Each day, from a given place, if I can see an identifiable star rising, it will always seem to rise over the same point on the distant horizon. (This will be on its eastern half. It will culminate due south of the pole for anyone living in the northern hemisphere. And in view of the context, this qualification need not be repeated.) If it can be seen setting, the star will similarly always seem to set over a fixed point on the western horizon. If the star is sufficiently important to me, I might choose to remind myself of those points of rising and setting, perhaps by such irregularities as hill-tops or isolated trees; or I might choose to mark the directions in which they lie by setting up pairs of posts or other markers relatively near at hand. I should not have to revise the alignments of such markers materially during my lifetime, unless I wanted extreme accuracy of a sort that need not be considered here. (The word ‘alignment’ will usually be used here to refer to two or more terrestrial objects lined up on a celestial object, and not exclusively to sets of three or more terrestrial objects in line, which is an unnecessarily narrow archaeological usage.)
I might choose to direct my buildings—say the main axis of my church—in the same way. Reversing the order of discussion, however, is a hazardous undertaking: the fact that the orientation of someone else’s church happens to produce an alignment with an astronomically interesting event does not necessarily imply that the orientation was deliberate. Deciding between deliberate and accidental alignments is one of the central problems of this book.
Just as with the stars, I may notice the Sun rising at a recognizable place on the horizon, but in this case, as the days go by, that place will seem to change. In midsummer, the Sun in the eastern half of the sky will rise over its most northerly point of the horizon. It will attain its most northerly point of setting on the horizon’s western half at the same season. I could mark these directions as I did those of the rising and setting star; and as the year progressed and the days shortened, I should notice that those distant points of the Sun’s rising and setting move southwards, and that in midwinter they reach to their furthest points south. Again I could mark those southern extremes in one way or another. The markers (both near at hand, or one near and one distant) would then be aligned on four critical phenomena, namely midsummer and midwinter risings and settings of the Sun.
FIG. 1. The directions of the rising and setting Sun at Stonehenge, around 2000 BC, at the times of summer and winter solstice. The directions for the equinoxes are also marked, but are barely distinguishable from the (broken) east–west line. In all cases it is supposed that the upper limb of the Sun (first or last glint of the Sun) is seen on the distant natural horizon.
Approximately halfway between the directions of sunrise at precise midsummer and midwinter (that is, at the solstices), is the true direction of east; and true west is similarly more or less mid-way between the extreme directions of setting. The actual sizes of the angles depend on various factors, and in particular on the geographical place (or more precisely the geographical latitude) and on the irregularities of the actual horizon. In Fig. 1 the angles are drawn for Stonehenge at a nominal date of 2000 BC. The angles are not precisely divided into equal parts by the east–west line, for reasons explained more fully in Appendix 2.
As early as Neanderthal man—say thirty or forty millennia ago—there were burials aligned accurately east–west, which suggests that some or other celestial body was in the thoughts of those responsible for organizing the rituals of death. A grave excavated at L’Anse Amour, in Labrador, incorporated what were evidently ritual fires arranged to the north and south of the body, which was laid in an east–west direction. The east–west and north–south lines seem to have been key directions in the placement of later burials in many parts of the globe, but—religion apart—how is this tendency to be interpreted? East and west are the directions of the rising and setting Sun at the equinoxes, but they are not easily established, and the positions of the fires might rather be thought to suggest that in the Labrador case the critical directions were north and south, the line having perhaps been decided by the Sun’s midday position. A body with head to the north might have been regarded as lying towards the pole, the region where stars do not move. Granted more sophistication, east and west might have been regarded as midway between the Sun’s extremes of rising and of setting. Alternatively, the four cardinal points of the compass might have been settled not by reference to the Sun but to the daily rotation of the stars: a star culminates (reaches its highest point) on the meridian, just as does the Sun, and the meridian also bisects the directions of a star’s rising and setting. Culminations are not easy to settle precisely, since the altitude of the Sun or star is changing least rapidly then; but this does not mean that culmination was not uppermost in the thoughts of those who chose these directions for burials. Then again, at various periods of history certain bright stars have risen and set due east and west, so that alignment might have been on them. Skeleton directions that have so often been interpreted in solar terms can all too easily be reinterpreted in numerous ways, without our presupposing any particularly sophisticated techniques of observation. Which of these alternatives should one favour?
FIG. 2. The absolute extreme directions of the rising and setting Moon at Stonehenge, around 2000 BC, assuming that the Moon is fully visible, and just touching the natural horizon.
The evidence, based not on skeleton positions (which are often dubious) but on the forms of tombs and other structures, is that all of these ways of considering the cardinal points of the horizon, east and west, north and south, are likely to have been familiar in late Neolithic Europe. It is all too easy to become hypnotized by the idea of observation of the Sun and to forget the stars, but there is strong evidence from the period before the first phase of Stonehenge that observation of the stars was then important, perhaps even more important. In some early cultures from which written records survive—in Egypt, for instance—the direction of north was significant, and was found from observations of the stars circling the pole, or of a particular star near the pole at that time. (This was not the star that now serves us as the Pole Star, which in the remote past was well removed from its present position.)
The directions of the Moon’s places of rising on the eastern horizon and setting on the western also change with time, but the pattern of change is much more complicated than in the case of the Sun. The details are put aside for the time being (they are treated more fully in Chapter V and in some detail in Appendix 2), but again there are four absolute extremes of direction, just as with the Sun. The angle separating the northern and southern extremes of the Moon’s rising and setting is greater than in the case of the Sun. The angles in question, which depend as before on several factors, actually fluctuate in the course of time in a way that at first seems erratic. Alexander Thom and others have suggested, however, that the pattern of change lent itself to an analysis of quite extraordinary penetration by the people of the Bronze Age, or even earlier. For the time being, Fig. 2 will suffice to give an idea of the absolute extremes of lunar direction at the latitude of Stonehenge.
The earliest written astronomical records—notably the Egyptian, Babylonian, and Greek—reveal a preoccupation with risings and settings and periods of visibility generally. They show a concern with what was to be observed at the horizon, and with intervals of time between special events in the heavens, and their recurrences. This is not to say that there was necessarily a concern with directions towards points of rising and setting, for there are other ways of using horizon observations. Consider, however, a passage from a Mesopotamian astronomical text compiled early in the first millennium BC and known as MUL.APIN:
The Sun which rose towards the north with the head of the Lion turns and keeps moving down towards the south at a rate of 40 NINDA per day. The days become shorter, the nights