Why base 12 is the minimal sufficient arithmetic · the dozenal polygon · what changes when we stop counting on decimal fingers
Before algebra, before geometry, before zero — there was division. Take a whole. Cut it into equal parts. Which cuts are clean? Which leave an infinite tail of remainders?
This is not an abstract question. A mason cutting stone, a merchant dividing goods, a calendar maker splitting a year — all of them need divisions that terminate. A fraction that does not terminate in your base is a division you cannot complete. You have to stop somewhere and accept the error.
Decimal arithmetic — base 10 — is a historical accident of anatomy. Ten fingers. Two prime factors: 2 and 5. That is all it buys you. One third is 0.333… in base 10 — a fraction that never closes. So is one seventh, one ninth, one eleventh. More than half of the simple fractions of everyday measurement are non-terminating in the base we inherited.
A fraction \(1/n\) terminates in base \(b\) if and only if every prime factor of \(n\) is also a prime factor of \(b\). This is the complete criterion.
In base 10: prime(10) = {2, 5}. So 1/3 does not terminate (3 ∉ {2,5}). 1/6 = 1/(2·3) does not terminate. 1/7, 1/9, 1/11, 1/13 — all non-terminating. The base 10 arithmetic we use daily cannot cleanly express the thirds, sixths, sevenths, and ninths that appear constantly in geometry, music, and physical measurement.
Base 12: prime(12) = {2, 3}. Adding the factor 3 is the single change that matters most. Now 1/3 terminates as 0;4 (dozenal notation), 1/4 as 0;3, 1/6 as 0;2. The six denominators that appear most in physical division — 2, 3, 4, 6, 8, 12 — all terminate cleanly.
Twelve has six divisors: 1, 2, 3, 4, 6, 12. No number smaller than 12 has six divisors. This is the minimum — the smallest base that simultaneously contains the arithmetic of halves, thirds, quarters, and sixths.
The six divisors of 12 are precisely the denominators that matter in practical measurement: the half note and the whole, the compass rose (N/S/E/W plus the diagonals), the hour and the clock face, the egg carton, the inch ruler. These are not coincidences — they are the consequence of living in a geometry where circles divide naturally into halves, thirds, and quarters, and 12 = LCM(1, 2, 3, 4) is the smallest integer divisible by all four.
That is the arithmetic identity at the base of base 12. It is not a coincidence of history. It is the minimal closure of the four most basic divisions of unity.
The table below shows which simple fractions terminate in base 10 (decimal) versus base 12 (dozenal). A highlighted ✓ marks the first base where each fraction terminates.
| fraction | factors | base 10 | base 12 | decimal form | dozenal form |
|---|---|---|---|---|---|
| 1/2 | 2 | ✓ | ✓ | 0.5 | 0;6 |
| 1/3 | 3 | — | ✓ | 0.333… | 0;4 |
| 1/4 | 2² | ✓ | ✓ | 0.25 | 0;3 |
| 1/5 | 5 | ✓ | — | 0.2 | 0;2497… |
| 1/6 | 2·3 | — | ✓ | 0.1666… | 0;2 |
| 1/7 | 7 | — | — | 0.142857… | 0;186A35… |
| 1/8 | 2³ | ✓ | ✓ | 0.125 | 0;16 |
| 1/9 | 3² | — | ✓ | 0.111… | 0;14 |
| 1/10 | 2·5 | ✓ | — | 0.1 | 0;12497… |
| 1/11 | 11 | — | — | 0.0909… | 0;1111… |
| 1/12 | 2²·3 | — | ✓ | 0.0833… | 0;1 |
Base 12 gains: 1/3, 1/6, 1/9, 1/12 — all multiples of three. Base 10 gains 1/5 and 1/10. For geometry, physics, and music, the gains from base 12 outweigh those from base 10 by a significant margin. The fifth (1/5) and tenth (1/10) matter in decimal coinage. The third (1/3) and sixth (1/6) matter everywhere else.
The regular twelve-sided polygon — the dodecagon — is the geometric expression of base 12. It is the polygon whose symmetry group is generated by a 30° rotation, and it contains, as inscribed regular polygons, every shape whose divisibility lives in base 12.
The dodecagon contains:
All three inscribed polygons — the shapes that arise in crystallography, tiling, and the geometry of close-packing — live naturally inside the dozenal polygon. There is no coincidence here. The dodecagon is the base-12 object; its inscribed polygons are the divisors of twelve, expressed geometrically.
The dm³ framework rests on a small set of critical thresholds. In base 10, most of them are non-terminating. In base 12, several clean up immediately — because the geometry of the system is built on thirds and halves, not fifths.
| constant | meaning | decimal | dozenal | note |
|---|---|---|---|---|
| ε₀ = 1/3 | fold singularity (F-operator) | 0.333… | 0;4 | terminates in one digit |
| τ = 2 | global attractor (dm³) | 2 | 2 | same in all bases |
| μmax = −2 | Lyapunov bound | −2 | −2 | same in all bases |
| 1/6 | one-sixth threshold (hex symmetry) | 0.1666… | 0;2 | terminates |
| 1/4 | quarter-turn (rotation gate) | 0.25 | 0;3 | terminates (shorter) |
| 1/12 | one dodecal step | 0.0833… | 0;1 | terminates in one digit |
| 72 (E₆ roots) | minimal vectors, E₆ lattice | 72 | 60 | six dozens exactly |
| 51840 (|W(E₆)|) | Weyl group order | 51,840 | 26,000 | 26 gross exactly |
The fold singularity \(\varepsilon_0 = 1/3\) — the central constant of the dm³ F-operator — which produces an infinite decimal expansion in base 10, is exactly 0;4 in dozenal. One keystroke. The geometry was always dozenal; the base-10 notation was obscuring it.
More telling: 72 (the number of minimal vectors in the E₆ root lattice) is 60 in dozenal — six dozens, written with the same two-digit elegance that Babylonians used for their base-60 sexagesimal system. The Weyl group order 51840 is 26,000 — twenty-six gross, perfectly divisible by 12³.
The shift to base 10 is recent in the history of measurement. The Babylonians used base 60. The Romans used twelfths — uncia, the twelfth of an as, is the origin of both "inch" and "ounce." The English foot is twelve inches. The dozen and the gross (twelve squared = 144) persisted in commerce long after decimal notation arrived in Europe, precisely because they are more useful for division.
We did not lose base 12 because it was worse. We lost it because the French Revolution standardized the decimal metric system in 1795, and administrative uniformity won over arithmetic convenience. The meter is easier to define; the foot is easier to divide.
Base 12 is the minimal sufficient base — the smallest number that terminates halves, thirds, and quarters. But "minimal sufficient" raises a natural question: what comes next? If we add the factor 5, we get base 60. If we add 7, we get base 420. If we keep extending — adding one prime at a time — we climb a ladder.
Each rung of that ladder is a different arithmetic universe. Each adds exactly one new prime to the terminating set, at the cost of an increasingly large base. The trade-off sharpens quickly: the marginal value of each new prime falls, but the base grows without bound.
The Prime Ladder — and what you find at the top — is the subject of WP46.
§6 above tabulates ε₀ = 1/3 — the F-operator fold singularity — as terminating in a single dozenal digit, 0;4, against an expansion that never closes in base 10.
WP66 §7.2 is where that observation becomes load-bearing. ε₀ is the parameter WP44’s Open Problem O1 asks to be calibrated; the calibration has never been performed for any nodal set; and the Andes–Amazon corridor is the first case with measured precursors against which to attempt it. A constant that accumulates rounding error from its first digit matters most in the one place where nobody has yet fixed its value.
This is also why this paper belongs to the WP39–45 climate arc rather than beside it. Its surface subject is arithmetic. Its §6 is a table of that arc’s constants, and the claim it makes about them is a claim about precision in the parameter deciding whether the fold fires.
Arc membership corrected in WP66 §1.