Vāstu - The Wall or the Circle: 32 Gates Two Different Ways
Vāstu divides the boundary of a plan into thirty-two positions, and the doorway is read by which one it occupies. The 32 Gates sets out what those positions mean and which are favourable.
This article is about something that piece leaves alone: there are two different ways of finding the position, they are used interchangeably, and on a square plan they do not agree. The disagreement is not a matter of interpretation or of lineage. It is geometry, and it can be computed exactly.
The classical division is of a wall
The classical thirty-two are the outer ring of the eighty-one-pada grid: cells lying along the four sides of a square, seven between each pair of corners plus the four corners themselves. They are found by dividing a side into equal parts.
Bṛhat Saṃhitā 52.42 ff. is explicit about the procedure. Take the length of the wall the door sits in, divide it into equal parts, and read off which part the door falls in. The Vāsturatnākara worked example follows the same rule: an eastern wall of some twenty-seven hasta is divided by nine, and the door's pada is whichever ninth it lands in.
Nothing in that operation involves an angle. It is a measurement taken along the wall with a cord, and it would work identically on a plan with no centre marked at all.
The modern division is of a circle
Modern applied Vāstu finds the same thirty-two a different way. Stand at the centre of the plan, divide the full 360° into thirty-two equal sectors of 11.25°, and the sector the door falls in is its gate. The sixteen zones are the same construction at 22.5°, and the arithmetic is consistent: 32 × 11.25° = 16 × 22.5° = 360°.
Practitioners who work this way are taught a tight tolerance — an error of five degrees is nearly half a gate, and one or two degrees is the permissible slack. That precision is real, and it is precision about a quantity the classical texts never mention.
They are not the same partition
Take a square plan and put the centre where the diagonals cross. Divide one wall into eight equal lengths, as the classical rule directs, and ask what angle each of those equal lengths subtends at the centre:
| Position along the wall | Angle subtended by an equal eighth |
|---|---|
| Segments 1 and 8 — at the corners | 8.13° |
| Segments 2 and 7 | 10.30° |
| Segments 3 and 6 | 12.53° |
| Segments 4 and 5 — at the mid-wall | 14.04° |
The eight equal lengths subtend angles from 8.13° to 14.04°, a spread of 1.73 to 1. Not one of them is 11.25°. They sum to 90°, as they must, but they are unequal, and they are unequal in a fixed pattern: narrow at the corners, wide at the middle of the wall.
Run the construction the other way and the same divergence appears from the other side. Equal 11.25° sectors, swept from the centre, cut the wall into lengths of 0.199, 0.215, 0.254 and 0.332 of the half-side — a spread of 1.67 to 1. And the 22.5° boundary, which a practitioner would take to mark the half-way point of a wall's quadrant, lands at √2 − 1 = 0.4142 of the half-side. Not at 0.5.
Where this bites
The two methods agree exactly at two places on each wall: the corner and the mid-point. Everywhere else they diverge, and the divergence is largest in the segments nearest those two agreement points — which is a counter-intuitive result worth sitting with, because it means the error is not small near the landmarks a practitioner uses to orient.
Put plainly: a consultant marking eight equal segments along a wall with a tape and a consultant sweeping 11.25° sectors from the centre with a compass will assign the same door to different padas across most of its possible positions. Both will be confident. Both will be following a rule they were correctly taught.
Which one governs
This article does not settle it, and it should be said clearly that the question ought not to be settled by preference.
What can be said is what each rests on. The linear division is what the classical texts describe, in the words they use, with a worked example that divides a wall by nine. The angular division is what the modern applied literature teaches, with a stated tolerance and a coherent internal arithmetic, and it has the practical advantage of being measurable with an instrument a consultant already carries. It is not a corruption of the classical rule; it is a different rule that produces thirty-two of something.
What follows for anyone reading a plan is narrow and practical. A reading should say which construction it used. Two readings of the same house that place the door in different padas are not necessarily in conflict about Vāstu — they may simply have measured two different things, and neither will know it unless the method is stated.
An open question
Whether any classical text describes an angular division of the plan is, as far as this project's sources go, unestablished. The texts consulted here describe a linear division and nothing else. That is not proof of absence, and if an angular warrant exists it would change the picture entirely — so it is recorded as open rather than answered.
The figures above were computed for a unit square with the centre at the intersection of the diagonals, using exact trigonometry, and are reported with the construction that produced them.
Measure and Number in Vāstu is a five-part series in the OccultSanctum library. Our daily tools — live panchang, muhurat and chart work — are in Vedic Samay, free on the App Store and on Google Play.