An Answer Written in the Stars

That an astronomical method was used to orient the pyramids received strong, if unexpected, support in the 1980s when historians discovered that among most of the Giza pyramids, the departure of a pyramid's eastern edge from a true north–south line correlated strongly with the accession date of the king for whom each was constructed. Which is to say that the direction of north as determined by the Egyptian method varied systematically as the centuries went by. The ready explanation for this is once again precession of the equinoxes: The early Egyptians must have applied some method of using the stars to find the north celestial pole without realizing that the pole is not fixed, but rather drifts slowly through the heavens.

In November 2000, Kate Spence, an Egyptologist at the University of Cambridge, published a seminal paper in Nature in which she suggested a method by which the pyramid builders determined what they thought was north. She also showed that the resulting orientation errors varied as a function of time—just as predicted by precession. Moreover, by fitting the time–linked precession errors to the slight deviations of each pyramid, she revised their building dates. Instead of 2554 B.C., her data suggest the Great Pyramid was constructed between 2485 and 2475 B.C.

click for full image and caption

The method proposed by Spence involved two stars on opposite sides of the celestial pole. She had to choose them by trial and error, since the pole drifts into different star fields as millennia pass. For the period of interest, Spence found that the stars named Mizar (Zeta Ursa Majoris) and Kochab (Beta Ursa Minoris) would have appeared to revolve around the pole on almost (but not exactly) opposite sides, so that a line joining them would always pass very nearly through the pole. When these two were aligned vertically, the pyramid builders might have hoisted a long plumb line and fixed it at the moment when the two stars both lay on the line. The point where the vertical line touched the ground would indicate north.

One idiosyncrasy of this method was that because these two stars were circumpolar (they never set), they could be seen from Egypt year–round. Thus, at some date during the year Kochab would have appeared above Mizar at meridian transit (when they would have been vertically aligned), but six months later Mizar would have topped Kochab. Early in the pyramid era, the pole was really slightly west (or east, depending on which star was uppermost at the time) of the line. Because of precession, the opposite was true late in the era. Support for Spence's theory came from two pyramids whose deviation from true north was of the expected magnitude but opposite sign. The explanation was that all the pyramids except these two had been set during the time of year when Kochab was above Mizar—these two must have been set six months later (or earlier), when Mizar surmounted Kochab.

Like many groundbreaking papers, this one quickly became the center of arguments and proposed improvements. Spence accepted a small but significant correction by extending the pole displacement to an azimuthal displacement, but she seems not to have been enthused by other proposals to use different stars in a different way. The method still has some practical problems. For one, the plumb line would have to be very long to reach high enough to be seen against the upper star, especially because the observer would need to be far away from the line to achieve sufficient accuracy. And it would have been difficult to see the line at all against a dark sky. Nevertheless, the explanation for the two pyramids with errors of reversed sign supports the basic idea. As centuries went by and the errors grew, later builders may have realized the problem and abandoned the method or used different stars. Thus, the failure of Spence's scheme among later pyramids is not necessarily a valid critique. My own inexpert view is that whether she is proved right or wrong, Spence's basic idea marks a major breakthrough in dating these pyramids.