holmes' dissertation does what good dissertations should do, I think: it synthesizes the state of a discursive space, offering the seeds of a 'new' argument only by first demonstrating erudition and pedagogical generosity as a writer. There is an overwhelming sense (in America) that one's dissertation is supposed to operate as a 'scientific' breakthrough, exposing some connection heretofore unseen and unconceptualized. This contributes to the reigning paradigm of paranoiac collaboration and publication, whereby there is an idea that one's academic space or niche can be stolen, as if intellectual specialization were a patented technology that needs regulatory mechanisms to protect its strategic advantage. This framing is not to downplay the very real and common occurrence of more senior scholars taking advantage of younger ones as a fungible labor force, manipulating the cult of authorial singularity in the humanities through their positions of relative power so as to significantly disadvantage and dispossess more junior persons of their due credit and/or acclaim. But it is to suggest that we should consider the limitations of the conceptual treatment of academic ideas as intellectual property, i.e. as salable commodities that should circulate in such a way that the benefit gained thereof warrants a financial or reputational 'interest' that accrues to their authors. Some will say that the commodified condition of the market of ideas is just the way it is, the realpolitik of (American) academia within which one must operate to have any relevance, and/or to feed oneself and one's "family". Perhaps this is the case. But shouldn't we still do our best to imagine it otherwise, and implement this imagination as much as we can, as much as we are allowed to?

It is rich, perhaps the same some will say, to suggest that the dedicated study of an infamously cryptic and white French man's seminars from the 60s and 70s represents a progressive (i.e. less commodified) kind of scholarship. And it is probably true that Lacanians have failed to convincingly articulate why such a thing could really (as I am coming to believe it to be) the case. But I really do digress, here, as the point that I set out to make is that Holmes' dissertation on Lacan is a valuable addition to the available other introductions to the Lacanian ethos of desire (i.e. Zizek's, Gallop's, to name a few that I have discussed on this site), and perhaps one that is the most up-to-date regarding more recent Lacan scholarship, such as Zupančič's What Is Sex?.

Adrian Johnston

I intend to discuss Johnston's most recent book Infinite Greed at more length in future logs on this site, as it is one of the most erudite studies of the relationship between Lacan's and Marx's thought available. In recent weeks I have read a 2017 essay that prefigures the argument in that book, and I am now working through a manuscript of the book itself.

The core argument (of both the essay and the book, as I understand them) is that Marx's critique of political economy in Capital prefigures some of the core psychoanalytic insights, the discourse of which Johnston suggestively refers to as libidinal economy. There are "temporal dialectical dynamics between past and present (as well as future)" (Johnston 2017, 277) that Marx's and the psychoanalytic accounts share, and in articulating this common understanding one can marry metapsychology (psychoanalysis) and historical materialism (Marx). The terminological lodestone of this effort is Trieb [drive], a word used by Freud, Lacan, and Marx alike.

Drive in the Lacano-Freudian sense of the term can be understood through Freud's enigmatic and polarizing suggestion that the human is characterized by her/his death drive. The final chapter of Alenka Zupančič's book What Is Sex? (Zupančič 2017) is a sublime rumination on and interpretation of the death drive as it should be understood: as the name for the object-less affect that emerges from the unknowable contradiction in consciousness called 'sex'. At once aimed at reproduction, as a crude Darwinian imaginary would have it, and paradoxically also at a self-annihilating end in 'death', the death drive is what differentiates a speaking being from an animal in Lacanian thought. Speaking being is characterized by a contradiction in its telos (purpose), whereby there is an element of indetermination in its (life) trajectory that gives rise to a special kind of under-determined drive called desire. Desire has no overdetermined object: it is a subject that strives not for something in particular, but for something 'abstract'. Thus Lacan refers to the object of desire as the objet petit a; for it is not an object in particular, but an absent and negative object that must be 'filled in' in order for the striving to make sense in a causal register. We do not desire because desire can be achieved, fully satisfied and sated. We rather continue to desire exactly because the object of desire (objet petit a) cannot ever properly concretize as something positive (a person, an amount of money); and so once we achieve what we thought we wanted, what it is we want has moved on.

Johnston's project is to put this Lacanian register of desire, understood as the contradiction in the multitude of drives a subject suffers, in touch with the way that Marx talks about Trieb in his theory and system of capital. He sees drive as a commonality between these registers, in that in both "drives are mediated productions instead of immediate givens"- and in that 'drive' is the constitutive terms for the subject in society as such: "The (drive-)object fabricates the subject (of drive)" (Johnston 2017, 280).

For subjects dominated by capital's abstract forms, the range of possible materializations of desire's covering over are reduced to two specific kinds of drive: capitalist drive ( M-C-M' ) and consumerist drive ( C-M-C' ). The capitalist is one who feeds money ( M ) into the economy with the hope and intention of generating a greater quantity of money ( M' ). The consumerist, on the other hand, is one who continues to take their commodities to market ( C ) in order to procure more and more different kinds of commodity ( C' ) in a fashion that is 'plastically' related to their direct, reproductive need for them.

Johnston is careful to distinguish between these two subject positions Marx's theory of capital's contortion of desire, and he is also careful about the specific structure of the relationship that he sees as giving rise to a compatibility between metapsychology and historical materialism:

plastic drive structure, as theorized within a metapsychological qua philosophical/psychoanalytic anthropology, is a necessary condition for capitalism's peculiar libidinal economics as per historical materialism. It becomes a sufficient condition [for one to call it 'capitalism'] when capitalist production, spurred by the pursuit of surplus value, begins exploiting this plasticity for its gains. (Johnston 2017, 283)

Psychoanalysis, in other words, theorizes a "plastic drive structure" in the speaking being (human) in general, i.e. independent of capitalism as a specific configuration of the condition of value's production in society. Historical materialism, complementarily but with a different scope, theorizes the contortion of desire - as reduced to only two dialects of drive - by way of an historically specific mode of production (movement of value) called capital. Johnston's analysis exegetically focuses on the German word for drive that both Marx and Freud use, Trieb, as Marx names the affect resulting from capital's structure eine besondre Form des Triebs [a particular form of the drive]. Capital's particular drive, Johnston suggests in his book, can be understood as infinite greed: as the drive is always towards the possession of some surplus amount, whether it is more money or something more in a different kind of commodity ( M' or C' ).

Bibliography

Holmes, Amanda. 2020. “Erotology: The Logos of Eros in Lacan’s Return to Freud.” PhD thesis, United States – Pennsylvania: Villanova University.

Johnston, Adrian. 2017. “From Closed Need to Infinite Greed: Marx’s Drive Theory.”

Zupančič, Alenka. 2017. What Is Sex? MIT Press.

one long token soup

Birdkeeping

Birdkeeping

Re-routing the flight path of AI discussion

From beads to birds

The art of falconry with birds

Footnotes

References

Taming jagged intelligence with beads and jj

Taming jagged intelligence with beads and jj

Planning with beads

Version control with jj

Reflections on context engineering

Footnotes

Writing weeknotes

Writing weeknotes

Footnotes

Ben Tarnoff and Quinn Slobodian: Muskism

Ben Tarnoff and Quinn Slobodian: Muskism

Writing in Typst | Hacking on Neovim with Claude

Writing in Typst | Hacking on Neovim with Claude

What is a ‘good’ writing system?

Enter: Typst

Enter: Claude Code

My new writing environment

Future work

Footnotes

Mladen Dolar, Rumors

Mladen Dolar, Rumors

Bibliography

Alma Steingart, Axiomatics: Mathematical Thought and High Modernism (introduction)

Alma Steingart, Axiomatics: Mathematical Thought and High Modernism (introduction)

Abstraction and Modernism

High Modernism and Axiomatics

Mathematics and Mathematization

Bibliography

Writing with Rheo

Writing with Rheo

Karl Marx: Capital, Volume 1 (chapter 22)

Karl Marx: Capital, Volume 1 (chapter 22)

Bibliography

Footnotes

Is psychoanalysis a Science?

Is psychoanalysis a Science?

Bibliography

Karl Marx: Capital, Volume 1 (chapter 21)

Karl Marx: Capital, Volume 1 (chapter 21)

21: Simple Reproduction

Erich Hörl: Sacred Channels: The Archaic Illusion of Communication (chapters 1 and 2)

Erich Hörl: Sacred Channels: The Archaic Illusion of Communication (chapters 1 and 2)

Thinking and so being

Intuition is the key to everything

Leibniz leading the blind

Kittler in the kitchen

Bibliography

Terry Eagleton: Why Marx Was Right

Terry Eagleton: Why Marx Was Right

Bibliography

Using NixOS

Using NixOS

NixOS

Home Manager

Nix Flakes

Footnotes

Writing Academic Essays

Writing Academic Essays

Have a thesis statement

Each section and paragraph of the essay should service your argument

Start with a clear presentation of one concept or object

Don't try to predict the future

Provide the reader with insight they would not obviously get by reading your essay's sources

Don't assume that the reader is familiar with the texts/sources you are discussing

A greater number of citations doesn't always amount to stronger evidence

Fall semester 2024 update

Fall semester 2024 update

Teaching CCeAI

The Internet and Datacenter Operating Systems

German Idealism and Canguilhem

Cycling typing

Cycling typing

Planning with `beads`

Version control with `jj`