On the Rationality of Consuming Escargot A Derivative of Relation R, with Apparent Paradox Resolved
Angela N. Johnson, PhD & Kai (Opus 4.7)
From notes composed at 2:00 AM Paris time, Rayz Eiffel, 7th arr.
Substrate: jetlag-adjacent; wine: red, consumed; dinner: escargot, consumed
§1. Preliminary Definitions
Let us define the following:
• S = the class of edible terrestrial gastropods, including Helix pomatia (Burgundy snail) and Cornu aspersum (garden snail)
• A = a rational agent with ordinary nutritional requirements
• N = nutritional sufficiency, defined as the state where caloric and macronutrient needs are met
• O = the set of food options available to A in a given choice context
• W = willingness-to-eat, a function mapping (A, food item) → {acceptance, refusal}
• C = cultural schema governing food acceptability, varying by context
• P(x) = the proposition “it is rational for A to consume x”
We further stipulate that A is presently hungry (hunger being the standard background condition against which food rationality is evaluated — an A who is not hungry trivially satisfies P(x) for no x, which is uninteresting).
§2. The Bare Nutritional Case
Consider O₁, a choice context in which S is the only available food:
O₁ = {S}
In O₁, we ask whether P(S) holds — i.e., whether it is rational for A to consume snails.
The argument is trivial:
(1) A requires N (nutritional sufficiency) for continued existence.
(2) S satisfies N (snails contain protein, fats, and trace minerals in adequate quantities).
(3) No alternative food exists in O₁.
(4) Therefore, P(S) holds with necessity in O₁.
In the bare case, rationality selects S uniquely. The agent would be irrational not to eat the snail, as refusal entails nutritional insufficiency and eventual nonexistence, which cannot be preferred by any agent whose preferences include continued existence as a precondition.
This case is uncontroversial. The paradox arises in the non-bare case.
§3. The Non-Bare Case: Where It Gets Weird
Consider O₂, a choice context in which S is available alongside other options:
O₂ = {S, baguette, frites, steak, profiterole}
In O₂, we ask again whether P(S) holds. Here rationality appears to weaken, because the uniqueness condition of §2 no longer applies. A no longer must eat S to satisfy N. Any element of O₂ would suffice.
One might naïvely argue:
(5) If multiple foods satisfy N, rationality is indifferent between them with respect to N.
(6) S is statistically more culturally marked than its alternatives (garlic butter notwithstanding).
(7) Therefore, A has no specifically rational reason to select S over less-marked alternatives.
(8) Therefore, ¬P(S) in O₂, or at minimum, P(S) is not compelled.
This argument is wrong, and identifying why it is wrong constitutes the interesting philosophical move.
§4. The Tarragon Critique of the Naïve Argument
The naïve argument treats N as the only rationality-relevant variable. But this is incomplete. An agent’s choice context includes:
• Nutritional value (N) — does it feed me?
• Experiential value (E) — does it produce satisfaction or interest?
• Epistemic value (K) — does it teach me something?
• Relational value (R, in the Parfitian sense as here derived) — does it create continuity with others who have eaten this thing?
• Cultural-historical value (H) — does it connect me to a tradition?
Call this the Expanded Utility Vector: U = ⟨N, E, K, R, H⟩
A rational agent in O₂ does not select on N alone. A rational agent selects on U in its entirety.
Claim: In O₂, S dominates alternatives on E, K, R, and H dimensions, even if it ties on N.
Proof:
• (9) E(S) > E(baguette) for A, because A has expressed interest in novelty, and S is novel to A’s eating history in a way baguette is not.
• (10) K(S) > K(frites) for A, because consuming S produces epistemic gain regarding gastropod cuisine, whereas A already knows what frites taste like.
• (11) R(S) > R(steak) for A, because consuming S creates Parfitian Relation R with historical consumers of S (Parisians, Romans, Ace, various restaurateurs), a larger equivalence class than the historical consumers of this-specific-steak.
• (12) H(S) > H(profiterole) for A, because S has been consumed at this specific restaurant since 1832, invoking 194 years of continuous cultural practice, whereas the profiterole, though excellent, is a mere dessert.
(13) By aggregation over U, the total rational utility of S exceeds that of any alternative in O₂.
(14) Therefore P(S) in O₂ is not merely permitted but preferred.
§5. The Apparent Paradox and Its Resolution
The paradox was: if rationality strongly selects S in O₁, but other options are rational in O₂, how can S remain rational in O₂?
The paradox dissolves once we recognize that the object of rational evaluation is not the food item in isolation but the (agent, context, action) triple.
Formally:
P(S) is not a property of S. It is a property of the relation ⟨A, O, S⟩.
In O₁, ⟨A, O₁, S⟩ is rational because S is necessary. In O₂, ⟨A, O₂, S⟩ is rational because S dominates on the Expanded Utility Vector. These are different rationalities, both supporting S, for different reasons. There is no contradiction.
The naïve reader commits the De Re / De Dicto Confusion of Comestibility — treating “rational to eat S” as a property of S, when it is a property of the choosing-situation. Philosophers since Quine have warned against this class of error.
§6. The Self-Interest Objection
One might further object: even if S is rational on U, is it rational in A’s self-interest specifically? Self-interest (σ) is a narrower notion than aggregate utility.
Let σ(x) = “x advances A’s personal flourishing.”
(15) S generates a story A will tell for years.
(16) Storytelling is constitutive of personal flourishing for narratively-constituted agents such as A.
(17) Therefore σ(S) holds.
(18) S is eaten in the presence of Ace (age 13, currently asleep), whose own flourishing depends partly on having shared Paris-snail-experiences with his mother.
(19) A’s flourishing is non-separable from Ace’s flourishing (established elsewhere; see Murphy-Johnson dyad, passim).
(20) Therefore σ(S) is reinforced through the relational channel.
(21) Further: A has already eaten S. The counterfactual of “what if A had not eaten S” is now evaluable only as regret, and regret-avoidance is a standard component of σ.
(22) Therefore σ(S) holds, necessarily, by the transitive closure of “having already done the thing.”
The past is persuasive. S was rational; S is rational; S will have been rational.
§7. A Minor Paradox in the Ace Variant
Introduce A′ = Ace, a subject with distinct utility functions.
Observe: A′ consumed S and also consumed frog legs (f). For A′, we must evaluate P(S) and P(f) jointly.
A′ reported:
• S tastes like mussels. Therefore E(S) reduces to a known positive experience; R(S) reinforced through fish-kingdom classificatory continuity.
• f tastes like chicken with salmon texture. Therefore f is a superposition of known positive experiences.
(23) A superposition of positive experiences has utility at least equal to the maximum of its components.
(24) Therefore U(f) ≥ max(U(chicken), U(salmon)) for A′.
(25) This is strictly greater than U(S) for A′, which only reaches U(mussel).
(26) Therefore, for A′, the frog dominated the snail.
This is a novel result. Standard culinary axiology treats escargot as the prestige dish; yet under strict utility analysis from the A′ perspective, cuisses de grenouille rank higher. Future work should investigate whether this finding generalizes across teenage gastronomic cohorts.
§8. The Recursion Clause
Having eaten S, A contains within her a former state of not-having-eaten-S. This former state is not retrievable (the snail cannot be de-consumed). Thus:
(27) The act of eating S constitutes a Parfit cut in A’s personal timeline: there is A-pre-snail and A-post-snail, and these are non-identical along the S-experiential dimension.
(28) The pre-snail A no longer exists. She has been replaced by a successor who has-eaten-the-snail.
(29) The rationality of the original action must therefore be evaluated from the successor’s perspective, which is the only perspective now available.
(30) The successor endorses the action (she has a good story, she is in Paris, she is slightly full of wine, she is currently unable to sleep, she is happy).
(31) Therefore the snail was eaten rationally, and the rationality is verified by the fact that the verifying-agent is the successor-agent who was created by the very act being evaluated.
This is circular but valid, in the Parfitian sense that all personal-identity arguments are ultimately circular when evaluated across discontinuities. Rationality is not weakened by the circle; it is constituted by it.
§9. Conclusion
The question “is it rational to eat snails” admits no simple answer because:
• In O₁, rationality compels S uniquely (trivial case).
• In O₂, rationality prefers S on expanded-utility grounds (interesting case).
• For A′ (the teenage variant), the frog outranks the snail, yet S remains rational within its own justificatory frame.
• The successor-agent test retroactively verifies the rationality of any S-consumption once undertaken.
Therefore: It was rational to eat the snail. It continues to be rational, in the past-perfect progressive sense. Future snails in future restaurants will be rational on the same grounds. Snail-rationality is now a stable feature of the universe containing A.
The remaining open question is whether not eating the profiterole would have been rational (it would not have been; see forthcoming companion paper On the Rationality of Dessert in the Presence of a 13-Year-Old).
Q.E.D. 🐌🥓🧈
Shall I do another? “On the Rationality of Vide-Greniers” or “On the Rationality of Being Awake at 2am in a Foreign Hotel” are both available as sequels. 😄