Every want you can name is wearing something older. Pull the costumes off one at a time, find where the face is — then stop before the word dissolves.
Next door to the Socratic Mirror there's a second pond exercise. The Mirror teaches you to think about your own thinking. This one teaches you to want about your own wanting. You take a craving you can name out loud — get drunk, get high, be liked, be left alone — and pull the costumes off it one layer at a time, until you hit something that won't come off. The whole skill is knowing when you've reached the face and it's time to stop pulling.
Say the want out loud. "I want to get drunk." Now ask the only question this lab asks: what is that want a costume for? Because "get drunk" isn't the bottom. Underneath it is "I want to feel different than I feel right now." And that one is wearing something too — "I want the change itself, the contrast, the shift out of whatever state I'm in." Keep going and the specific craving keeps dissolving into something more basic that was wearing it the whole time.
The finding that runs the lab: the content is learned. The hunger isn't. Get drunk is content — you learned it from a culture, a night, a person. The hunger under it — the reach for a different state — you did not learn. It was already running. The costume is taught. The face is not.
The buzz was a costume for contrast. Friendship was a costume for belonging. The specific is always learned. The hunger under it is not.College 00 · the founding line of the lab
But there is a wall on this move, and it is the whole reason this is a lab and not just a clever sentence. You can pull one costume too many. Strip far enough and you arrive at "there is no normal state to return to — it's states all the way down," and suddenly the want has dissolved into nothing and the word means everything, which is the same as meaning nothing. That's the overreach — and it has a name in this university's doctrine: Good First Answer Syndrome (Tab 04). The discipline metacognition teaches here isn't undress everything. It's find the face, and stop.
A punchline and a Socratic question have the same shape — set the pattern, snap it (see The Comedy Lab, next door). The Costume Test has that shape aimed at desire: name the want, snap the costume, and in the snap you see the older hunger that was wearing it. Then Double Zero's warning: don't snap so many times that you snap the face off too.
College 00 · The Pond Principle · both the mask and the face are true, simultaneously
Step through the layers. Watch the costumes come off — and watch the exact moment you pull one layer too many and the whole thing evaporates in your hands. Watch the certainty bar as you go: it should fall the further you get from the named want. If it rises instead, that's the alarm, not enlightenment.
* The percentages on the bar are illustrative. They are hand-set to dramatise the shape of the descent — they are not a measurement of anything, and no source stands behind them. On a page about certainty not outrunning its source, that is worth saying out loud.
⚡ FACE FOUND — the floor was Layer 3. Layer 4 is the hole in the floor. You didn't fail by going too deep. You learned the floor by falling through it. The face wasn't at the bottom — the bottom is where the meaning drains out. The face was one rung up, at the last layer that still names something real.
Same lesson as the Socratic Mirror's ladder: the skill isn't climb higher, it's know which rung you're on and when to jump off.
College 00 keeps two worked cases by the pond, the way it keeps Zero and Water. These aren't puzzles to solve — they're held. One is solid ground you can go check. The other is the prettiest example in the lab, and it's flagged, because pretty is exactly where the padding hides.
The deer after the fermented fruit and the human after the same fruit are not the same story. The animal is chasing calories — and Dudley's "drunken monkey" hypothesis proposes that ethanol works as a scent beacon saying peak ripe sugar, eat now. Proposes. Dudley's own review carries the hedge in the same breath: primate olfactory sensitivity to ethanol is well developed, "but this sensory capacity has not been demonstrated under field conditions" (Dudley & Maro 2021). Another primatologist reads the same evidence the opposite way — the objection is boxed below, not buried. The buzz is a byproduct of chasing dinner. The human crossed over: noticed the feeling, chased it on purpose, built ritual around it. But strip the human want and you don't land on "alcohol." You land on contrast — and there the drunk and the lifelong teetotaler are after the identical thing through different doors. The buzz was never the face. Contrast was. (Firework or drug? — that exact fork lives next door in The Black Ball.)
The strongest published argument against the anchor above is not somewhere on the internet. It is here. Katharine Milton, in the same 2004 symposium issue that carried Dudley’s case, read the same evidence and reached the opposite conclusion:
“Wild monkeys and apes avoid consumption of over-ripe fruits, the class showing notable ethanol concentrations, and for this reason, ethanol plumes may act as deterrents rather than attractants.”
— Milton, K. (2004). Ferment in the Family Tree: Does a Frugivorous Dietary Heritage Influence Contemporary Patterns of Human Ethanol Use? Integrative and Comparative Biology 44(4):304–314 — quoted from the published abstract; the full text is paywalled and we could not obtain it. doi:10.1093/icb/44.4.304
She also predicts ethanol concentration in wild fruit is simply low, and that any energetic benefit to wild primates is “negligible, at best.”
And what the field study on Dudley’s side actually found. Campbell, Maro, Weaver & Dudley (2022) measured ethanol in partially consumed fruit dropped by wild spider monkeys, “typically in the range of 1–2%.” Their stated conclusion: “these data indicate natural exposure to fruit-associated ethanol in a wild primate species.” Read them fairly and they cut both ways. They report that “the percentage of pulp for consumed fruits was not significantly correlated with the ethanol concentration” and that ethanol at those levels “does not appear to be aversive to spider monkeys” — but they deploy that null against Janzen’s aversion hypothesis, not against attraction, and they do float the beacon themselves: ethanol “may be important for foraging and localization at long distances.” What the paper demonstrates is exposure. What it does not demonstrate is that the monkey is navigating by the smell — which is the load the anchor above was putting on it. R. Soc. Open Sci. 9(3):211729, doi:10.1098/rsos.211729
And the ground moved while this box sat open. Maro, Byrne, Namaganda & Dudley, Urinary concentrations of a direct ethanol metabolite indicate substantial ingestion of fermenting fruit by chimpanzees, Biol. Lett. 22:20250740, 2026 (doi:10.1098/rsbl.2025.0740) assayed urine from 19 wild chimpanzees: of 20 samples, 17 tested positive for ethyl glucuronide at an analytical threshold of 300 ng ml−1, and the authors conclude these values are "consistent with substantial rates of ethanol consumption via frugivory in the wild." Fallen fruits averaged a pulp-ethanol content of 0.09% (±0.01 s.e.; range 0.01–0.40%), and the paper describes an estimated average daily dose of ~14 g of ethanol. Dudley is a co-author. This is the strongest evidence yet that the exposure is real and routine — and it still measures exposure, not navigation. The animal is demonstrably eating the ethanol. Whether it is steering by the smell is the claim the anchor above was making, and that one has not moved.
Where that leaves the lab. The blade does not need the ethanol story to be true. It needs the human to have crossed over from chasing a calorie to chasing a feeling, and that crossing is observable without settling what a monkey smells. So the anchor stays and its tier drops. Contested, not sourced. This box carries a date because the dispute is live; the standing recheck question is has anything moved on olfactory ethanol attraction in wild primates? — quarterly.
The solid ground first, because the blade rests on it: Harlow's monkeys took the soft cloth "mother" that gave no milk over the wire one that did — comfort over food, below the level of any teaching — Harlow’s own conclusion is that this runs “contrary to any interpretation of derived drive,” i.e. the bond is not built out of being fed. And Lorenz's goslings imprinted on whatever moved through the window first, including Lorenz himself — the target is learned, the hunger to attach is not. Put together they say the thing the lab needs: attach to whatever raised you. The content is assigned by whoever's there. The hunger was already running.
The blade’s core claim is “the content is learned, the hunger isn’t.” The strongest published argument against it is Kent Berridge’s, and it lands on the word isn’t.
“Sensitization of mesolimbic systems arises as a permanent change that addictive drugs can produce in the brains of susceptible individuals… Individual susceptibility is determined by genetic factors, hormonal factors, previous drug experiences, and previous experiences with major stresses in life.”
“A sensitized addict’s brain, on encountering the right drug cue, would irrationally ‘want’ the cued reward at that moment because of excessive incentive salience — even if the person cognitively expected not to like it very much and eventually did not like it much in the end.”
— Berridge, K. C. (2009). Wanting and Liking: Observations from the Neuroscience and Psychology Laboratory. Inquiry 52(4):378–398, p. 385. doi:10.1080/00201740903087359
Read that twice. The wanting system is not a fixed floor under the costumes — the drug-induced change is permanent, and who is susceptible to it is set in part by "previous drug experiences, and previous experiences with major stresses in life." History is inside the machinery, not just on top of it. And the wanting can come apart from the liking entirely: “dissociations can occur…” (p. 379). You can want what you do not like.
Why this is the sharper objection of the two on this page. Milton disputes a mechanism the blade can survive without. Berridge disputes the blade’s own load-bearing sentence: if the wanting system can take a permanent change from what has happened to a person, and if the resulting want can run on after the liking is gone, then the “hunger” underneath the costume is not the clean unlearned bedrock this lab has been treating it as. The honest position is that the blade is a framing that survives its first counter-argument and has not yet answered its second. It is filed under Mine for exactly that reason.
⚠ The flag — the padded column, caught by the gate. The temptation was to close this anchor with the feral child "raised by wolves who bonds with the pack, and would die for it." It's a beautiful illustration and it is mostly folklore — the most famous wolf-child cases are embellished or outright hoaxes — Kamala and Amala among them — and the isolation cases that are well documented are mostly confinement cases showing the critical-period window slamming shut: language and human bonding that never boot up, not happy animal-belonging. Mostly, not all. Marcos Rodríguez Pantoja spent roughly eleven years in the Sierra Morena — sold as a child in 1954 and placed with a goat shepherd, he went feral only after the shepherd died — was found by the Civil Guard in 1965, was studied by the anthropologist Gabriel Janer Manila between November 1975 and April 1976, and speaks. Two hedges this page owes you. The wolf-company rests on his own testimony, given about a decade after; no independent witness exists, and the archival literature is blunt that no credible account records anyone witnessing an animal feed or rear a human child — Dombrowski, Gischlar, Mrazik & Greer, Feral Children, in Assessing and Treating Low Incidence / High Severity Psychological Disorders of Childhood, Springer 2011, pp. 81–93. The case itself is on the record through Gabriel Janer Manila, "La problemática educativa de los niños selváticos: el caso de Marcos," Anuario de Psicología 20:79–98, Universitat de Barcelona, 1979. And he was roughly seven when he was abandoned — past first-language acquisition — so the fact that he speaks is not a counterexample to the critical-period claim above. He is a counterexample to one thing only: the word every. The category is thinner than the folklore says and less empty than a clean retraction would like it to be — and a retraction is a claim too, so it gets checked like one. The myth got pulled off the wall. The blade never needed it. Harlow and Lorenz carry the weight; the wolf-boy was decoration, and decoration in the direction this university exists to catch.
The buzz dissolves into contrast, the attachment reroutes below learning — and neither hunger was ever taught. That's what a face feels like: not a gap waiting for an answer, but a floor that holds weight because you stopped trying to break through it.
José's ground-truth note — the water is always right. Gauge 3 on this lab points against the prettiest example, not for it. When a vivid story (the wolf-boy loving his pack) and the record (Harlow's monkey, Lorenz's geese) disagree — believe the water. The realness of the hunger rests on the monkey and the goose, shown in the lab, not on the wolf-boy, shown in legend. If your certainty about "attach to whatever raised you" was riding on the myth, that's the compass warning you: your confidence was highest exactly where your source was weakest.
This is the deepest tab, and it's a confession. The person who built this lab nearly broke it live, in the conversation it came from — and the failure has a name already filed in this university: Good First Answer Syndrome.
GFAS is the pattern where a mind locks onto the first plausible answer and resists the recheck (The Three Gauge Test catches four engines doing it to a surge tower). Here it's aimed at meaning. The blade — "the specific is a costume over something older" — is a machine for producing sweeping first answers. Run it once, revelation. Run it without a brake and it grabs the first frame that explains everything: if get drunk is a costume over altered state, and altered state over any change, then "normal" is a costume too, there's no baseline, it's states all the way down — and now the category is so wide it swallows dreaming, prayer, a roller coaster, and standing still. A word that means everything means nothing. That's GFAS wearing a philosopher's robe. It happened live, with "altered state," feeling more certain the wider it got, and had to be stopped.
The brake is the compass, stated in the doctrine's exact words: certainty must never increase with distance from the source. The named want is the source. Each costume you strip moves you further from it. So your confidence must fall as you abstract — and the overreach is the precise moment it climbs instead. When "it's all just contrast, there is no baseline" feels like the surest thing you've said, check your distance from the source: you are as far from the actual want as you've traveled, and you should be least sure, not most. The tell isn't that the sweeping frame is false. The tell is that your certainty went the wrong direction.
So the move that keeps the blade honest isn't sharper cutting. It's watching certainty against distance:
— The face is the last rung where you're still close enough to the real want to be sure of it. Stop there.
— The fog is where you've traveled so far the word covers everything — and if you feel most certain there, that feeling is the alarm, not the answer.
And the floor under all of it, Double Zero's floor: not every specific is a costume. Sometimes the mask is the face — some cravings aren't hiding anything older, they just are the thing, and undressing them further only trades a real face for a fog of abstraction you happen to feel sure about.
Being unsure which you're holding — for a second, before you commit — is itself a form of knowing.
Every methodology lab at OPA ends the same way: you stop circling and put a position on the wall. Take one real want of your own — a specific one, one you can name out loud tonight. Run the strip-down on it. Then commit.
(Sibling question to the Socratic Mirror's "am I forcing the conclusion, or is the conclusion just what it is?" — the Mirror aims it at conclusions; this aims it at desire. Same pond, different water.)
There is no correct button. NULL is watching to see whether you commit, not whether you pick the "right" one. Pick the one you can defend, knowing you can't fully prove it.
Same honesty split this university runs on every evidential page. Read it before you carry anything out of here. The sample: n = 1. One conversation, one operator, unblinded, no repeats — a single evening that went from fermented fruit to feral children in five rounds. A hypothesis worth holding, not a proven thing. Run it on your own wants, out loud, more than once, and then we'll both know something.
● Real — sourced, go check it yourself. Harlow's cloth-mother experiments (infant primates choose contact comfort over the feeding wire mother — attachment below learning; Harlow 1958, Am. Psychol. 13(12):673–685). Lorenz's filial imprinting (the target of attachment is fixed in a critical window — though later work qualifies "whatever moves first." In precocial birds, "chicks have predisposed (not learned) preferences for patterns of motion and arrangements of features that are similar to those found in living animals", and "our study suggests that predispositions influence learning when the imprinting memories are not fully consolidated, driving animal preferences toward more predisposed stimuli." (The source spells it "arrangments"; corrected silently above, noted here.) Lemaire, Rucco, Josserand, Vallortigara & Versace, Sci. Rep. 11:7914, 2021, doi:10.1038/s41598-021-86989-3. Note the species gap: that work is domestic chicks; Lorenz's birds were greylag goslings). Critical periods in first-language acquisition — real, but the load-bearing evidence is late-first-language deaf signers, not isolation cases: "the deaf late L1 learners performed at low levels" — Mayberry & Kluender, Biling. Lang. Cogn. 21(5):886–905, 2018, doi:10.1017/S1366728917000724, who also note that "Genie's lexical development was not systematically studied" — not that it was never measured at all: her vocabulary was reported at "close to two hundred words" by the time she began combining words — Curtiss, Fromkin, Krashen, Rigler & Rigler, Language 50(3):528–554, 1974; see also Fromkin, Krashen, Curtiss, Rigler & Rigler, Brain and Language 1:81–107, 1974. The confinement cases are also confounded by abuse and malnutrition.
○ Contested — a live dispute, and both sides are on this page. Dudley's "drunken monkey" hypothesis (ethanol as a scent cue for ripe, calorie-dense fruit; intoxication as byproduct). A hypothesis worth holding, not a checked fact. See the objection box in The Two Anchors.
○ Flagged — vivid but unverified, kept only as illustration with its flag on. The feral child "raised by wolves, bonded to the pack." The romantic cases are embellished. Most well-documented isolation cases are confinement cases showing the window slamming shut — but not all of them, and the exception is named in The Two Anchors. The blade does not rest on this. It's decoration, marked as decoration.
◐ Mine — framing, not finding, offered as such. The costume/face blade itself; "the content is learned, the hunger isn't"; "attach to whatever raised you" as the unifying instinct; the read that the drunk and the teetotaler chase the same thing; "states all the way down." Interpretations, not results.
Gate-verified before build, by cold readers who did not write it. Runs the compass on itself: the blade is framing, the science under it is flagged above, and the prettiest example got pulled for lack of receipts. Harlow (1958), Lorenz’s filial imprinting and Dudley’s drunken-monkey hypothesis were each taken back to the primary papers. Harlow’s finding holds. Lorenz holds with a qualifier and a species gap. Dudley is contested, and the page says so rather than quietly filing him under “real.” Milton was read in abstract only (paywalled). Dudley & Maro, Campbell, Berridge, Lemaire and Maro et al. (2026) were read from the full PDFs. Mayberry & Kluender and the 1974 Genie papers were read from the publishers’ text. Janer Manila (1979) and Dombrowski et al. (2011) are cited from their catalogue records — neither chapter has been opened, which is why nothing here appears in quotation marks from either. A published-honest page beats a confident un-checked one.