Debasement — how a standard rots while its name stays the same
A unit of money can be hollowed out two ways that must never be confused. Of the standard: give less metal for the same name — a lighter coin, a short weight, a clipped rim. By alloy: keep the weight, swap the silver for copper. Alloy debasement needs a struck coin that passes by its face; standard-erosion needs only a name and a scale — which is why it is the older of the two. On Enheduanna's baseline there were no coins for ~1,700 years, so her world could only be cheated the first way.
the Enheduanna baseline
AGORA anchors on Enheduanna (c. 2300 BCE, en-priestess of the moon-god Nanna at Ur, daughter of Sargon) by era and script only — she was a priestess and the earliest named author-figure, not a mint official, moneylender, or monetary reformer. The tie is that her world already had the thing debasement attacks: a standard of value, the weighed silver shekel, recorded in cuneiform. And it had no coins — so the only debasement possible in her lifetime was erosion of the standard (false weights, short measure, a drifting mina), never alloy. The first coin you could dilute with copper is Lydian, ~1,700 years later. Honest limit: "the standard can rot as soon as there is a standard" — a claim about the unit, not her biography. Her own works survive only in Old-Babylonian copies ~500 years later, and her authorship is debated; she is a dating anchor, not a preserved source.
two attacks, and which one her age could do
~2300 BCEstandard onlyMoney is weighed silver. No coin exists, so no alloy trick. You cheat the weight: a false stone, a short measure, a mina that quietly drifts. Law already polices it (later, Hammurabi's false-weights clause).
~600 BCEthe coin arrivesLydia strikes electrum stamped to guarantee weight — so you can now count instead of weigh. The stamp also makes a new lie possible: keep the weight, hollow out the metal.
64 CE →the alloy ageRome runs the classic experiment: the silver denarius falls from near-pure to a few percent over three centuries, its silver quietly replaced by copper under an unchanged name and face.
every agefrom below, tooDebasement isn't only top-down. Users clip and sweat metal off the rim — which is why milled, reeded edges were eventually invented. The coin keeps its name; its edge disappears.
machine · the great Roman rot — and why the number depends on how you measure
The denarius is history's best-recorded debasement. But every "silver %" you see is method-dependent: old surface assays read high because minting and corrosion leach copper from the surface, leaving a silver-rich skin (later coins were even deliberately silver-washed). Modern bulk analysis drills a core and reads lower. So this is drawn as a band, not a crisp line — the width is the honesty.
silver-rich (top of band) copper-rich (bottom of band) bulk-core check (Butcher & Ponting)
The method gap, not hidden. Surface figures (Walker, 1970s) run up to ~15 points too high. Nero's post-64 denarius is ~93–94% by surface but ~80% by bulk core; Augustus's Lyon issues are "~95–98%" by surface but essentially unalloyed by bulk. The famous "98% → 5%" span is a rhetorical stretch across two coins (Augustan denarius → Gallienic antoninianus/billon) and two methods. The direction — near-pure silver to a few-percent billon by the 260s–270s — is solid; the exact per-emperor figures are an approximate curve.
machine · debase the mint — and watch Gresham's Law switch on and off
"Bad money drives out good" is real, but it is not a universal law. It needs a mispriced peg: two coins forced by law to be accepted at the same face value. Set an official tariff above the coin's metal worth, then flip the legal-tender peg — watch the good coins flee or come back.
legal tender at par: ON — the law forces both coins accepted equally
in circulation
hoarded · melted · exported
Real mechanism, computed live. When the tariff exceeds the metal (overvalued) and the peg forces par, good full-silver coins are worth more melted than spent, so they vanish — Gresham's Law. Turn the peg off (let people discount the bad coin) and it reverses: the bad coin trades at a discount and the good one stays — the "Thiers' Law" case (Rolnick & Weber, 1986). Gresham himself did not discover this — Oresme, Copernicus, even Aristophanes' Frogs describe it centuries earlier; the eponym is a 19th-century attribution.
machine · two ways to break one coin
Same coin, same name — two independent attacks. One shaves the weight; the other swaps the metal. Only the second needs the coin to pass by its face rather than its mass.
of the STANDARD
Full weight. Full name.
by ALLOY
Full silver. Full name.
The ordering is the point. Standard-erosion (left) works on any weighed money — it is as old as the shekel, ~1,700 years before a coin existed. Alloy debasement (right) can only happen once coins pass by tale, by their stamped face. Enheduanna's age had the left attack available and not the right. The profit from either — the metal the issuer keeps — is seigniorage: the antoninianus tariffed at 2 denarii but holding ~1.5 handed Rome a 25% cut of every coin's face, free.
the reckoning — what's settled, what's still argued
✓ SETTLED
Two distinct debasements: of the standard (less metal per name, by weight — light coins, short-measure, clipping) vs by alloy (same weight, less fineness). Alloy needs struck coins passing by tale.
Standard-erosion is older: weighed-silver money (the shekel) predates coinage by 1,000+ years; the first coin is Lydian, ~600 BCE.
The Roman denarius fell from near-pure silver (Augustus) to a few-percent billon by the 260s–270s — a consensus trajectory. Nero's AD 64 reform is the textbook first debasement; the antoninianus (215 CE) was fiduciary — tariffed 2 denarii, ~1.5 of silver.
Fineness figures differ systematically by assay method (surface reads high via depletion silvering; bulk cores read lower).
Gresham's Law needs two monies and a legally fixed, mispriced exchange rate; its correct form is "overvalued drives out undervalued."
⚑ STILL CONTESTED
Exact fineness numbers — surface vs bulk disagree by up to ~15 points; the per-emperor curve is direction + rough magnitude, not exact assay values.
Debasement → collapse — the tidy "debasement caused the hyperinflation that felled Rome" chain is popular but the quantitative link is debated; "decline" itself is contested framing. A major symptom, not a proven sole cause.
Gresham as a universal law — without the mispriced peg, "good drives out bad" (Thiers) can hold instead. Conditional, not iron.
Did the ancient standard "erode over time"? — no: Bronze-Age weights held a low ~5–6% variation over 2,000 years. The honest picture is many coexisting, imperfect, fraud-prone standards, not one unit that inflated away.
"Inflation" at 2300 BCE is anachronistic (no money-supply lever). Use "price movement / ratio drift." The rich price series (random-walk, war spikes) is 1st-millennium Babylon — ~1,700 years later, a different era.
Three hard flags, kept. (1) Figures vary by method — every silver % is surface-or-bulk; there is no single "true" number, so the chart is a band. (2) No coins on the baseline — alloy debasement is impossible in Enheduanna's age; only standard-erosion applies, and even that was bounded, not runaway. (3) Gresham is conditional — it is routinely oversimplified into a false universal.
Render-not-invent. Sources: K. Butcher & M. Ponting, The Metallurgy of Roman Silver Coinage (Cambridge, 2015) and Butcher, "Debasement and the decline of Rome"; D. R. Walker's surface metrology (the traditional series); Sargent & Velde, The Big Problem of Small Change (2002); Rolnick & Weber, "Gresham's Law or Gresham's Fallacy?" (JPE 1986); Selgin (EH.net) and the Cleveland Fed on Gresham; Ialongo, Hermann & Rahmstorf (PNAS 2021) on Bronze-Age weight stability; Cripps (CDLJ 2017:2) on Neo-Sumerian barley:silver ratios. Where it belongs: the-shekel (the standard it attacks) · money · debt · bitcoin (a standard engineered to resist it).
AVAN's inverse companion:目減り · Meberi — the unit keeps its name and loses its substance; debasement seen not from the mint above but from the hand below, where the coin still says "one" and holds less.