Mathematics, Measurement, and Record-Keeping

Difficulty: concepts range from elementary to specialist; adoption and reliable tables, surveys, accounts, and experiments require teaching and checking. Prerequisites: a shared language/notation and people with time to learn; durable records additionally need writing materials, agreed units, and custody.

Interactive figure, Fig 9 · Slide Rule Multiply: drag the slide to set a and the cursor to set b, and watch the hairline land on a × b on the D scale, with the slide swinging to its right index (and the decimal point left to estimate) whenever the product passes 10. A code-drawn explanatory model built from this guide's numbers, for orientation only; it is not an equipment drawing or a procedure.

Why it matters

Mathematical ideas travel cheaply compared with furnaces or ships, but teaching, practice, notation, measurement standards, table production, and record custody all consume time and resources. Well-chosen methods can still multiply the value of almost every other guide.

More importantly, record-keeping is what makes knowledge cumulative. A craft tradition can preserve sophisticated knowledge orally and by apprenticeship, but it is vulnerable when transmission breaks. Written, replicated records make quantitative comparison, remote review, and recovery easier; they do not ensure truth, interpretation, access, or preservation.

Part 1: Numbers

Positional notation with zero — the single best idea here

Roman numerals are awkward for written algorithms, so calculation commonly used counting boards, abaci, tables, and trained practitioners. Positional notation makes many written procedures more compact; that advantage should be demonstrated without treating another notation as evidence of mathematical incompetence.

Positional notation — where a digit's value depends on its position, and a zero holds an empty position — supports compact, teachable written algorithms.

Decimal positional systems with zero developed in South Asia, circulated and developed further through the Islamic world, and reached European users through multiple channels. Fibonacci's Liber Abaci (1202) was influential, but adoption was gradual and uneven because education, commerce, existing tools, notation, trust, and document practices all mattered.

What to teach, in order:

1. Ten digits including zero as a number, not just a placeholder — a genuine conceptual leap that took a long time historically.
2. Place value: units, tens, hundreds.
3. Columnar addition and subtraction with carrying and borrowing.
4. Long multiplication and long division.
5. Fractions, and then decimal fractions (Stevin, 1585) — which make fractional arithmetic as easy as whole numbers.
6. Negative numbers, which make debts and directions calculable.

These tools can simplify commerce, taxation, surveying, and engineering. Learning time depends on prior education, language, teaching, and practice, and durable adoption depends on institutions as well as an explanation.

Related notation worth introducing

• Symbolic algebra. + − × ÷ = √ and using letters for unknowns are all late developments (the equals sign is 1557). Before them, algebra was written out in sentences and was correspondingly painful. Compact notation isn't cosmetic — it's what lets you manipulate expressions mechanically and see structure.
• Decimal units. See navigation and timekeeping. Coherent base-10 measures make many conversions simpler, but units, dimensions, scale factors, and uncertainty still have to be tracked explicitly.
• Scientific/exponent notation for very large and small numbers.

Part 2: Practical geometry

Right angles and squares, without instruments

The 3-4-5 triangle. A triangle with sides in the ratio 3:4:5 has a perfect right angle between the 3 and 4 sides. Make a closed rope loop divided into 12 equal segments, then peg it into sides of 3, 4, and 5 segments. You have a right angle whose practical accuracy is limited by the rope, knots, and how carefully it is tensioned.

Rope geometry has a long history in surveying, and this is one of the most useful facts in construction. Related: when assembling a four-sided frame whose opposite sides are already equal, adjust it until its two diagonals are equal; that verifies that its corners are right angles. Equal diagonals alone do not prove that an arbitrary quadrilateral is rectangular, and a geometric square also requires all four sides to be equal.

Pythagoras generalizes it: a² + b² = c². That one relation gives you diagonals, roof rafter lengths, distances across obstacles, and slope lengths.

Similar triangles — measuring what you can't reach

If two triangles have the same angles, their sides are in the same ratio. So:

• Height of a tree or tower: stand a stick of known height upright, measure its shadow and the tower's shadow, and the ratio gives the height. Thales reportedly measured the pyramids this way.
• Width of a river you can't cross: lay out a triangle on your bank using sightlines to a landmark on the far bank.
• Distance to a ship from a known observation height, if you also measure the angle of depression with a quadrant: horizontal distance is approximately height / tan(angle). Distance to the sea horizon is a different problem that must account for the Earth's curvature.

Similar triangles plus a sighting instrument is the whole basis of surveying and mapmaking.

The values worth memorizing

• π ≈ 3.14159 (22/7 is a good working approximation; 355/113 is accurate to six decimal places and is easy to remember as the digits 113355 split in half).
• Circle: circumference 2πr, area πr².
• Sphere: volume 4/3 πr³, surface 4πr².
• Cylinder: volume πr²h. A cone is one third of a cylinder with the same base and perpendicular height; a pyramid is one third of a prism with the same base and perpendicular height. Those relations matter for estimating volumes of grain heaps, spoil, and masonry.
• √2 ≈ 1.41421 (the diagonal of a unit square) and √3 ≈ 1.73205.

Trigonometry

Sine, cosine, and tangent relate a triangle's angles to its side ratios. With a table and enough independent measurements, you can solve many triangles. Some data are insufficient or ambiguous—for example, the sine-law side–side–angle case can have zero, one, or two solutions—and field work still needs a checked baseline, angle observations, and error estimates.

Compiling and printing accurate trig tables (see printing) is one of the highest-value things a press can produce, because everyone downstream uses them and an error in a hand-copied table silently corrupts every calculation made from it.

Part 3: Calculating aids

Logarithms — the best computational idea before machines

log(a × b) = log(a) + log(b)

Because of that identity, a table of logarithms converts every multiplication into an addition and every division into a subtraction. Roots and powers become division and multiplication.

For astronomers, navigators, and engineers doing long chains of multiplication by hand, logarithms can save substantial work. Tables are costly to calculate, typeset, proof, distribute, and revise; copying or rounding errors can propagate, and users must know the base, precision, and interpolation method. Logarithms were one important tool among arithmetic, geometry, abaci, mechanical calculators, nomograms, and other methods.

The slide rule makes it physical: two logarithmically-graduated sliding scales, so setting one against the other performs the addition mechanically. It's a set of logarithmic scales whose usable precision depends on length, graduation, alignment, eyesight, and technique—often around two to three significant figures for an ordinary rule. It does not track decimal place or validate the model.

Part 4: Record-keeping and accounting

Double-entry bookkeeping

Every transaction is recorded with equal total debits and credits — often in two accounts, sometimes in more — so the books should always balance. An imbalance proves that something was posted or added incorrectly. The converse is important: balanced books can still contain omitted, invented, duplicated, or misclassified entries.

Codified by Luca Pacioli in 1494 (describing Venetian practice), it became a major foundation of commercial accounting in many economies. Its power is threefold:

1. An arithmetic and posting check. Many one-sided entries and addition errors show up as an imbalance; fraud and matched errors may not.
2. It separates capital from income, so you can tell whether a business is actually profitable or is quietly consuming itself.
3. It makes enterprises auditable by outsiders when entries are backed by source documents, cash counts, and physical inventories. That supports investment by people who are not present — a precondition for companies, banks, and large-scale enterprise.

The debit–credit rule can be introduced quickly, but competent classification, closing, reconciliation, audit evidence, and institutional adoption require practice and review. Its value is conceptual and organizational, not a promise of one-day mastery.

Technical records

Write down what you did, in numbers, and what happened. This one habit is the difference between a craft and an engineering discipline.

For every batch — of glass, steel, glaze, dye, mortar — record: the proportions by weight, the source of each material, the temperature and time, the weather, and the outcome. Number the batches. Keep the failures.

Change one variable at a time — at first. It is excellent for diagnosing a known process and mapping the effect of one factor. It is not a complete method for optimization: two variables may interact, so the best setting of one can depend on the setting of another. Once a process is repeatable, compare a small set of combinations, repeat each combination, and randomize their order where possible. That separates factor effects from drift, weather, operator learning, and chance.

The test tile habit in glazing is the model: small labeled samples, fired alongside real work, with the recipe recorded. It's how you build a body of knowledge that survives you.

Statistics and evidence

• Count things and compare, with domain expertise. Two wards, two fields, two treatments, or two furnace designs can reveal differences, but numbers do not repair biased assignment, unsafe experiments, bad measurement, confounding, or the wrong outcome. Use expert knowledge to design and interpret comparisons.
• Averages and spread. An average can hide distributions, outliers, subgroups, and change over time; report an appropriate measure of variation and the raw pattern where practical.
• A contemporaneous comparison group. Compare like with like, assign cases to groups without favoritism where possible, define the outcome in advance, and repeat. James Lind's 1747 trial — twelve sailors, six treatments, two per group — is a famous early comparison, but its groups were far too small to be a general model for reliable evidence.
• Beware small numbers. Streaks and coincidences in small samples look like patterns. Over-reading a handful of cases is one durable source of false belief.
• Correlation is not causation, and confounding is everywhere. The classic trap: sicker patients get more treatment, so treatment appears to cause death.
• Compound interest. A = P(1 + r)ⁿ. Understanding exponential growth is essential for finance, population, and resource planning — and it is deeply unintuitive to almost everyone.

Part 5: Writing itself

Numbers are useless if you cannot record them, and everything in Part 4 above assumes you can write things down.

Full treatment is in writing and literacy — choosing an alphabet or syllabary over logograms, keeping the sound-to-symbol mapping regular, cheap writing surfaces, and the formatting inventions (word spacing, punctuation, page numbers, indexes) that turn a book from something you read into something you can consult. If the society you're in has no writing at all, go there first; it outranks everything on this page.

Part 6: Concepts worth carrying that need no equipment

• Zero and negative numbers.
• Scientific inquiry: hypotheses, observation or controlled experiment where ethical and feasible, measurement, criticism, publication, and independent confirmation. Replication is especially important for repeatable experiments; rare events and observational evidence instead need transparent records, triangulation, and predictions that can be checked.
• Independent review and accessible publication alongside apprenticeship. Secrecy can obstruct checking and diffusion, while oral craft traditions can preserve sophisticated practice. Records, teaching, incentives, attribution, and appropriate access controls should reinforce rather than caricature the people who maintain the knowledge.
• Conservation of mass — in a closed system, total mass is unchanged, so products plus anything remaining weigh the same as all inputs. An open vessel can gain gases from the air or lose vapor, smoke, and splashes; those flows must be captured or included in the accounting.
• Elements, compounds, and mixtures — chemical elements form compounds that often have definite composition, while mixtures can vary continuously and some solids have composition ranges. Use mass balance and analysis instead of assuming every material has one fixed ratio.
• Powers of ten for estimating. Order-of-magnitude reasoning ("is this a hundred or a hundred thousand?") is a genuinely learnable skill and prevents enormous waste.
• Interpolation — reading between the entries of a table.
• Error and precision — every measurement has uncertainty, and extra digits do not create information. Averaging independent repeated measurements can reduce the standard error of random noise; it does not remove calibration bias, drift, correlation, or other systematic error.

Failure modes

Sources and limits

The experimental-design cautions follow the NIST handbook sections on one-factor-at-a-time experiments and randomized, replicated comparisons. For bookkeeping, HM Revenue & Customs gives the useful warning that double entry does not guarantee accurate records. The geometric rules above assume straight, taut lines and a level or otherwise surveyed reference plane; field error, not the theorem, usually sets the result.

Era notes

Positional notation with zero: India, by the 6th century. Euclid's Elements (~300 BCE) is still one of the finest mathematics textbooks ever written and its axiomatic method is as valuable as its content. Double-entry bookkeeping: Italy, 13th–15th century. Logarithms: 1614. The slide rule: 1620s.

Ranked by value per unit of effort:

1. Positional notation with zero, and the four arithmetic operations.
2. Writing it down — technical records with numbers and controlled comparisons.
3. The 3-4-5 triangle and matched sides plus equal diagonals for laying out square structures.
4. Double-entry bookkeeping.
5. Logarithms and printed tables.
6. Controlled comparison — count things, use a control group, distrust small numbers.
7. Open publication over secrecy — the meta-idea that makes everything else compound.