A collection of Sanskrit texts (c. 800–200 BCE) giving precise geometric instructions for constructing Vedic fire altars. They contain some of the earliest known explicit statements of the Pythagorean theorem, methods for constructing squares equal in area to rectangles or circles, and early approximations of √2. The mathematics is practical and exact — required because the altars had to be the right size for religious rituals to work.
Imagine you need to build a fire altar for a religious ceremony. The rules say the altar must have exactly the right shape and area — if you get it wrong, the ritual won't work. How do you make a rectangle exactly the same area as a square? How do you double the size of a square altar? How do you construct a perfect right angle without a ruler?
The Śulbasūtras — "rules of the cord" — are ancient Indian manuals that answer exactly these questions, using geometry you can do with ropes and pegs. They date to roughly 800–200 BCE, and they contain remarkable mathematical results:
These are not abstract mathematical exercises. They are practical instructions for building altars — which means ancient Indian priests needed to do real geometry with ropes, pegs, and sand over 2,800 years ago.
| Text | Author | Date (approx.) | Vedic school |
|---|---|---|---|
| Baudhāyana | Baudhāyana | c. 800–600 BCE | Taittirīya |
| Āpastamba | Āpastamba | c. 600–450 BCE | Taittirīya |
| Kātyāyana | Kātyāyana | c. 300–200 BCE | Śukla Yajurveda |
| Mānava | Unknown | c. 600–300 BCE | Maitrāyaṇīya |
BSS 1.12: "The cord stretched along the diagonal of a rectangle produces an area which the vertical and horizontal sides together produce."
In modern notation: if a rectangle has sides a and b, then the diagonal d satisfies d² = a² + b². This is the Pythagorean theorem for rectangles, stated in words. The text also lists specific triples:
BSS 1.61–62 gives the diagonal of a unit square (i.e. √2) as:
The actual value is 1.4142135…. This is accurate to 5 decimal places. The formula is not explained — it appears as a statement of fact, suggesting it was derived by some process not recorded in the text.
Adding two squares: BSS 1.50: "If you want to combine two squares into one, make a rectangle with sides equal to the two square sides. The diagonal of that rectangle gives the side of the combined square." In algebra: if we want c² = a² + b², then c is the diagonal of the rectangle with sides a and b — the Pythagorean theorem used backwards.
Squaring the circle (approximate): BSS gives a construction to make a square approximately equal in area to a given circle. The approximation used corresponds to π ≈ 3.088, less accurate than Āryabhaṭa's later value but still a practical construction.
Dating Indian mathematical texts before 500 CE is difficult because manuscripts were copied repeatedly, internal chronological references are rare, and Indian traditional dating does not always align with historical evidence. The Baudhāyana Śulbasūtra is assigned to c. 800 BCE on the basis of linguistic analysis (the Sanskrit style is earlier than texts datable to the 6th century BCE), comparison with Vedic texts of known approximate date, and the absence of features that appear in later texts. The range c. 800–600 BCE is accepted by the majority of specialists (Plofker, Hayashi, Srinivasan). Some scholars argue for dates as late as 500 BCE for Baudhāyana.
The Śulbasūtras state the theorem but do not provide a deductive proof in the Greek sense. Indian mathematical tradition of this period emphasised correct results and reliable methods (upapatti — justification) over the Euclidean axiomatic framework. The question "was this a proof?" reflects a modern, Greek-influenced definition of proof. The statement was known to be true and used systematically — whether that constitutes a "proof" depends on the definition applied.
The value 577/408 is a convergent of the continued fraction for √2. Plofker (2009) notes that the sequence 1/3, 1/12, 1/408 could be derived by successive correction steps (halving the error at each stage), or by a known Babylonian-style iteration method (Heron's method). The text gives no derivation. Whether the value was derived independently or borrowed from earlier sources is unknown. The same value appears in later Indian texts without additional explanation.
The specific Pythagorean triples in the Śulbasūtras overlap with those in Babylonian texts (Plimpton 322 lists (3,4,5), (5,12,13), (8,15,17) triples). This raises the question of independent discovery versus transmission. Current scholarly consensus (Robson, Plofker) is that the overlap is plausibly coincidental — these are the simplest triples and would be discovered by any culture doing systematic construction geometry — but the question is not fully settled.