Art As Engineering

A Conversation with Gabriel Duquette and The Sublemon

Gabriel Duquette is a co-founder of Liposuction (tagline “aesthetics without all the fat”). He started the site with The Sublemon, who studied art at Yale before contributing to Carcinistion and Ribbonfarm. Both are interested primarily in “retro-engineering” and applying “epistemic hygiene” to matters of taste and aesthetics. They write loosely within the community — and from the intellectual framework — of postrationalism (alongside critics like Sarah Perry and Venkatesh Rao).

Key terminology they’ve coined or championed:

Compression: “AI scientist Jürgen Schmidhuber suggests the idea of ‘compression’ as the explanation for both why art exists and why it is pleasurable. The gist of Schmidhuber’s concept of compression is that the human brain is itself a kind of hard drive with a limited amount of space. Given that the brain is space-limited, it makes sense that information that uses that space efficiently might reward the brain with pleasure. It’s in our interest, in other words, to find patterns so that we can get rid of extraneous data and use our brain for more things. This reward system explains why things like stereotypes (all people are X) or religion (everything happens because of X) feel good; it also explains why we’re drawn to symbolism, metaphor, and succinctness.”

Bandwidth: The amount of information a work or medium can host. Novels, movies, plays, operas, all have high bandwidths. Individual songs, poems, and paintings have lower bandwidths.

Chords: Aesthetic fit. “Chords are elements combined in a way that is appealing, but not because the combination describes reality. Chords exploit the many evolved sweet spots of the senses… Chords can be consonant or dissonant — the sum of their parts can elicit pleasure or irritation, or even revulsion. Chocolate and peanut butter fit better than chocolate and ketchup.”

Maps: Abstract fit. “Maps describe what exists. They exploit the evolved need to understand how reality behaves. They can be aesthetically pleasing but they put the task of abstraction first. Maps “fit” when they achieve compression — when they eliminate redundancies in a pattern of real-world relationships without sacrificing essential features. Poor map fit is usually due to bad compression (irrelevant information that feels like fat on the bone) or outright misrepresentation (features that don’t appear in the abstracted territory). The Wire fits better than unedited surveillance camera footage or CSI: Miami.”

See also: Sarah Perry’s “Gardens Need Walls: On Boundaries, Ritual, and Beauty”

* conversation has been re-ordered and condensed from the original (compressed even, for optimal aesthetic and communicative value) 

SR: How did the community surrounding cultural criticism sites like Carcinisation, Sublemon, and Lipoblog form? What are its ties to rationalist publications like Ribbonfarm and Slate Star Codex?

Duquette: I moved from LessWrong to Twitter in 2012 and started talking to Sarah [Perry] and Rev and Misha and a few others. Carcinisation is pretty dead; Sam Burnstein started it in the hope that we’d all blog more, but we were too flaky (with the exception of Sarah obviously). I don’t really read SSC but I respect Scott a lot. Same with Ribbonfarm and Venkat.

Sublemon: Gabe and I met in 2013 and he introduced me to his Twitter friends. I mostly started Sublemon to “find a voice” outside of academic styles of writings, and it’s deliberately quite casual as a result. Carcinisation feels like it was a stepping stone before Sarah (and then I) began writing for Ribbonfarm.

Duquette: The shared priority between postrationalism and rationalism is caring about reductionism. Being reductive can be good, but it can also serve as a beacon to handwringers for useless nuance or, rather, not-as-useful nuance.

I suppose if I gained anything from rationalist writings, it’s a keener sense of when words aren’t pointing to a thing in reality, or aren’t pointing to that thing as well as they could. Watching for circular definitions, unfalsifiability, etc. (“art is the pursuit of beauty,” or “art is the ineffable” being examples).

SR: “Art is the ineffable” being similar to the idea of aesthetic autonomy? The idea that there is something special about art inherently as art; that it cannot be reduced into social and economic factors.

Duquette: This is where the rationalist side kicks in and I have to ask what “inherently” means. That seems like a religious exemption.

Sublemon: Right, “inherent” sounds like a magical word to me. Art might be an unusual behavior, but it’s reducible in various ways. And not just social or economic ones.

Duquette: If we all agree that there’s no supernatural component, then aesthetic experience is a thing that happens in human brains.

SR: Pivoting here temporarily — There seems to be a lot of value in bringing well-developed outside perspectives to a subject like art, where the conversation is very much dominated by one particular approach (that of academia and the Ivory Tower). There are frequent accusations made against academia and art history departments of insularity, incomprehensibility, or widespread, unconscious biases which limit the potential for insight. Besides merely being a secondary and outsider perspective, are there specific things you think the current art criticism mode lacks or could gain from a rationalist approach (and vice-versa)? Do you ever feel the dangers of insularity in rationalist or postrationalist communities? I’m thinking a bit of Sarah Perry’s piece on semi-permeable boundaries, “Gardens Need Walls.”

Duquette: It might be simpler to ask what in academic art discourse is worth preserving, and what specifically are the dangers of insularity.

SR: Well, you have the Galápagos effect, where insularity and isolation can lead to interesting, original evolutions of ideas. That’s the positive side. You also can lose grounding or reference points — there’s no stabilizer, no outside force to “check” you. Without someone questioning macro-level priors, practices, and assumptions (because everyone in the community, via self-selection, agrees on them), it’s easy to imagine the ship steering unknowingly off-course.

Sublemon: With jargon at least, there’s both value and danger. Words take on conceptual loads that allow for increased complexity and specificity; the tradeoff is accessibility and clarity (to others and to oneself). I think the main difference between Gabe’s and my approach (vs. the academic approach) is that our interest in criticism and philosophy is mostly practical. We both enjoy art that works.

SR: Practical to what ends? Works or is successful at doing what? Anything? Are there priorities in “ends” or “effects” that an art can do successfully/work well at doing?

Duquette: With Liposuction, I’m currently trying to build a new toolkit for people to identify problems in things they experience, so that feedback between makers and audiences can be more fruitful. As for what art can or should do, I think there are very very many valuable “engineering targets” or “ends” as it were. Probably not infinite but.

SR: It seems like contemporary artworld thought is primarily oriented to using art as a basis for thinking about interesting ideas — that the critical role is complementary, one of almost completing the larger artistic/philosophical/creative process. This is as opposed to “getting at the essence” of something, or trying to reverse-engineer how something works.

Duquette: I see art as basically being about experiences and communication. I’d like to see a greater variety of experiences and high-fidelity communication, all accessible to everyone, and reverse-engineering successful art to find out what makes it “work” is essential to that project. Serious engineering isn’t, say, sitting around discussing the original Wright Brothers’ biplane Flyer I (as art historians and academics do with art) — it’s building modern planes that people can safely fly across oceans in.

SR: But the goal of engineering is to make planes fly faster, while the goal of art is just to explore a set of ideas through form.

Duquette: “Exploring ideas through form” is a goal that one can be better or worse at optimizing for/achieving.

Sublemon: It’s great to know the history of how things have been engineered before you start engineering something else, so they’re not completely incompatible pursuits, but definitely fundamentally different ones. I will say, though, that the only significant way I would change contemporary art criticism would be getting it to be more explicit about the rubric by which it’s judging. If criticism cared about only one thing (be it engineering or history etc) it would be boring. I’d just like to see [critics] articulate what it is that they do care about, and why.

SR: So, reverse-engineer what successful art is doing, find trends between works, identify those common traits that make it successful and effective, etc., and all in the service of promoting “better art” in the future —  “better” in the sense of artists having a better toolkit of techniques that allow them to achieve their intention?

Sublemon: Yes, that sounds about right, though “better” also in the sense of artists having a better idea of what their intentions even are.

Duquette: A successful end result would look something like more, and more diverse, Coca-Cola formulas or KFC original recipes. Unlike academic art, mass experiences have to work; they’re meticulously engineered accordingly. Precise emotional rendering in movies is another example, especially when it’s very hard to articulate the emotion (and even better when the rendering is done with great economy).

SR: And building a toolkit of concepts and ways to discuss art ostensibly helps not just artists but critics and audiences as well.

SR: Out of curiosity, does postrationalism contain any economic critique of art?

SR: To expand on that — PostRat/Rat seems very interested in social currency and motivation, but economic currency and motivation and its role in things seems absent from art critique. Is this just a matter of, “we can only devote our energy to so many areas of inquiry,” or is it a matter of believing financial value doesn’t incentive or affect artistic creation as much as social value? Also, Gabe, you’ve used examples corporate success eg. Coke/KFC/Chipotle. Isn’t there a degree to which marketing and money and financial incentives (cheap cost of fast food) are playing a larger role in incentivizing consumption than the actual aesthetic “success” of the food?

Sublemon: If you’re referring to people in the postrationalist corner of the Internet in general, and not postrationalists who write about art specifically, then there are a lot who are interested in economics. I don’t want to claim anyone belongs to some kind of “identity” of postrationalism against their will, and I’ll rescind the association if they object, but I know Matt Simpson, Byrne Hobart, and obviously Venkatesh [Rao], for example, are all economics or business people. Purely by virtue of overlap with the tech world it seems there are a lot of people in the community who know or talk about business.

For me in particular, economics and how economics affect art isn’t a primary area of interest (though not because there’s nothing to be interested in there). It may well be that overhauling how studios decide what to greenlight, or figuring out how to get resources (time, money) into the hands of the talented-but-disadvantaged, would be a better way of getting more good art into the public eye than writing tetchy blog posts. In fact, that’s a pretty likely possibility which I may have to engage with at some point. In the meantime, I don’t have the authority to write about it.

SR: I’m curious your thoughts on a utilitarian or consequentialist art criticism, in which ideas about art and criticism maximizing societal good are explored. One might hypothetically believe that contemporary art culture has the best artistic minds working in insular experimental traditions which serve only the purposes of its narrow academic/artworld establishment; that the equivalent of a Sistine Chapel — and the effect of beauty and religious awe which the Chapel inspires, which is a net good for humanity — is much less likely to be made today because of this. One might, moreover, find this to be socially detrimental. Therefore the net “effects” of an art culture should be considered in assessing its value and whether it’s headed in a “good” or “bad” direction. This doesn’t seem too far off from the beliefs you and Perry might hold, though I don’t mean to misrepresent your views.

Sublemon: I’m pretty hedonistic when it comes to art. If it makes people feel good then I’m glad it exists. I disagree with people about quality but I would never want someone to feel bad for liking what they like.

As far as social good goes, I consider well-made things to be a social good. I consider truthful things to be a social good. Conveniently, I also consider truth and functionality to be aesthetically good. A beautiful building or a great computer usually improves the lives of those who interact with it. Similarly, art that is used for political action is flawed when it lies, and is good (or at least, not-flawed) when it tells the truth. Good both aesthetically and socially.

SR: Isn’t the implication here, though, that artistic value judgments are dependent on the ideological leaning of the critic (given that the non-political side of the artworks)? Even two reasonable, open-minded viewers have very different ideas about what is politically/morally true or false. For example, a piece of B.L.M. performance art commenting on police brutality might seem true (and therefore good) to many progressives, but would appear false (and therefore bad) to most conservatives, or even to those who believe (as say, Scott does re: police shootings) that the issue is general structural poverty and not on-the-ground policing methods. Is there a way to rise above ideological belief to evaluate the goodness of political art?

(The alternative, I suppose, is that by “truth” versus “lie,” you just mean whether a person is genuine in their expression, but there are plenty of genuine personal expressions which don’t improve people’s lives or lead to social good.)

Duquette: I evaluate it on the basis of how many points of view it’s able to integrate. If an artist chooses to work in a medium that disincentivizes the integration of many POVs, I more or less reflexively disregard it. It’s important to remember that one needn’t reckon with things just because they exist. [note: see Bakhtin’s dialogic vs. monologic, cf Haley’s concept of artistic bandwidth]

SR: I feel like a lot of art, and especially literature, has two sensibilities or artistic priorities that tend to war with one another. There’s subtlety, which is definitely held as superior to being on-the-nose. And then there’s clarity: if the author can’t communicate their basic ideas in language, they’re doing a bad job of communicating. Contemporary literary praise of subtlety seems similar to praising a skillful compression. Maybe there’s a difference here, and it’s worth pausing to address that. I’m sure social signals and context play a role too, since the ability to “get it” or understand subtle communication makes you “in on the joke.”

Sublemon: Subtlety to me basically means “more information.” Economy means “with less.” Clarity means “all the information was received.”

SR: Good compression is packing in as much information with as little material in a way that is still received by the audience.

Duquette: And non-obvious signals can be economical and maximally evocative when chosen well. The skill of a world-class journalist is the skill of noticing every detail or idiosyncrasy and then knowing which handful to pick for inclusion in the piece. These details being the tips of the iceberg that make it easiest to infer the shape of the submerged part and the whole (instead of tediously describing the entirety).

SR: There’s a nifty parallel here — I’m sure you’re aware of this already but — where the concept of compression is itself good compression (boiling a concept down to a concept-handle).

SR: Okay, so here social context comes into play. Social contexts create shared understandings. Compression takes advantages of shared understandings; you’re guessing which details or iceberg tips will evoke the whole for a given audience of readers. So good compression for one audience can be terrible compression for another, if the audience is mismatched to the artist.

eg language: every word I’m writing here could be explained and defined, but we have the shared knowledge base of the English language and therefore we can compress communication massively. Obviously this changes when I interact with a non-English speaker.

Sublemon: I have a post on symbolism for Lipoblog that I’ve been sitting on for months, which talks about just that.

SR: Well, I’m curious then — cultural fragmentation is a very real phenomenon. People are dividing up into subcultures and everyone’s shared knowledge base/set of references is quickly splitting up. This makes art production for large audiences much harder.

Sublemon: Right. Though I want to move away from the line of thinking that, since communication requires shared contexts, you shouldn’t even try to transcend subcultures.

SR: Okay, so let’s assume communication is still possible across subcultures with enough effort, using a combination of intuitive references with whatever shared knowledge bases still exist (there are currently many, who knows how many there will be). If we see good art as being subtle versus on-the-nose, or compressing well — where economy is seen as graceful and beautiful in the same way an efficient basketball play is beautiful — then even if it’s possible for art to communicate across many subcultures, is it possible for it to be seen as beautiful or good art?

Duquette: You’re making that sound far-fetched to accomplish. I’m not sure why.

Sublemon: Why wouldn’t it be possible?

SR: Well, for an example, Lena Dunham’s Girls — within the subculture that orbits this sensibility and lifestyle, viewers know that Dunham is parodying herself (and Greater Brooklyn). There is good, clear communication happening between subculture and subcultural artist, and Dunham doesn’t have to over-satirize or put huge scare quotes around every scene in order to convey tone and meaning. But to, say, conservative Iowan viewers, Girls doesn’t come across as a parody — they assume Dunham isn’t self-aware, that she really stands for the things her character does in the show, and there’s been a massive pushback against Girls by people who misread it. There’s a communication breakdown. Now, if Dunham were to put huge scare quotes and satire disclaimers all over Girls, all viewers in America would know it’s a parody. But it wouldn’t be “artful” or graceful or beautiful in the way that it currently is. It would be bad satire. Too obvious.

Sublemon: To be fair I think a lot of people have difficulty judging the object/content of Dunham’s satire.

SR: Well, maybe another, simpler example is the use of the sarcasm tag “/s” on Internet forums. Because it’s hard to know, especially with highly populated forums/threads, whether everyone reading your post is going to share your same priors/perspectives, you have to mark it clearly as sarcasm tag. But when you do this, even though everyone is on the same page about your post’s tone and intent, the joke has been thoroughly killed; there’s no longer an artful ingenuity, subtlety, or humor.

Duquette: The ability to detect subtlety is not evenly distributed. Also, people tend to use cultural goods socially foremost. So it’s partly about which “team” you’re on, not whether the thing works in other ways.

Sublemon: It’s true, some things will only seem beautiful to people with a niche, shared context. Though there are other kinds of shared context, so I don’t see why beauty wouldn’t be perfectly achievable in those other contexts.

Duquette: I’m personally ambivalent about the atomization of culture. It’s a brain drain on the mainstream and it shows. But it’s ultimately a better use of scarce resources.

[see “The Melancholy of Subculture Society”  — Gabe]

SR: That’s interesting. Better use of scarce resources how?

Duquette: People can get more bang for their buck, belonging-wise, since there are more communities and therefore more community roles. The absolute amount of artistic variety also increases, which slightly increases the absolute amount of high-quality art. But it mostly flattens everything into “pretty good I guess,” which means less low-quality art (unless you count libraries full of fanfic or whatever). Basically, nobody ever has to put up with anything they don’t like — ever. But people are often wrong about their own preferences, so the peaks get cut off.

Sublemon: Regarding resources, there’s the resource of how much work you have to do to be understood in the compression way, and how much work you have to do to be accepted socially. As far as Gabe’s comment on preferences, I don’t know if it’s so much that people are wrong about their preferences as it is they’re bad at imagining new things that will fall within their preferences (because being good at that requires, effectively, artistic talent).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at

%d bloggers like this: