Why Do Temples Face a Particular Way? Buildings Aligned to the Sky

A building has an orientation whether anybody intended one or not, so the question is never whether a structure points somewhere but whether the direction was chosen and why. Deliberate alignment to the sky is well established for some structures, with sunlight entering a chamber only on particular days, with sightlines to the horizon marked by stones, and with construction texts describing the procedure. Alignment can equally follow from the terrain, from the shape of the site, from the direction of a road, from an existing building or from a convention within a culture that had nothing astronomical behind it. Distinguishing those requires more than measuring one building and noticing that it points at something.

The candidate targets are numerous, which is part of the difficulty:

• •Sunrise or sunset at the solstices, which are the commonest and most convincing cases
• •Sunrise at the equinoxes, which is harder to determine observationally than people assume
• •The extreme risings of the moon, which follow a cycle of about 18.6 years
• •The rising of a bright star or a recognisable group
• •Cardinal directions, established by the sun or by the pole
• •A sacred direction unrelated to the sky, including towards a city, a mountain or a shrine

The field has a serious problem with false positives and the reason is statistical. Given a building, a wide horizon and a long list of possible targets including solar, lunar and stellar events, some alignment will be found by chance, and the looser the tolerance allowed the more certain that becomes. A single structure aligned to something is therefore weak evidence. What makes a claim strong is a population, meaning many structures of the same type showing the same alignment where chance would produce a scatter, which is exactly what has been demonstrated for several groups of monuments and what has failed to appear for others. Precession also matters, since the rising position of a star shifts over centuries, so any stellar claim requires the date of construction and gives a different answer for a different date.

Establishing an alignment properly requires more than a compass bearing and the procedure is worth describing. The direction is measured by survey rather than magnetically, since magnetic north moves and differs from true north by an amount that varies with place and date. The horizon altitude in that direction must be recorded, because a sun rising behind a mountain appears at a different bearing from one rising over a flat sea, and ignoring that is a common error. The date of construction must be established independently, since the sun's solstice positions shift slightly and stellar positions shift considerably over centuries. And the point from which to measure must be justified, since a long building offers many possible sightlines and choosing the one that works is exactly the error the whole exercise must avoid.

Several alignments are supported well enough to be uncontroversial. A passage tomb in Ireland admits sunlight along its passage to the chamber only within a few days of the midwinter solstice, which is a designed effect of a structure over five thousand years old. Egyptian temples show orientations clustering on solar events and on the cardinal directions, with construction rituals describing a stretching of the cord to establish the alignment. Many Christian churches face east, which is a documented convention with a stated theological reason. Islamic buildings face a specified direction determined by calculation, and the methods for finding it drove substantial work in mathematical geography. In each case the evidence combines measurement with a text or a population rather than resting on one building.

Every building has an orientation, so the question is whether a direction was chosen and why, and terrain, roads and convention all produce alignments with nothing astronomical behind them. A long list of possible targets guarantees chance matches, so strong claims rest on many structures sharing one alignment. Passage tombs admitting midwinter sunlight and eastward churches are among the cases that hold up.