Wine in front of me

by Collin Lysford

A perennial tis.so discussion topic is anti-inductivity, situations where observing a pattern makes it less likely to occur in the future. Blueberry-picking is inductive: when you’re looking for blueberries and you find some, you’ll update your views on “is this a good place to look blueberries?” to be more positive. But when you go home and see that same location has gone viral on SuperAwesomeBlueberryLocations.biz.site, you don’t think “That’s a second piece of evidence this is a good place for blueberries! I’ll update even more positively!” No, you think “Shit, it’s now common knowledge there are blueberries here; I bet it’ll be picked clean by the next time I visit.”

Social deduction games like Mafia/Werewolf are one of the purest examples of an anti-inductive game. Any time you figure out a tell to distinguish a mafioso from a truth-telling townsperson and use it to get the mafioso killed, it creates future opportunities for the bad guys to trick you knowing that you view the world through that lens. An ostensible signal that’s been completely captured by these anti-inductive dynamics is called “WIFOM”, short for “Wine in front of me” after the classic scene from the Princess Bride:

Westley:

“All right: where is the poison? The battle of wits has begun. It ends when you decide and we both drink, and find out who is right and who is dead.”

Vizzini:

“But it’s so simple. All I have to do is divine from what I know of you. Are you the sort of man who would put the poison into his own goblet, or his enemy’s?”

WIFOM arises when you have a clearly imbalanced choice presented by an agent who can easily choose between multiple options and each side understands that the other side is heavily scrutinizing their choice. People usually drink from the cup in front of them, so that’s the obvious choice for a poisoner to pick. But since you’re aware that people usually drink from cups in front of them, obviously you’d want to avoid the obvious choice. But the poisoner also knows you’d want to avoid the obvious conclusion that people drink from the cup in front of them, and could have put the poison in your glass for that reason…etc.

Now let’s extend this example to Mafia. Westley is the one presenting Vizzini with the hidden information game, so he’s going to be our mafioso against townsperson Vizzini, who has no hidden information. Suppose Vizzini presented a compelling case on Westley and he’s slated for execution. Westley knows that the whole town is about to learn he’s a member of the mafia when he flips.¹ Right before the hammer falls, he comments: “Fair play. You got me, Vizzini.” Read literally, this statement gives the non-Vizzini townspeople extremely important information: Vizzini was not one of Westley’s buddies bussing² him for cred, but an actual townsperson. But of course, Westley knows that people will interpret it this way, and if Vizzini is actually his buddy, that false clear will give him a free pass to win the game. So maybe some townspeople think: Vizzini must be Westley’s buddy for him to make a comment like that! Let’s kill him next! But Westley also knows you’d want to avoid that obvious conclusion that those wry comments are about townspeople, and so could have said it about a townsperson to get them killed next…etc.

This is classic WIFOM, and like the titular example, it regresses without end. As such, when a mafioso make a seeming-slip in a situation where their death is assured, experienced players will usually say “that’s just WIFOM” and ignore it. Yes, technically speaking Vizzini is or isn’t Westley’s buddy, but trying to develop a good general heuristic of which to believe given “You got me, Vizzini” will see that heuristic instantly gamed. But wait — isn’t the whole enterprise of playing Mafia all a bunch of WIFOM, then? Won’t anything that gets used to catch Mafia players be instantly subject to this unproductive infinite regress? If WIFOM is the stuff that experienced players ignore in an anti-inductive game, then what don’t they ignore?

One tell I have a lot of faith in is something I call “caught for the wrong reasons”. Let’s suppose Westley’s alive and not under particularly much suspicion. He makes a comment like “Nice going, killer”, and Vizzini remarks: “Killer! You must be subconsciously thinking about the murder you committed last night, you dastardly mafioso!” Westley actually says killer all the time and it has no bearing on alignment, but no matter how much he tries to point this out, Vizzini keeps doubling down. Westley calls him an an idiot, but the other townspeople start taking Vizzini’s side, and soon Westley is ranting about how stupid EVERYONE is—

—Completely forgetting to pretend that he thinks some of his accusers could be mafia arguing in bad faith. Westley was so annoyed that Vizzini’s stupid theory would be “confirmed” by his flip that he vented completely honest frustration without mixing in fake suspicion. Good townspeople will wordlessly coordinate on “believing” stupid theories if they see glimmers of this dynamic to try to get a telling reaction. Let’s say they pull it off, and Westley’s about to be executed. Right before the hammer falls, he makes a wry little comment: “Genius work, Vizzini. You really know the magic words only bad guys say.”

Superficially this may look like it’s governed by the same WIFOM dynamics as last time: it’s confirming Vizzini as a townsperson, but Westley knows it looks like that, and so on. But I would treat this as strong (though not definitive) evidence that Vizzini is a townsperson. Why? Because we’re specifically in a “caught for the wrong reasons” situation, and the whole reason we pegged Westley as a mafioso is that he’s letting earnest frustration dictate his posting. If Westley was a more experienced player who knew about these dynamics, I’d be less inclined to take his last words as evidence towards Vizzini’s alignment. But also, if Westley had parceled out “caught for the wrong reasons” as a concept to the point where he could be anti-inductive in that specific frame, he wouldn’t be in this situation in the first place.

WIFOM dynamics require the person offering the choice to be able to replicate the other person’s internal monologue, which means they need a similar vocabulary, which means they need a similar ontology. This point is often missed by people studying game theory, since the formal rules of the game are assumed to be the ontology the players share. The rules are certainly part of the player’s ontologies, and we can see that a nominally anti-inductive game can collapse if one player doesn’t understand the rules, or if they convince the other player they don’t. But while games have ontologies “built in” owing to the verbs of interaction, experienced players will extend their ontologies beyond that to conceptualize higher-level patterns. This ontological asymmetry is what stops anti-inductive games from collapsing into unusable WIFOM. Two players may be looking at the same formal game state and the same stream of public information, but if only Vizzini resolves it into the state of “caught for the wrong reasons”, Westley can’t recreate Vizzini’s internal monologue in order to subvert it. Vizzini has representational privilege over Westley.

When you hear about the back-and-forth of “information” in an anti-inductive game, always remember: agents don’t act on information, they act on perceived states generated by motivated ways of seeing they develop over time. That development is oftentimes what we mean by “experience”, an advantage difficult to gain but also difficult to fool with WIFOM once gained.

strategic interaction games ontology generalized reading anti-inductivity strategy game theory deception rules representational privilege mafia mini-games concepts WIFOM

“Weaponized” words are imbalanced decision rules

by Hazard

Many words functions as labels for a decision rule, an “if X then Y” heuristic, a descriptive/prescriptive blend for acting in the world. To apply such a label to something is to simultaneously assert that it fits some descriptive criteria, and additionally that you think it should be treated a certain way. A lot of legal terminology is explicitly created to serve as decision rules; “this is what counts as a misdemeanor and this is how we treat people judged guilty of misdemeanors.”

Any given decision rule can be more or less useful for a given context, and more or less “imbalanced”; more heavy on the descriptive side or on the prescriptive side.

From C. S. Lewis in “The Death of Words”:

But surely there are words that have become merely complimentary—words which once had a definable sense and which are now nothing more than noises of vague approval? The clearest example is the word gentleman. This was once (like villain) a term which defined a social and heraldic fact. The question whether Snooks was a gentleman was almost as soluble as the question whether he was a barrister or a Master of Arts. […] This is one of the ways in which words die. A skilful doctor of words will pronounce the disease to be mortal at that moment when the word in question begins to harbour the adjectival parasites real or true. As long as gentleman has a clear meaning, it is enough to say that So-and-So is a gentleman. When we begin saying that he is a ‘real gentleman’ or ‘a true gentleman’ or ‘a gentleman in the truest sense’ we may be sure that the word has not long to live.

To be more… descriptive, than prescriptive, I’d say what’s happening is the word is becoming more and more of a prescriptive label. This is not always and everywhere a bad thing. “Guilty” is a word that’s almost entirely prescriptive. It’s what you use for a conclusion. The word “guilty” does not encode the process of determining guilt. That task is offloaded to the whole complex mess that is the legal system (or however conflict resolution works in your local scope). It’s useful to have a word like guilty.

You can often tell you’re dealing with a word who’s common usage has mostly come to mean the prescriptive part of a decision rule when you hear things like, “Everyone agrees that X is [value judgment], we just don’t agree on what counts as X.”

I’d say that when someone calls a word “weaponized”, they’re getting at how the prescriptive part of the decision rule has grown in severity while the descriptive part lags behind, not growing in sophistication to match the intensity of the new prescription. Perhaps the descriptive aspect is still anchored on dictionary definitions, ones that are decent for gesturing vaguely at a cluster, but don’t really cut it (according to the person making claims of “weaponization”) for the task of mediating conflict and deciding when a severe consequence should be dished out.

Imbalanced decision rules, ones with general or vague descriptions but intense, precise prescriptions make room for a maneuver I really don’t like, arguing definitions when you’re really trying to argue how to make a decision. You argue that “by definition, X is a Y, it’s pretty clear cut, I don’t see what there is to debate here” when your full logic is “by definition, X is a Y, and we’ve already decided that we treat Y like Z, so let’s Z”. The real underlying argument is about “should we use prescription Z for dealing with X?” but that real conflict is evaded in favor is trying to “short-circuit” the decision via definitional arguments.

C. S. Lewis decision rules meaning weaponized language

Euphemism treadmills and “Communication finds a way”

by Hazard

The word “stupid” used to more or less be a medical term. In some people’s living memory the word “retarded” was also a medical term, before it became the word that I and other kids used to make fun of each other in elementary school. A while back Stephen Pinker coined the phase “the euphemism treadmill” to point at the process of emotionally charged language getting replaced with more “neutral” language, and in time the new “neutral” language gaining the same emotional charge as the old language. A similar thing often happens over time for the language used to refer to different racial and ethnic groups. Think “negro” -> “black” -> “african american” -> “person of color”.

I mostly hear the term “euphemism treadmill” used to talk about something that happens to slurs. C. S. Lewis noted a more general version of this process in his essay, “The Death of Words”¹:

The truth is not simply that words originally innocent tend to acquire a bad sense. The vocabulary of flattery and insult is continually enlarged at the expense of the vocabulary of definition. As old horses go to the knacker’s yard, or old ships to the breakers, so words in their last decay go to swell the enormous list of synonyms for good and bad. And as long as most people are more anxious to express their likes and dislikes than to describe facts, this must remain a universal truth about language.

Lewis and I might argue about the finer points in his essay, so I’m going to reformulate is observation into my own version I can stand behind:

People will use the language they have available to express what they care about expressing², regardless of the original intent behind the language.

If people want to insult each other, words will become insults. If people want to say things are good and bad, words will become synonyms for good and bad. If people want weapons, words will get weaponized.

Underlying this “law” is a slightly more nuanced version of “meaning is use”. Communication has a lot of built in redundancies, and the context (verbal, extra-verbal, and non-verbal) surrounding any individual word informs the intended meaning of the word in addition to and separate from the meaning you might derive from the word in isolation. If enough people use a non-insulting word in an “insult slot” in their speech, eventually the underlying shared interpretive equilibrium will shift such that that word can be received as an insult without needing corroborating support from the surrounding context.

Stephen Pinker C. S. Lewis communication language meaning euphemism treadmill

When a muskrat’s a fish

by Feast of Assumption

On fast days in the Catholic liturgical calendar, the eating of meat is avoided. There’s been plenty of writing across the centuries¹ about this being no biggie for those on fine economic footing, but somewhat more of a hardship for those of meager means.

In Michigan among French-descended fur trappers, a missionary in the early 1800s noticed his parishioners were faring poorly, and allowed them to eat muskrat (after all, it lives in water!). This post gives a detailed saga, culminating with a modern “hey waaaaait, muskrats aren’t fish, shouldn’t we make the Michiganders knock this off?,” which was answered by the late Bishop Kenneth Povish of Lansing in 1987:

He referred to what he called the “Great Interdiocesan Doctrinal Debate” of 1956, during which he determined that although muskrat is a warm-blooded mammal and technically flesh, the custom had been so long held along Michigan’s rivers and marshes that it was “immemorial custom,” thus allowed under church law.

For the record, Bishop Povish didn’t much care for muskrat as a meal. He wrote that “anyone who could eat muskrat was doing penance worthy of the greatest of the saints.”

The post linked above gives a complete explanation of the dispensation and its history, but I found that it faceplanted on the description of the taste of muskrat. It tastes like a duck if a duck had no fat–it is quite lean, and its flesh tastes like algae. The best perparation I’ve found is in the vein of sauerbraten, where you give the muskrat an overnight sweet vinegar-based marinade. Once it’s been well marinated, muskrat tastes fine enough, but the bones:meat ratio is tiresome, and I don’t know a fix for that.

Family lore has it that (cicra 1930), Ojibwe coworkers would brag about having a muskrat available for dinner that week, but white coworkers wouldn’t.

muskrat categories made for man Catholicism ontology diet habitus Voltaire

Brains in your feet

by Collin Lysford

I.

Terry Prachett’s Night Watch is probably the single book I’ve re-read cover to cover the most times.¹ It’s short enough for one sitting and as fizzy and easy to read as almost all of Prachett’s work², but it’s telling a much more intimate and meaningful story about human suffering and time.³

A quick summary: His Grace the Duke Sir Samuel Vimes is chasing a murderous bastard named Carcer when they’re flung into a magic storm that sends them back in time to when Sam was just Sam, a snot-nosed kid freshly recruited to the Night Watch. Young Sam got his lesson in how the world works from one Sgt. Keel, a veteran copper who teaches the impressionable kid lessons about the world he’d never learn from the rest of the mediocre dregs in the Watch. But history is changed and Carcer kills Keel before he can teach anything. So Vimes (with the aid of the History Monks trying to set the timeline straight) must take on the role of Keel and teach Young Sam what the proper Keel had taught him, to make sure that this version of himself also grows up decent.

Overshadowing all of this is the Glorious 25th of May, a revolution that Vimes knows will claim innocent lives for only a few hours of freedom. He sees it brewing again and he knows how it went the first time. Vimes wants to change the future as little as possible to make sure there’s a future to come back to when the magic snaps back; but would Vimes be Vimes if he let it happen again while holding these memories of how it went before?

II.

In Memories are environmental indices, I make the point that memories aren’t crystalline collections of meaning itself, but pointers to environmental configurations where meaning is stored. This idea comes up continually throughout Night Watch:

“You been here before, Sarge?” said Sam as they turned a corner.

”Oh, everyone’s visited Ankh-Morpork, lad,” said Vimes jovially.

”Only we’re doing the Elm Street beat perfectly, Sarge, and I’ve been letting you lead the way.”

Damn. That was the kind of trouble your feet could get you into. A wizard once told Vimes that there were monsters up near the Hub that were so big that had to have extra brains in their legs, ’cos they were too far away for one brain to think fast enough. And a beat copper grew brains in his feet, he really did.

As Duke and Commander of the Night Watch, Vimes is a manager, not a beat copper. He does this job well if not spectacularly. But all of Vime’s clarity comes from his time as a beat copper; from the cobblestones themselves. And while His Grace needs to wear high-quality, durable boots, Sergeant-at-Arms John Keel’s first priority is sourcing boots thin enough he can feel the ground beneath him:⁴

It was good to feel the streets with dry feet again. And after a lifetime of walking them, he did feel the streets. There were the cobblestones: catheads, trollheads, loaves, short and long setts, rounders, Morkpork Sixes, and the eighty-seven types of paving brick, and the fourteen types of stone slab, and the twelve types of stone never intended for street slabs but which had got used anyway and had their own patterns of wear, and the rubbles, and the gravels, and the repairs, and the thirteen different types of cellar covers, and twenty types of drain lids—

He bounced a little, like a man testing the hardness of something. “Elm Street”, he said. He bounced again. “Junction with Twinkle. Yeah.”

He was back.

Vimes brings knowledge from the future into a past where it can be used constructively. But the knowledge didn’t come from the far future. There wasn’t a grand Theory of Coppering that was studiously developed and then brought back to enlighten the primitive people of the past. His knowledge of the eighty-seven types of paving brick only helped because the city was still made out of those bricks. Vimes brought back his indices, and thank goodness he didn’t go back so far — they were still pointing at something.

III.

Dr. Stone is another series about taking highly developed memories to a time that needs them, though it lacks Prachett’s deftness and emotional heft. In Dr. Stone, some mysterious force petrifies all of the humans in the world for thousands of years. The world regresses to the stone age, with children who were born post-petrification never knowing about the old world. But boy genius Senku remembers all of human scientific progress, and is determined to speedrun civilization back into being. The story feels like blitzing through a video game tech tree, an aesthetic Dr. Stone explicitly cultivates:

What it lacks are the eighty-seven types of pavement brick, the concrete understanding that bridges the gap between copper as a concept and copper in it’s use. Dr. Stone falls into the Rick and Morty trap where abstract “intelligence”, whatever that means, is trusted to be able to stand in for any manipulation of the environment. This just isn’t how human memory works, and Rick and Senku are ultimately larger-than-life comedy characters, a stupid person’s conception of what a smart person is.

But there’s one scene in Dr. Stone that’s worth highlighting⁵. Senku meets Chrome, a post-petrification child who’s an obsessive tinkerer. He has Senku’s drive, but not Senku’s knowledge. As Senku describes the old world, Chrome starts crying:

Senku: Why are you bawling all of a sudden? What’s up with your emotions?

Chrome: I’m not crying! Wait, damn it, I am! Who caused the petrification? If ever see them, I’ll kill them! How could we lose this? Our human predecessors spent millions of years slowly building this crazy, technological civilization, and it got wiped out in the blink of an eye?! I’m pissed as hell!

Senku: It’s not wiped out, idiot, The whole thing’s right here….two million years of human history is stored right here inside me.

Dr. Stone is right to give us Chrome’s awe of the chain of human knowledge and anger at it’s severing. It’s wrong to give us the idea of Senku, one bookish nerd that can somehow store two million years of history. As Feast of Asusmption points out in Monograph or aether, much of human understanding is transmitted via folklore while it’s relevant and lost when it isn’t. Our feet have brains, but not pens to write with.

And in Night Watch, the revolution fails again; the same graves are filled. Vimes was willing to throw his future away to save the people in this past, but he just wasn’t capable: his knowledge isn’t the sort of thing that saves revolutions and rebuilds civilizations. His wisdom is only good for putting a young copper on the right path and letting his lads die with their heads held high. Prachett was too smart to believe in nerds, and Vimes is too savvy to want this part of history monographed:

“They did the job they didn’t have to do, and they died doing it, and you can’t give them anything. Do you understand? They fought for those who’d been abandoned, they fought for one another, and they were betrayed. Men like them always are. What good would a statue be? It’d just inspire new fools to believe they’re going to be heroes. They wouldn’t want that. Just let them be. Forever.”

There’s a certain sort of knowledge that Prachett understood was too important to be lost to aether. But it’s not eternal knowledge left on a statue for nerds like Senku. Your feet still need to touch the stones.

memory environmental cognition literature Terry Prachett knowledge logistics

Glass

by Possible Modernist

Several of us recently opened up a long back and forth about the nature of truth, and ontological questions more generally. A key part of this conversation was unpacking the pragmatic perspective on meaning.

A nice example that occurred to me later is glass (the material) — specifically the question of what type of thing glass is. I can’t remember exactly where or when, but I’m sure I’ve heard various people claim that glass is in fact a liquid, not a solid. This seems to me like a classic example of an attempted scientific categorization that would defy common sense.

When we think of glass that we are familiar with, we probably think first of its transparency (and tangentially, I wonder if this part of the explanation for what makes people sympathetic to the liquid claim, given that liquids are generally more likely to be transparent than solids), but right after that we likely think of solidity, and brittleness, which are generally antithetical to things we classify as liquids. The claim, however, is that glass actually flows like a liquid — it just flows so slowly that we can’t notice it, or direct it, the way we can with typical liquids.

I’ll revisit this claim in a second, but we should note first that the claim itself is not completely implausible on its face. After all, people have likely encountered substances like molasses, which certainly flows, (it can be poured), and yet is much more viscous than, for example, water. As such, it’s not hard to imagine cranking that knob up much higher, to get a liquid that is so viscous that it barely flows at all; eventually, however, it would take on the shape of whatever container it is in (perhaps over thousands of years).

In addition, we know glass is clearly highly malleable under the right conditions, specifically when it is at very high temperatures; indeed, that is how it can be made to take on such varied forms during manufacturing. So again, if we assume that viscosity decreases with temperature, it could be plausible that viscosity is just very high at room temperature — too high to allow for easy manipulation, but still subtly present.

In the case of glass, it seems like a key piece of evidence that people point to is the fact that very old window panes appear to show a history of flowing over time. In particular, many of them are thicker at the bottom than the top, or contain stream-like patterns. According to Wikipedia, however, there is no solid (sorry!) evidence for this; rather, it seems more likely that such panes were simply manufactured that way, perhaps intentionally or perhaps due to limitations of older manufacturing processes.

Still, we are left with the question of what exactly is a liquid? Pragmatically it seems like a very intuitive category, but as always it is tough to define clear boundaries. We can try to make a definition that is more scientific precise in various ways, which will likely involve describing aspects of material that cannot be observed independent of technology, such as molecular composition. In the end, however, we are faced with the reality that “liquid” and “solid” are not fundamental, pre-existing categories. Rather, they are concepts that humans have come up with (along with “gas”) to describe common, observable, categories — broad basins of attractions that most things we encounter seem to fall into.

If we do make use of some more technical observations, we find that at an atomic level, most things we call liquids have loosely attached molecules that slosh around, allowing them to flow. Most solids, by contrast, have tightly packed molecules, generally in a crystalline lattice, as is the case in metals, minerals, ceramics, ice, etc. It turns out, however, that glass defies this categorization scheme. Glass is rigid and brittle like most solids, but it does not have the lattice structure that is found in most other solids.

As such, various finer distinctions have to be made. Many sources, including wikipedia, now categorize glass as an “amorphous solid”, as opposed to “crystalline solids”. Others have argued that glass should be called a “frozen liquid”. Which is correct? (And how should we arrange the hierarchy of categories?) Well, there may be arguments to be made one way or the other, but again we fool ourselves if we think there is only one, fundamental, scientifically meaningful set of categories that will apply everywhere and always, rather than multiple possibilities which are all compatible with the evidence.

Ironically, if we are left with “flowing” as a defining characteristic of liquids, we run into some more weird conflicts with intuition. It turns out that there is a category in geology called “rheids” which (if I understand this correctly) seems to describe solids — specifically rocks — that flow under pressure. To make matters even worse, this seems to be an absurdly clear case of a category that is defined by a purely arbitrary boundary. According to Wikipedia, “In geology, a rheid /ˈriːɪd/ is a substance whose temperature is below the melting point and whose deformation by viscous flow during the time of observation is at least three orders of magnitude (1,000×) greater than the elastic deformation under the given conditions.” Why three orders of magnitude? No doubt there is some motivation, but I declare it to be arbitrary. Someone chose that number, and it has stuck.

Further down on the page we learn that granite can be categorized as a rheid, meaning that it (like glass?) does flow very slowly over time, seemingly mocking the expression “solid as a rock”.

No doubt I am being somewhat cavalier in talking about many of these concepts, and perhaps experts in any given domain could find ways of aligning their usages such that everything would have a nicer internal coherence. The point, however, is to illustrate the kinds of difficulties we run into when trying to come up with a rigorous and defensible way of carving up the world. Any categorization scheme is like a kind of bet against the universe, daring it to come up with an exception, which it almost certainly will.

concepts categories science pragmatics materials matter

Disavowed Desire leads to Abstracting Desires, capping your ability to get what you want

by Hazard

A model, and an example.

A Model

Sometimes, I have desires. Sometimes, I disavow these desires. Ignore them, tell me self I don’t actually want what I want, feel averse to the fact that I want what I want, or even hide from myself that I want what I want.

Maintaining this separation requires some flavor of self-deception, and a key thing about the nature of self-deception is that it globally reduces my efficacy. Self-deception works by fragmentation and compartmentalization, preventing different parts of my mind from talking to each other, and importantly, not leaving a paper trail through conscious memory. The disavowed parts of me do have all sorts of abilities that “I” don’t, but I’m more or less blocking off being able to pursue any desire with the full strength of my combined body-mind as long as I’m fragmenting around it.

But there’s another way that I get shot in the foot with disavowed desire. Not only am I handicapping my full problem-solving capacity, but depending on the exact nature of the aversion/disavowal, “I don’t want X” has to be “believable” in some sense, to someone. Denying something to myself is almost always entangled with denying something to others. And the better and more reliable I am at getting the thing I want, the more obvious it will be to any one paying attention that I want it. This means that if I have a desire that I’m trying to satisfy while also hiding from myself that I have this desire, I can’t get too good at satisfying the desire, otherwise the whole ruse will unravel.

One way I’ve noticed this manifest in myself is through the “abstracting” of desire. The original desire will arise from some really concrete situation/source, but will quickly be generalized, abstracted, and decontextualized. This laundering of concrete desire through the abstract allows me to move in the direction of satisfying the desire (the abstract version still has some connection to the original disavowed desire), while also no longer being incriminatingly associated with what I’m trying to avoid. Unfortunately abstracted desires really SUCK for guiding my actions when push comes to shove. The more trade-offs and tough decisions I have to make the more I need clarity about the exact shape of my desire, a clarification that said abstracted desires resist in order to maintain plausible deniability.

An Example¹

Sometime in early 2021 I spent several weeks feeling like I really wanted to make a “cool and slick” project in the Tools For Thought space. I spent a few days on and off brainstorming ideas but nothing felt right and things that did feel right stopped feeling right soon after poking into them more.

At some point it hit me, that this whole endeavor was downstream of wanting to impress Conor Sullivan, the founder of Roam. The idea of trying to work at Roam had been floating around in my head a bit. I liked the product, I’d been getting into Clojure recently and knew they were a Clojure shop. And perhaps most importantly, or at least the immediate cause of all this, was the week before I saw someone tweet a demo of something cool they’d made and Conor replied with “You want a job Dan?”

I don’t recall what sparked this realization, but when I poked around at it it was very clear this was the grounding desire for the whole project. So why hadn’t that been clear to me earlier? Why didn’t I see that tweet and go “oh shit! Possible entry point for getting this job I’m pretty interested in! Let’s hit it”

This is the disavowal part. I’ve got a non-trivial amount of aversion to anything like “put effort into impressing a specific person”. There’s some useful instinct there. Too much focus on impressing others can lead to overfitting which makes life suck. But this wasn’t just a cool-headed decision to not putting energy into impressing Conor, it was aversion, and aversion is the starting point of things getting shunted out of awareness. The grounding desire formed after seeing the tweet, it was quickly disavowed, and then transmuted into a more abstract desire (Make a Generally Cool Project), which still pointed me roughly in the direction of satiating the original desire but was way less useful for guiding my actions and making design decisions.

Something I could have done was “okay, so I’m clearly trying to impress this specific person, and have several examples of what impresses him. I also probably won’t be able to summon the energy for the project if I don’t also think it’s cool. So let’s hunt around for the intersection of stuff I dig that I think Conor would also dig.”

But the fact that I had disavowed the “impress Conor and get a job” portion of the desire meant I was stranded in no mans land. I couldn’t just follow my own gut, because I was motivated by Conor, and I couldn’t just double down on Conor, because I was trying to ignore that part of my motivation, and I also couldn’t attempt any synthesis of the two directions.

Just to not end on a cliff hanger, that particular situation ended with me checking in on “do I actually want to try and change jobs right now?” and I got a resounding “Nope”, so I put the whole thing to rest.

self-deception disavowed desire aversion abstraction

Digitalization & reification

by Suspended Reason

Reality is “wave-like, continuous, contextual etc” [Kevin Simler]. And yet we turn the world into a set of nouns which imply global, eternal qualities. What’s going on here?

The post-cybernetic tradition (Bateson, Wilden, some of the Frenchies) would say: reality is analogue, and langauge is digital. “Social magic,” Bourdieu writes, “always manages to produce discontinuity out of continuity.” So we might think of the conversion of a continuous and fluid reality into discrete noun-identities as “digitalization.” Similar camps would refer to the de-indexicalization of local solutions—either by porting them to a new domain, or generalizing them outside their original domain—as “reification.”

Hopefully, these terms can serve as useful tools in beginning to theorize functional pragmatism.

digitalization continuous vs discrete analogue vs digital reification essentialism concepts functional pragmatism Pierre Bourdieu pragmatism

A number is not an explanation

by Neil

It’s almost old hat these days to bash the concept of “utility.” Thought experiments like the utility monster and the repugnant conclusion show the problems with generalizing utility into an ethical framework (and Peli Grietzer has some interesting recent tweets on the subject). But cutting deeper, pieces from @interpretantion (if you’re reading this, Alex, please update your SSL certificate) and our own Collin Lysford ask: in what sense is utility actually, like, a thing?

It’s revealing, I think, that we would never assert the predictions of a utility model over reality. Take Collin’s example of Hazard getting “more utility” from jumping higher and higher on a trampoline. When Hazard jumps too high and hits his head, we wouldn’t say to him: “no, you must be mistaken, it says right here that was the highest-utility jump yet. Are you sure you don’t feel happy?”

If I don’t enjoy my second apple as much as my first, that’s diminishing marginal utility, of course. But maybe I enjoy my second apple more than my first — maybe I needed some time to get into the right mood for apple-eating. Well, there’s a utility function for that too.

But if we’re just going to fit and re-fit a utility function to any possible statement of my preferences, then what’s the point of having the utility function at all? It seems to have literally zero predictive power. Somehow, making the problem a number seduces us into thinking we’ve explained something.

The concept of “status” runs into similar problems. Now, I do have to give some credit to the concept of status, because there are a cluster of body-language behaviors that can be identified as “things confident people do.” Shoulders relaxed, take up space, don’t speak too quickly, and so on. I do recommend learning healthy posture, and ideally, you can be be integrated enough with your body that these behaviors come naturally when you feel comfortable.

But status, like utility, is liable to being used as a post-hoc explanation that gives the illusion of being pre-hoc. In any situation where one person “owns” the other, we can say that the first person is higher-status — note how we’ve sneakily made status quantitative. But we’ve just restated the fact; it doesn’t explain why the situation played out the way it did. Nor does it provide any pragmatic value. “Be confident” is famously terrible advice.

Evo-psych explanations, in particular, tend to stretch status past the breaking point, using it as a catch-all for body-language, institutional prestige, reputation among peers, games of etiquette, even media recognition — all feeding into sexual success. But then the sentence “women prefer high-status men,” for all its red-pill mystique, collapses into either a tautology (women prefer men they prefer) or an uncontroversial platitude (women prefer men who are, like, good at stuff). And just like with utility, any specific formula for status must be endlessly caveated in the face of new data. If status is about power and wealth, how come broke musicians get laid? Now that must somehow also be high-status: ”well, actually, in the evolutionary environment…”

Obviously, it’s possible to do things to present yourself more successfully in social situations, and obviously, it’s possible to make decisions that make you more or less happy. But assigning a number to these things does nothing for us; it’s a kind of Wittgensteinian self-bewitchment. We make up a number because, we think, well, there must be a number — but must there, really?

utility status frames

Retroactive conceptualization

by Suspended Reason

[12:33 PM] suspended reason: idea I had on the dock today: retroactive reconceptualization

[12:39 PM] suspended reason: It comes out of two things:

1. The realization that many things I’ve historically wanted to call “retconning” were not really attempts at retroactively continuity, so much as just any attempt to produce new speech/text/narrative that retroactively altered the original text.
2. A conversation about strategies when you accidentally send the wrong person a text (e.g. that’s about them, or inappropriate for your relationship)—mine was “they don’t know when you realized you were texting the wrong person; immediately send a second text that alters the meaning of the first, while still appearing to be an accidental message to the wrong recipient,” so e.g. if you send your employee something about lingerie, you can chase it with an obvious spam link (“buy real ray-bans at tinyurl”), or if you shit-talk one friend to another friend (but send it to the first) you can chase it with something very complimentary, or that leverages ambiguities in your first text to put a different implied “spin” on it

[12:39 PM] suspended reason: I think people do this with text a lot

[12:40 PM] suspended reason: Maybe they make a commitment, or express a belief, or make a claim, and then a little later in conversation they may try to advance an alternative spin on that earlier text/speech by adding more speech to the first speech

[12:41 PM] suspended reason: Clumsy players do this explicitly—“Earlier, when I said X, I meant Y”—but skilled players do it implicitly, by adding information that a (clever listener) will pick up and use to retroactively update the meaning of a previous ambiguous utterance

[12:42 PM] suspended reason: To shift the “majority” interpretation (is there a word for this in like, probability/stats/Bayesian stuff? the event testified to most strongly by the evidence) of that previous text

[12:43 PM] suspended reason: Politicians do this constantly

[12:43 PM] suspended reason: People are forced to do this when old tweets come up that contradict present stances, or that are unflattering in a new political/cultural/moral landscape

[12:44 PM] suspended reason: And I think you even get it in television, or other serialized art—they want new affordances, and they can tell what possibilities have been “ruled out” by previous information, and which possibilities are still live/possible in the half-determinate structure of physics/reality/“rules of the game” they’ve constructed so far

retcon retroactive conceptualization strategic conceptualization strategic interaction

PvP v PvE

by Feast of Assumption

This post originally appeared on my blog, because it features Longtermism and Ethics, a Call-to-Action! and The Human Spirit! But the fellows at tis.so said “praise be, we finally have a pvp/pve intro post!” So you may wade through some Human Spirit!, but you’ll find frames and interaction here commingled.

I get ‘accused’ of being exceptionally credulous, straightforward, and non-torquey¹ in conversation on a somewhat regular basis; and when I ask “why though?” the response is “probably because your job is PvE instead of PvP.” And as we chew through the implications of a PvE mindset—where instead of viewing strategic interactions as person vs person, I view them as ‘me vs the universe hiding secrets’— most people come around to “and what if that’s good?”

Person vs Person, or Person vs Environment

When the primary interactions in your life are person vs person, you become accustomed to your triumph being at someone else’s cost, your achievement being negated by someone else’s loss. The situations in your world are zero-sum.

When the primary interactions in your life are person vs environment, that constraint isn’t there. When your sphere of influence is your own farm field, your excellence doesn’t come at the cost of your neighbor. And indeed, both in ‘midwestern farmer stereotype’ and in my observation, one of the most popular topics among farmers (after the weather) is “how are you going to do it better next year?” And they respond, candidly and honestly. And they learn from each other. They might all be selling the same product, but they do not feel like competitors.

This wasn’t remarkable to me until I worked outside of ag, and I found that “what are you doing to get better?” could be considered prying; and that answering might be discouraged as diluting competitive advantage. Solutions developed in a PvP, zero-sum mindset result in neutral or anti-social emergent outcomes: secrecy/dishonesty, focus on maintaining a delta between oneself and one’s competitors, considering peers as threats. Solutions developed in a PvE, non-zero-sum mindset produce neutral or pro-social emergent outcomes: knowledge transfer, exploration of waste that can be removed from systems, considering peers as allies, unlocking the universe’s secrets.

Why is PvP popular?

Some things just are zero-sum. There’s only so much water in the Colorado River. The developer can only hire one concrete company to pour the foundation for a new building. A hospital only needs one software provider. Someone will get hired, some company will get the contract—but somebody goes home empty-handed, and somebody’s dry.

Burton, Virginia Lee. Mike Mulligan and His Steam Shovel. 1939.

If 499 situations were zero-sum, it would be natural to prepare yourself for the 500th to be, too. But what if you noticed that people were resigning themselves to a zero-sum frame more often than they needed to?

Converting PvP to PvE: An Example

There are in fact lots of situations that appear PvP at a quick glance, but which can be reformulated to be PvE—or at least, to result in outcomes where both participants get some benefit.

This isn’t true of all interactions—sometimes there is a limited resource, and someone will lose. Often, though, there are interactions that appear zero-sum at first blush.

A farmer is buying seed corn for the spring. The farmer wants a low price, the seedsman wants a high price. This looks win-lose, right? And on a short timescale, it is.

But suppose you’re creative, and you’re in it for the long haul. You look at the environment the people are operating in. The farmer is thinking “I wish I could get the variety I want, at the quality I want, for cheaper.” And a seedsman is thinking “I wish I could know what my customers wanted 18 months in advance², so I could spin up production of the popular hybrids and not have produced excess seed for the unpopular hybrids.” Now our farmer and our seedsman can meet in advance, and lock in payment for future seed. The farmer can have a lower cost in exchange for the information he had anyway, and a commitment he is willing to make. The seedsman can tolerate getting a lower price, because she has less waste in her production system and fewer unwanted boxes sitting in her inventory. The farmer and the seedsman can tweak their environment to mutual benefit—the ‘loser’ was the waste generated by opaque demand—this was not zero-sum at all.

I use an ag example, but no matter the industry, there is no shortage of situations that could be solved to mutual benefit, but are misidentified as zero sum and solved with a winner and a loser. And what a waste that is.

Developing PvE Mindset

PvE thinking results in better outcomes across all areas it is applied to. My hypothesis is: the easiest way to habituate yourself to PvE thinking is to mostly solve problems that are only and obviously PvE. If no other person is involved, they can’t be losing. Start a garden, raise an animal, build yourself a woodshed. Build yourself a bookshelf, design a layout for your kitchen that saves you steps, practice changing a flat bike tire as quickly as you can. If you find yourself in a lot of PvP at work, when you’re at home, see if you can’t identify and practice solving problems in which nobody stands to lose.

Best of all (my hypothesis goes) is to start a person on PvE problems when they’re young. When most of a kid’s hours are spent in school or even sports, there’s inherent jockeying for the limited resources of attention, status, or being a starting player on the team. Replacing (or at least balancing) these with formative experiences of “how can you pull this weed and get the whole root out?” “How can we keep this pipe from rusting after we bury it?” “Can we think of a way to save the chickens from standing in their own food trough?” will result (or at least, they did for me) in adults who are straightforward, creative, and solution-oriented.

I don’t mean to say “growing up on a farm” is the only way to develop PvE mindset. (If that were the case, I wouldn’t be bothering to write this post!) Rather, I have met ‘city kids’ with PvE and ‘city kids’ with PvP. But, every farmkid I’ve ever met has led with PvE. Given the results of this natural experiment, my hope is to contemplate what it is about farm life that encourages PvE—and to discuss the tools that might make this mindset the prevailing way of living, no matter where formative experiences are had.

The creative, solution-oriented mindset described as PvE is one of the sharpest differences I observe between people who grew up near agriculture, and those who did not. Finding ways to develop this mindset in all people, and especially in children, should be one of the top priorities of longtermists, and of all who champion the human race.

Burton, Virginia Lee. Mike Mulligan and His Steam Shovel. 1939,

corn PvE PvP strategic interaction long-termism

Categories as heuristics, pt 2

by Suspended Reason

(Previously…)

RIP DCB says: let’s hear more about categories as heuristics!

Here’s my pitch:

Heuristics and frames

Let’s define a heuristic as a tactic for perception or action which is suited to a “type” of situation.

By “perception,” I mean the descriptive apprehension and determination of what “is”; by “action” I mean the (normatively guided) response to perception.

Note that we’ve also (implicitly) invoked Schutz’s theory of types.

To understand heuristics, we have to understand frames (or “frameworks” or “schemas”). Frames exist outside of and are complementary to heuristics. Heuristics author our frames: our sense of what “is” is a product of the tactics we use to monitor our environment. And frames author our heuristics—are the normative force which, emerging from our perceptions, guides our actions.

The circularity problem (similar to chickens and eggs) of heuristics and frames is resolved by reference to a biological “seeding” of heuristics. Our sensory organs themselves are perceptual heuristics: environment-fitted responses to an (experienced in the past, expected in the future) probability distribution of events, which becomes less and less effective as we continue along the distribution’s far tails and into the unknown. (Into events it is not optimized over, that its model is not “aware” of.)

Our strategy (bundle of tactics) for understanding what is, or what “type” of situation we are in, is composed of gradated bottom-up processing, from the “buzzing, blooming” level of pre-shapes and colors, to the attentional and cognitive prioritization of certain facets of our sensory field (over others), to the inference performed on these especially salient facets (e.g. regarding other agents’ intentions, the social status of an interaction, and similar high-level diagnoses).

We can think of the couplings between lower-level salient perceptual aspects, and their higher-level inferential meanings, as surrogates (or “cues”).

A concept, for instance, is something like a a complex of (perceptually salient) surrogates for identification (“differences which make a difference”). The concept’s easily observable characteristics become a premise for inferring or assuming other, less observable, but pragmatically relevant properties. A poker player may develop a concept of when his opponent is bluffing, based on surrogates such as hand trembles, which allow him to infer (the more pragmatically relevant matter of) whether said opponent has a strong hand. Inevitably there will be circumstances—whether or not they come up in a given game—when the opponent’s hand trembles and yet his bet is honest (for instance, if he has recently quit drinking), or when his hand is bad but his hand steady (for instance, he is mistaken about the state or quality of his own hand).

And when (as in the action-oriented model) we speak so as to have an effect on another agent, we will employ a heuristic (“tactic”) for accomplishing this effect. (Even if the transaction, and desired effect, are as simple as greeting a friend hello, making a doctor’s appointment, or asking a woman for her phone number.) These utterances will work best in more “typical” (i.e. center-of-distribution) situations, to which the heuristic is best suited, and work least in atypical situations.

Heuristics are fundamentally associative because they link situation to tactic; they are scoped, indexical, contingent solutions—couplings of “is” and “ought.” The de-indexicalization of surrogates is sometimes called “reification.”

Heuristics in adversarial (language) games

Heuristics fail when they have pressure put on them. As we have discussed, heuristics work well in a central set of situations, and then fail as we diverge from typicality. In strategy games, players are therefore incentivized to push the distribution of events towards those which are more atypical for opponents, so that their tactics (heuristics) will be least effective. If a fencer is strongest defensively, then force him onto the offensive. If America has zero nuclear icebreakers (because it has invested military resources optimizing different types of problems…), and Russia has half a dozen, then Russia should be aggressive pushing strategies and situations where a lack of nuclear icebreakers is a serious handicap.¹

Conceptual analysis predictably failed to derive satisfactory, robust definitions of concepts because it put a heuristic (the concept) under adversarial pressure. Philosophers could always imagine some scenario, some situation, in which a posited definition, or set of conditions, failed to hold.

Successfully working with and around heuristics is, rather, a deeply cooperative practice. Miller’s Law for communication: First assume it’s true; then figure out how. Because heuristics are fitted to frames—for instance, there is a relationship between an agent’s goals and his utterances, or his diagnosis of the present interactive game and his linguistic tactics—interactants can use their interlocutors’ deployed heuristics—which appear to them in the form of surrogates—in order to model higher-order intentionality (“project”) and understanding (“model”). Then this higher-order understanding of project and model can be used to disambiguate and make meaningful further of the interlocutors’ deployed heuristics (observed surrogates).

categories concepts heuristics ontology frames surrogates perception ACiM interpretation empowerment optimization conceptual analysis Miller's Law typification