Introduction

The astronomical interpretation of stone circles and standing stones has a long history; in the late seventeenth century, for example, Martin Martin, writing about the Ring of Brodgar and the Stones of Stenness in Orkney, records that ‘Several of the Inhabitants have a Tradition that the Sun was worshipped in the larger, and the Moon in the lesser Circle’ (1716, 365). In recent years a scientific basis for interpretation of the archaeological evidence in astronomical terms has been proposed by Alexander Thom (Thom, A. 1967; 1971; Thom & Thom 1978). In 1965 he began his introduction to megalithic astronomy: ‘Much has been written for and against the astronomical significance of the stone circles, stone alignments, etc., which are scattered throughout these islands and indeed much further afield. There is, however, universal agreement that the erectors, herein called for convenience Megalithic man, marked the rising and setting points of the solsticial Sun’ (Thom, A. 1965, 1).

Introduction

The valley of Barbrook on Big Moor in Derbyshire was the site. of a major prehistoric settlement (Radley 1966, Burl 1976). The moor contains the remains of a number of stone circles, earthen enclosures, standing stones and cairns, some of which have been shown by Lewis (1966) to have existed in Neolithic and Early Bronze age times. One of the stone circles, Barbrook 1, has been claimed by Thom (1967) and by Barnatt (1978) to have been used by prehistoric man for astronomical purposes. In addition, Thom (1967) has suggested that Barbrook 1 and another nearby stone circle Barbrook 3 (also known as Owler Bar) were constructed according to an accurate geometrical method (Thom type B circles).

In order to make an objective assessment of the astronomical and geometrical claims two projects were carried out:

1. (i) We have accurately surveyed possible sightlines following the criteria suggested by Cooke et al. (1977) and have assessed their significance.

2. […]

INTRODUCTION

The stone circles of Britain have undergone detailed study and statistical analysis over the past decade. The progenitor of this work, Professor A. Thom, claims that these monuments are set out to accurate geometric designs with the use of a standard unit of length, the ‘megalithic yard’ (MY) equal to .829m or 2.72 ft (Thom 1967). This claim has been investigated in two recent statistical analyses which both concluded that only the stone circles from Scotland lent some support to Thom's theories (Kendall 1974; Freeman 1976). Both these studies as well as Thom's used about 200 circles whose diameters were estimated by Thom.

Stonehenge has a history of construction and use extending over more than two thousand years, from about 3100 BC to about 1000 BC or later, on the evidence of radiocarbon dates. As an aid to the understanding of the sequence of building, and of the possible astronomical alignments incorporated in each phase of construction, the principal components of each period are summarised below, with the relevant dates corrected according to R.M.Clark (1975), and are illustrated in figs. 1 and 2. A more detailed description can be found in Stonehenge (Atkinson 1979).

The earliest structures at Stonehenge are three pits found during the extension of the car park. Their excavators (Vatcher & Vatcher 1973) believed them to be postholes for massive tree-trunks, although their form and cross-section differ from those of other excavated postholes at Stonehenge. The late C.A. (‘Peter’) Newham suggested that they could have served as astronomical foresights (Newham 1972); but this must be discounted because pine charcoal from two of them has yielded radiocarbon dates of 7180 be ± 180 and 6140 be ± 140 (L.Vatcher, personal communication).

An international conference on ‘archaeo-astronomy’ is an interesting experience for a historian of astronomy of the traditional school. Even if he is somewhat acquainted with the literature on the subject, and perhaps to some extent has also used it for teaching purposes, he must feel very much an outsider. He is familiar only with certain types of sources in the form of documents or instruments and not used to the strange array of material presented here in evidence of a startling hypothesis, of a prehistoric prelude to an historical account which until now has begun with the earliest written records. The immediate effect is a feeling of perplexity or confusion which only gradually gives way to a set of more structured impressions. Certain parts of the material seem to indicate possible fields in which the possibility of fruitful research cannot be dismissed without further investigation. A number of more or less obvious pitfalls reveal themselves as almost inevitable consequences of the very nature of the material evidence. Furthermore, the dangers connected with any kind of interdisciplinary study become clearly apparent. And last, but not least, one becomes increasingly aware of a series of methodological questions which must be answered if the whole subject is to be salvaged from the rocks, lurking under the surface in the form of arbitrary assumptions, vicious circles, and unjustified inferences.

Of the many megalithic sites claimed by Professor Thom as remains of prehistoric observatories none has proved more contentious than the proposed midwinter solstitial site at Kintraw (A. Thom 1971, pp.37–40). This consists of cairns and a menhir which Thom claims indicates a horizon foresight, the col between Beinn Shiantaidh and Beinn a' Chaolais on the island of Jura, 45 km away. Since the view of this col from the area around the menhir is partially blocked by the intervening ridge of Dun Arnal, Thom suggests that preliminary observations were made from a ledge on a hillside to the north east of the site, and overlooking it. It is the question of the validity of this claim which has produced such a vigorous debate (MacKie 1976, 1981. Patrick 1981).

The subject of statistics has, on the whole, a pretty bad public image. Even among scientists it is usually regarded as a necessary evil, a hoop through which one's data must jump before getting accepted for publication. Archaeoastronomers don't even seem to have got this far, since most papers and talks at conferences seem oblivious of the need for taking any kind of statistical view of the data presented. I very much hope that two recent excellent works (Heggie 1981 a, b) will rapidly change this attitude and I strongly recommend all readers of this article to consult both these references. They make nearly all the points that I should have wished to make, so I need not repeat them here.

We do not need statistics to help assess the evidence for midsummer sunrise at Stonehenge or midwinter sunrise at Newgrange.

INTRODUCTION

The present paper is a rewritten version of a review article (Heggie 1981b) which was designed as an introduction to problems of megalithic astronomy for participants at the conference in Oxford. However the resemblance is largely confined to the title and the structure. Its general purpose here is to provide a background for many of the other papers in the present volume, but the opportunity is also taken to argue the case for the use of a particular methodology in the examination of much of the evidence on megalithic astronomy, and to consider how it may be applied to one or two important bodies of data. Another purpose of this paper is to draw the attention of readers to a number of recent results which have a bearing on the subject and are perhaps not mentioned elsewhere in this volume.