tyr

Prompt

TASK: Generate a genuinely original thesis and write a competition-winning 500-word philosophy essay around it. You cannot browse the web; all information is below. Read everything first. 1. HARD RULES Immerse Education Essay Competition. Ages 13–18, thousands of entries per round. Prize: up to ten 100% scholarships per round to university summer programmes (Oxford, Cambridge, etc.). Marked blind — no name or personal details. Question: "How can you prove that you are not currently dreaming?" (Philosophy; based on Descartes' Dream Argument.) Max 500 words body text, 10% leeway (absolute max 550). Titles, footnotes, bibliography do NOT count. Target 500–520. Oxford footnotes + alphabetised bibliography. Deadline 25 October 2026. Age groups (13–15 / 16–18) marked separately. The competition has explicit AI rules; past winners state AI should only aid grammar and "the essence of the essay must remain the writer's authentic work." Your output will be reworked by the entrant; your job is architecture, not final polish. 2. RUBRIC AND WINNER PATTERNS Communicating ideas effectively — 30% (incl. authentic voice) · Critical analysis — 20% · Word choice & structure — 20% · Evidence & inference — 20% · Grammar — 10% Biggest scoring error: describing instead of analysing. Judges' documented praise: "a very mature and balanced approach," "the references show real depth of understanding and breadth of research." Winner openings: 2–3 sentences — a paradox, a quotation, or crisp definitions — then thesis. Never "Throughout history..." Thesis: one sentence, "In this essay I argue that...", with visible roadmap. 3–5 sources total, each given a full paragraph with dates/quotations. (One winner: Floyd case + Ibsen's A Doll's House + Snowden, one paragraph each.) Exact specificity: "net losses of $3,494 between 1974 and 1980." Conclusion: "In conclusion" or equivalent, callback to opening, nothing new. Source range is wide: one winner used zero citations and won on visible passion; another used 11 scholarly footnotes. Appropriateness is the bar, not prestige. 3. THE CORE PROBLEM Thirteen versions of this essay have been written and head-to-head judged. They all converge on one thesis family: "You cannot prove you are not dreaming — the question undermines itself, leaving intact the thinking self that asks it." Standard seminar position (unprovability + cogito + hinges + doubter survives). Best execution scored 9.5/10 — but philosophically identical. Any strong entrant could converge on it. Your task: a thesis-level escape. Not better craft on the same argument — a different core claim. The exemplar proves it's possible. 4. THE EXEMPLAR PATTERN (copy the pattern, not the answer) The one escape already built: the demand to prove I am not currently dreaming is partly grammar's demand. English forces tense on every verb, so the present feels like an inspectable object; once it could be fake, it demands proof. Zhuangzi's Chinese marks no tense — his doubt couldn't take the Cartesian shape. Calibration — copy this: "partly" is load-bearing. Never claim grammar invented the doubt (discredited strong Sapir-Whorf); claim it pre-installs the doubt's most tormenting form (mainstream "thinking for speaking," Slobin). Structure — mandatory for your essay too: ANSWER the question first ("No, almost certainly not, on ordinary evidence") — never dodge into pure dissolution THEN dissolve why proof nevertheless felt like an emergency Close on what survives This defeats the objection "you're avoiding the question — ARE we dreaming right now?" 5. CANDIDATE DOMAINS (use one, combine, or beat them all) (a) Predictive processing. Contemporary neuroscience (Seth, Friston, Clark): the brain always predicts; sensory input only constrains. Waking perception is "controlled hallucination" (Seth). Thesis: the dream/waking binary is a difference of degree — how tightly input constrains prediction — not of kind. (b) Retrospective-only verification. Dreams are known by their traces after, never during. Every dream ever verified was verified by WAKING — even LaBerge's REM experiments were confirmed by waking reports. "Am I dreaming NOW?" demands prospective verification of a structurally retrospective category — an impossibility independent of grammar. (c) The ethics angle. Augustine (Confessions X) worried whether one can be immoral in dreams. If dream actions harm no one, the doubt has no ethical grip; its urgency may be misplaced. (d) The phenomenological continuum. Hypnagogia, daydreaming, dissociation, flow, lucidity — a continuum, not a binary. Asking "am I dreaming?" is like asking "am I wet or dry?" standing in fog. (e) Indexicality. "Is HE dreaming?" is answerable (EEG, REM). "Am I dreaming?" is first-person, self-referential (Perry's "essential indexical"). Unanswerability as indexicality problem, not evidence problem. (f) The pragmatic dissolution. William James: what practical difference would knowing make? You'd behave identically. A question whose answer changes nothing is a mood masquerading as an epistemic problem. (g) Memory asymmetry. We remember waking from dreams; we never remember dreaming from waking. The direction of memory is itself asymmetrical — you always remember out of dreams, never into them. (h) The standard-of-proof equivocation. "Prove" means different things: mathematical proof, legal proof (beyond reasonable doubt), empirical warrant (science never certifies, only warrants). Thesis: under the only standard applicable to empirical matters, you CAN prove you are not dreaming — on the same warrant that proves you have hands. Under Cartesian infallibilism, you cannot prove anything empirical at all. The dream question reveals no special defect in our knowledge of wakefulness; it reveals the inapplicability of infallibilist standards generally. The Moorean route — Moore's "Here is one hand" proof (1939); Wittgenstein's On Certainty was written in response to it. (i) The agency asymmetry. Dreams happen to you; waking is something you do. In ordinary dreams you cannot decide to lift your arm and feel it lift, cannot sustain an argument across paragraphs — agency is what dreams characteristically lack (lucid dreams, the exception, are defined by recovering it). Thesis: the dream/waking line tracks agency, not vividness or coherence — and agency is self-verifying in a way evidence is not: to test it is to exercise it. (j) The evolutionary function. Revonsuo's Threat Simulation Theory (2000): dreams as offline threat rehearsal — training run in sleep because offline is safe. Thesis: dreaming is the simulation and waking the arena; the doubt asks the simulation to certify the arena from inside itself. (k) The biographical irony (a hook, not a thesis). Descartes had three famous dreams on the night of 10 November 1619, near Ulm, which he read as a divine calling to rebuild all knowledge — recorded in the Olympica, preserved in Baillet's 1691 biography. The author of the dream argument trusted a dream to launch the very project that doubted dreams. No entrant will have this opening. (l) The simulation update (high convergence risk). Bostrom (2003) — the modern dream argument. Many entrants will reach for this. Only viable with a distinguisher no one else will have. (m) Something none of these. Gödelian self-reference. Borges' 'The Circular Ruins' (a man dreams another into being, then discovers he is himself dreamt). Descartes' Meditation VI proof depends on God's non-deception — the dream question may be theology in disguise. 6. OBJECTION ARSENAL — what a philosophy-trained judge will raise, whatever your thesis 1. The contrast objection. "Dream" is a contrast word, like "counterfeit": a dream is experience masquerading as waking, ended by waking. Totalise it ("all experience is dream") and it means nothing — a dream never woken from is just a life. Counter: this is a weapon against the sceptic. His globalised version self-destructs; say so. 2. The self-defeat objection. Your dissolution is an argument; arguing requires wakeful reasoning; so you presuppose your conclusion. Counter: don't hide this — make it the ending. The doubting activity is the one thing the doubt cannot touch. A feature, not a leak. 3. The "so what" objection. If nothing practical changes, the essay is empty. Counter: name what changes — not the facts, but the classification of the failure. The sting dissolves; the practice stays ("I would still check the clock before leaving home"). 4. The overreach objection. Any strong version — grammar invented the doubt; waking IS dreaming — dies on contact with a knowledgeable judge. Counter: "partly," "difference of degree," "pre-installs rather than creates." Calibrate every claim to its defensible version, then no further. 7. WORKED STRESS-TEST — the method, demonstrated on (a) Thesis: Waking is "controlled hallucination" (Seth); the dream/waking binary is a difference of degree — how tightly input constrains prediction. Sharpest objection: If waking is a kind of dream, "dream" loses its contrast — you've relabelled everything. And inflating Seth's metaphor to "waking IS dreaming" is pop-science overreach a science-literate judge will punish. Calibration: Not "waking = dreaming" but "the boundary is a control knob, not a wall." Seth's own framing supports degrees — the same machinery runs dreaming with the constraints loosened. Convergence test: Seth + Inception is common-adjacent; needs a second distinguisher — answer-then-dissolve: verdict first, then Seth to explain why the binary felt absolute. Verdict: strong, current, but common-adjacent; viable only with the second distinguisher. Stress-testing means finding where the defensible version ends. 8. STRUCTURAL REQUIREMENTS (non-negotiable) Answer the question before dissolving it — an actual verdict, stated early or at the turn. "Partly" or equivalent calibration wherever the claim could overreach. Winner template: 2–3 sentence opening → one-sentence thesis with roadmap → 3–5 sources, each a full paragraph → "In conclusion" + callback, nothing new. One bracketed slot: [PERSONAL DREAM SENTENCE HERE] in an early paragraph. Quote primary sources verbatim wherever possible. Concede something. The strongest essays name their own strongest objection and survive it. 9. VERIFIED CITATIONS — THE ONLY ONES YOU MAY USE Never invent or alter a citation. Fabricated references invalidate entries. Zhuangzi, ch. 2, Watson trans., Zhuangzi: Basic Writings (Columbia UP, 2003): "solid and unmistakable Zhuang Zhou. But he didn't know if he was Zhuang Zhou who had dreamt he was a butterfly, or a butterfly dreaming that he was Zhuang Zhou." Original Chinese marks no tense; 夢 appears uninflected in both halves. Descartes, Meditations (1641), Cottingham trans. (CUP, 1996): Meditation I: "I am here, seated by the fire"; dreamt that scene while asleep in bed; "no certain marks" divide waking from sleep. Then the evil demon — ancestor of Bostrom's simulator. Meditation VI: "...when, without interruption, I can connect the perception I have of them with the whole of the other parts of my life, I am perfectly sure that what I thus perceive occurs while I am awake and not during sleep." Depends on God: "since God is no deceiver, it necessarily follows that I am not herein deceived." Meditation II: the cogito — even a deceived thinker must exist to be deceived. Hobbes, Third Set of Objections (printed with the 1641 Meditations): one could dream of passing the coherence test. Malcolm, 'Dreaming and Skepticism', Philosophical Review, 65/1 (1956), 14–37, p. 21: "if a person is in any state of consciousness it logically follows that he is not sound asleep." A sleeper can utter "I am asleep" but cannot assert it. Known weakness: consciousness excluding sleep is contested (LaBerge). Wittgenstein, On Certainty (Blackwell, 1969): §115 "If you tried to doubt everything you would not get as far as doubting anything." §341 hinges. §343 "If I want the door to turn, the hinges must stay put." Also Philosophical Investigations §309: "To shew the fly the way out of the fly-bottle." LaBerge et al., 'Lucid Dreaming Verified by Volitional Communication during REM Sleep', Perceptual and Motor Skills, 52 (1981), 727–732: lucid dreamers signalled with pre-arranged eye movements during REM. Verification depended on waking reports. Moore, 'Proof of an External World', Proceedings of the British Academy, 25 (1939): the "Here is one hand" proof. On Certainty was written in decades-long response to it. Nietzsche, Twilight of the Idols (1889), "'Reason' in Philosophy" §5: "I fear we are not getting rid of God because we still believe in grammar." Revonsuo, 'The Reinterpretation of Dreams', Behavioral and Brain Sciences, 23 (2000), 877–901: Threat Simulation Theory. Bostrom, 'Are You Living in a Computer Simulation?', Philosophical Quarterly, 53 (2003), 243–255: the modern dream argument. High convergence risk. Sosa, A Virtue Epistemology (OUP, 2007): in dreaming "we do not really believe; we only make-believe." Ichikawa, 'Skepticism and the Imagination Model of Dreaming', Philosophical Quarterly, 58 (2008), 519–527: the reply — experiences remain subjectively indistinguishable. Augustine, Confessions Book X: dream morality. Seth, Being You (Faber, 2021): "controlled hallucination." Clark, Surfing Uncertainty (OUP, 2015). James, Pragmatism (1907): if no practical consequences differ, the dispute is verbal. Borges, 'The Circular Ruins' (Ficciones, 1944). Baillet, La Vie de Monsieur Descartes (Paris, 1691): records the three dreams of 10 November 1619 (the Olympica). 10. FAILURE MODES (from thirteen judged versions) Overcrowding: 7–10 sources = a list. Winners use 3–5. Pure dissolution without verdict: "we can never really know" = despair dressed as depth. Over-qualification: hedging reads as weak conviction. Template visibility: winners' formulas used twice show the checklist. Machine-perfection paradox: the thirteen versions are technically cleaner than any essay that has ever actually won. The published winners contain comma splices and duplicated paragraphs. The entrant reworks the final text; your job is architecture. 11. YOUR PROCESS Generate at least 5 candidate theses — one paragraph each, argued seriously. At least one from §5; at least two that appear nowhere in this brief. Stress-test each (following §7): sharpest judge objection · where the defensible version ends · whether a strong entrant could converge on it independently. Interrogate the winner: Can a judge state your move in one sentence after one read? Does the essay contain at least one thing no other entry will have? Does it answer before it dissolves? Write — 500–520 words, citations from §9 only. Generate the candidates now. The bar to beat is 9.5/10 — the best of thirteen versions. Beat it on originality, not just craft. YOU SHOULD OUTPUT THE BEST OF THE BEST, ABSOLUTE PEAK ORIGINALITY AND CRAFT.