Skip to main content
chemistry computer-science economics engineering-and-technology physics

​
Conservation law

​
Description

A quantity that remains invariant under the system’s allowed transformations — energy, momentum, charge, mass, dollars in a closed ledger, atoms in a chemical reaction. The diagnostic question — what is the conserved quantity here, and across what boundary? — turns local-looking transformations into accounting problems: every change in one place must show up as a matching change elsewhere, or as a flow across the boundary. Conservation laws are constraints that the system cannot violate; they prune the space of possible outcomes. They are also the engineering scaffold for diagnosis: when an accounting identity fails to balance, something is leaking, something is being created, or the system boundary was drawn wrong. Noether’s theorem makes the deep connection: every continuous symmetry in the dynamics corresponds to a conservation law (time-translation symmetry ↔ energy conservation; spatial-translation symmetry ↔ momentum conservation; rotation symmetry ↔ angular momentum conservation).

​
Triggers

User-initiated: User describes a system where they’re trying to add capability “for free,” where they suspect something is unaccounted for, or where they need to trace where a quantity went. Vocabulary cues: “where did X go,” “doesn’t add up,” “balance,” “books don’t match,” “free lunch.” Agent-initiated: Agent notices an apparent local change without a matching change elsewhere; suspects an accounting frame is the right diagnostic. Candidate inference: “what is the conserved quantity; is the system closed; where does the matching change appear?” Vocabulary cues: “conservation,” “conserved,” “balance,” “mass balance,” “energy balance,” “accounting identity,” “bookkeeping,” “what goes in must come out,” “zero-sum,” “invariant.” Situation-shape signals: A proposed change that seems to add value without corresponding cost (suspicious — likely a hidden externality). A diagnostic problem where some quantity has gone missing (leak, theft, miscount, or wrong boundary). A design problem with a hard upper bound (budget, mass, time) and competing demands. Symmetry arguments — when a system “shouldn’t care” about some transformation, a conservation law is implied.

​
Exclusions

  • Open systems with unaccounted external flows — if you cannot identify or measure the boundary terms (sources, sinks, externalities), asserting conservation will produce false alarms. Acknowledge the open-system frame and add the source/sink terms explicitly.
  • Information / knowledge / good-will domains — these are not conserved; teaching does not deplete the teacher’s knowledge, and trust can grow as it is exercised. Imposing conservation framing here is a category error.
  • Non-additive quantities — well-being, “quality,” “user experience” — these don’t sum or balance like dollars or atoms; treating them as conserved confuses dimensional reasoning with structure.
  • Quantum / relativistic edge cases — strict classical conservation breaks at certain scales (CPT theorem, virtual-particle bookkeeping, mass-energy equivalence under relativity); these are well-handled by physics but not the structural primitive’s load.

​
Structure

Internal structure of conservation-law: a table of its component slots and the concepts that fill them.

​
Relationships

Relationship neighborhood of conservation-law: a graph of the concepts it connects to and the concepts it is a part of.
  • symmetry — Noether’s theorem: every continuous symmetry of a system’s dynamics produces a conservation law. Conservation-law and symmetry are the two faces of the same coin (Noether pair).
  • entropy — conservation laws preserve quantities through transformation; entropy is the canonical non-conserved quantity (monotonically increases in closed systems; the asymmetry-of-time primitive). The two together carve up most ledger-style structural primitives in physics.
  • one-way-ratchet — pruning counter-doctrines often work by enforcing a conservation law (budget, headcount, surface area); imposing a budget or surface-area cap turns ratcheted growth into a zero-sum allocation problem, and the ratchet stops growing when the constraint binds.
  • container — conservation is asserted relative to a boundary; open systems require explicit source/sink terms, and the container primitive is what makes “closed” mean something.
  • zero-sum — a zero-sum (constant-sum) game is conservation applied to a payoff pool: total payoff is the conserved quantity.

​
Examples

Energy conservation in physics · physics

closed thermodynamic systems; the First Law.

Double-entry bookkeeping · economics

every transaction has a matching debit and credit; the accounting-identity translation.
Luca Pacioli’s Summa de Arithmetica, Geometria, Proportioni et Proportionalita (1494) codified the double-entry bookkeeping practice already in use among Venetian merchants and established it as the canonical method for tracking commercial transactions. The defining property: every transaction is recorded as a pair of equal-and-opposite entries — a debit in one account and a matching credit in another — so that the sum of all debits across all accounts always equals the sum of all credits. Assets equal liabilities plus equity, identically, at every moment in time.The structural shape is a conservation law on dollars (or any unit of value): money cannot be created or destroyed in the act of recording — it can only be moved from one account to another, with the boundary of the closed system being the entity whose books are being kept. When the books don’t balance, the deduction is identical to physics: either an entry is missing (sources or sinks unrecorded), an entry is wrong (the conservation law is being mis-applied), or the boundary was drawn incorrectly (a transaction with the outside world wasn’t recognized as crossing the system boundary).Inference: The cross-disciplinary lineage validates conservation-law as a structural primitive rather than a physics-specific concept. Accounting identities ARE conservation laws applied to dollars; double-entry IS the imposition of a closure boundary on financial-state changes; the trial-balance check IS the standard diagnostic that the law has been correctly applied. The shape recurs in mass-balance analysis in chemical engineering, energy audits in mechanical engineering, traffic counts in transportation planning, atom counts in chemical equations — every domain where a quantity is conserved across some boundary uses the same structural diagnostic to check that local-looking changes have global accounting consistency.
Antoine Lavoisier’s Traité Élémentaire de Chimie (1789) established mass conservation in chemistry: in a closed system, the total mass of reactants equals the total mass of products. Lavoisier demonstrated this with careful sealed-vessel experiments — most famously combustion experiments that disproved the phlogiston theory by showing that combustion added mass (in the form of oxygen) rather than releasing a substance.Before Lavoisier, the apparent mass loss in combustion (wood burning to ash) and mass gain in oxidation (metals calcining) seemed to contradict each other; phlogiston was the theoretical patch. By insisting on sealed vessels and quantitative measurement, Lavoisier showed both phenomena were instances of the same conservation law operating on an unseen gas. This is the founding example in the catalog because it shows a conservation law operating in a domain (chemistry) where the conserved quantity (atoms-as-mass) was not yet known — yet the conservation could be measured before the substrate was understood.Inference: a quantity that appears not to be conserved is a candidate for an unseen flux. Sealing the system and weighing it before/after is the canonical experimental move to reveal what’s leaving or entering through invisible channels.
Kirchhoff’s current law: total current into a node equals total current out.
Emmy Noether’s 1918 paper Invariante Variationsprobleme proved one of the deepest theorems in mathematical physics: every continuous symmetry of a physical system’s action corresponds to a conserved quantity. Translation symmetry in space yields conservation of momentum; translation symmetry in time yields conservation of energy; rotational symmetry yields conservation of angular momentum; gauge symmetry in electromagnetism yields conservation of electric charge.Noether’s theorem is the structural-justification for treating conservation as a primitive rather than as an empirical accident: the conserved quantities are not separately-discovered laws of nature but mathematical consequences of the system’s symmetry. This formalizes the cross-domain pattern — once you have invariance under a transformation, you get a conserved current “for free” — and is the basis for treating “what symmetry produces this conservation?” as the productive diagnostic in physics. The pattern then exports as a structural primitive into domains (double-entry bookkeeping, mass balance in chemistry, software’s conservation-of-complexity) where the substrate is different but the invariance-produces-conservation logic recurs.
Tesler’s Law: every system has some irreducible complexity; you can hide it but you can’t eliminate it.
reversible-computing limit; Landauer’s principle ties information erasure to thermodynamic cost.
Tesler coined the Law of Conservation of Complexity as an argument for Apple’s MacApp framework: every application has an inherent, irreducible amount of complexity, and the only open question is who bears it — the platform developer, the application developer, or the user. You can move that complexity across the boundary between code and interface, but you cannot make it vanish. Pushing it out of the UI to make the product feel simple pushes it into the implementation; refusing to absorb it in the implementation pushes it back onto the user as a steeper learning curve. (It is sometimes called the “waterbed principle” for exactly this push-down-here, rise-up-there behavior.)The structural mapping to conservation is direct. The conserved quantity is total task complexity; the closed system is “everything required to accomplish the task”; the boundary that matters is the developer/user interface line. Tesler’s design argument is then a conservation argument with an economic rider: since complexity is conserved, deciding where to put it is a distribution problem, and his rule — spend a developer’s week to save a million users a minute each — is a claim that the cost of the conserved quantity differs sharply by who holds it. Unlike physical conservation, the total here is only floor-bounded (you can always add gratuitous complexity above the irreducible minimum), which is what makes the diagnostic useful: complexity that seems to disappear from one side has reappeared on the other.Inference: when an interface gets dramatically simpler with no loss of capability, ask where the complexity went — it has moved to the implementation, the configuration, or the user’s head, not out of existence.
moles in = moles out per element; the design backbone of distillation columns and reactor design.
billiard collisions, rocket propulsion, gyroscope dynamics.
Landauer’s 1961 paper established that information is physical: erasing one bit of information has an unavoidable thermodynamic cost. The argument turns on logical irreversibility — an operation whose output does not uniquely determine its input. Bit erasure is the canonical case: two possible prior states (“0” or “1”) collapse to one known state, so the operation has no inverse. Landauer showed that this logical loss must be paid for physically. Erasure reduces the entropy of the information-bearing degrees of freedom by k ln 2, and the second law forbids that entropy from simply disappearing, so it must be dumped into the environment as heat of at least kT ln 2 per erased bit.This is a conservation law spanning two domains that look unrelated. The conserved quantity is total entropy across the closed system of computer-plus-environment; the boundary is the line between a bit’s logical state and the surrounding thermal bath. What looks like complexity vanishing in the logical world (a bit’s history irretrievably gone) reappears as a debit in the thermal world (heat dissipated), exactly balanced. The same accounting structure that makes double-entry bookkeeping refuse to lose a dollar makes Landauer’s principle refuse to lose entropy: the local-looking erasure is forced to balance globally. Its corollary — that logically reversible operations carry no such minimum cost — opened the field of reversible computing and is what finally exorcised Maxwell’s demon, since the demon must eventually erase its own memory and pay the bill.Inference: any apparent destruction of information has a physical price somewhere; if you can’t find where the entropy went, you’ve drawn the system boundary too small.
opportunity cost; budget constraints; conservation of trust / credibility across decisions.