On the basis of a recently updated data base of Sardinian Megalithic monuments, we perform two independent statistical analyses on several groups of these structures and confirm the existence of differences between the southern and northern halves of Sardinia, indicating that there seem to be different traditions of orientation in these two parts of the island. We then compare the orientation of these groups with other groups of contemporaneous sacred monuments in the central Mediterranean basin. The differences observed in Sardinia could be related to differences in the orientation customs between the north and south shores of the Western Mediterranean.
See for example, ZeddaM.HoskinM.GralewskiR.MancaG., “Orientations of 230 tombe di giganti”, Archaeoastronomy, no. 21 (1996), S33–54, for the tombe di giganti; ZeddaM.HoskinM., “Orientations of Sardinian dolmens”, Archaeoastronomy, no. 22 (1997), S1–16, for the dolmens; ZeddaM.BelmonteJ. A., “On the orientation of Sardinian nuraghes: Some clues to their interpretation”, Journal for the history of astronomy, xxxv (2004), 85–107, for the nuraghi; and BelmonteJ. A.ZeddaM., “From Domus de Janas to Hawanat: On the orientations of rock carved tombs in Western Mediterranean”, in ZeddaM. P.BelmonteJ. A. (eds), Light and shadows in cultural astronomy (Cagliari, 2007), 39–46, for a preliminary study of the Domus de Janas (2005).
2.
Such as the sun, the moon, the Southern Cross area of the sky, etc. Also relevant would be: González-GarcíaA. C.FerrerL. CostaBelmonteJ. A., “Solarist vs. Lunatics: Modelling patterns in megalithic astronomy”, in ZeddaBelmonte (eds), Light and shadows in cultural astronomy (ref. 1), 23–30.
3.
ZeddaM. P., Astronomia nella Sardegna preistorica (Cagliari, 2013). Actually, the review and later publication of this book, with its huge series of new data, has been the trigger initiating the work presented in this paper. The data of these, and other, Sardinian monuments are available on request from the second author.
4.
For similar approaches, see: González-GarcíaA. C.BelmonteJ. A., “Statistical analysis of megalithic tomb orientations in the Iberian Peninsula and neighbouring regions”, Journal for the history of astronomy, xli (2010), 225–38; and González-GarcíaA. C., “Profiting from models of astronomical alignments to unveil ancient cosmologies in Europe and the Mediterranean”, in SprajcI.PehariP. (eds), Ancient cosmologies and modern prophets (Anthropological notebooks, xix (2013), supplement), 49–66.
5.
Each of these typologies is extensively analysed and discussed in Zedda, op. cit. (ref. 3).
6.
The data come from Zedda, op. cit. (ref. 3), Tables 3.3 and 3.4.
7.
As shown in ZeddaBelmonte, op. cit. (ref. 1, 2004).
8.
TandaG., “L’ipogeismo in Sardegna: Arte, simbologia, religione”, in Atti del Congresso Internazionale L’ipogeismo nel Mediterraneo Orogini, sviluppo, quadri culturali (Sassari, 2000), 399–426. Also relevant, although somewhat outdated, is LilliuG., La civiltà dei Sardi, dal paelolitico all’età dei nuraghi (Turino, 1988).
9.
Analysis of the orientation of the tangent lines of the peripheral towers of complex nuraghi showed a statistically significant preference towards the points from which the sun and the moon rise and set at the solstices and the lunistices. See ZeddaBelmonte, op. cit. (ref. 1), and ZeddaM. P., “When astronomical meaning goes beyond orientation and becomes architectural design”, in ZeddaBelmonte (eds), Light and shadows in cultural astronomy (ref. 1), 31–8. It is also important to note that in the more sophisticated nuraghi, there are seldom additional windows in the upper levels of the nuraghe (and the central tower in the complex ones) that keep the same orientation of the gate at groundfloor.
10.
See DepalmasA., “Il Bronzo medio della Sardegna”, “Il Bronzo recente della Sardegna”, and “Il Bronzo finale della Sardegna”, in Atti della XLIV Riunione Scientifica Istituto Italiano di Preistoria e Protostoria: La preistoria e la protostoria della Sardegna (Florence, 2009), 123–30, 131–40 and 141–60.
11.
See Zedda, op. cit. (ref. 3). Also relevant would be also ZeddaM. P., Archeologia del paesaggio nuragico (Cagliari, 2009).
12.
See Zedda, op. cit. (ref. 11), and ZeddaBelmonte, op. cit. (ref. 1). Surprisingly, the geographical distribution of the two dialectal versions of Sardinian (Campidanese and Lugudorese) coincides with the geographic boundary of the two styles of nuraghe, despite both being descendents of the Latin spoken on the island after the Roman conquest in the third century b.c.
13.
The Epanechnikov kernel is optimal given that it does not have tails towards infinity like a Gaussian, where part of the estimate would be diffused. EpanechnikovV. A., “Non-parametric estimation of a multivariate probability density”, Theory of probability and its applications, xiv (1969), 153–8.
14.
A key choice is that of the bandwidth of the kernel function. Depending on the number of monuments, the spread in azimuths and the errors implied in the data acquisition process, one may use larger or smaller values for the pass-band. It is important to note that for other kernels there are theoretical or practical prescriptions, but for the Epanechnikov kernel this is based mostly on a trial and error method.
15.
González-GarcíaBelmonte, op. cit. (ref. 4). See also BelmonteJ. A.González-GarcíaA. C., “The ‘genetic’ analysis of Iberian dolmens: A test of the idea in the central Pyrenees”, Journal for the history of astronomy, xliii (2012), 227–32. Also relevant will be: González-GarcíaA. C.BelmonteJ. A., “Temple orientation across the Mediterranean: A comparative statistical analysis”, Mediterraean archaeology and archaeometry, in press.
E.g. EverittB. S., Cluster analysis (London, 1995). See also: González-García, op. cit. (ref. 4).
19.
In our statistical analysis we consider only azimuths and not declinations. In principle such a choice seems surprising when considering orientations at different latitudes, such as the northern and southern shores of the Mediterranean. However, the choice is justified by the intention to include the effect of the rising and setting directions, which is lost when using declinations. Besides, it should be noted that, for instance, the most extreme positions of the sun at winter solstice would differ by as much as 3½° between latitudes 36° and 43°. The variance is of course larger at latitudes further north, but these lie outside the geographical scope of this paper.
20.
Data on all these series of monuments can be found in HoskinM., Temples, tombs and orientations: A new perspective on Mediterranean prehistory (Bognor Regis, 2001), and references therein. Two exceptions are the Corsican dolmens and the square talayots of Majorca, for these, see: Thury-BouvetG.SantucciJ.-F.KhoumeriE. H.OttaviA., “Orientations of Corsican dolmens”, Journal for the history of astronomy, xxxvii (2006), 299–306; and Aramburu-ZabalaJ.BelmonteJ. A., “On the orientation of the square talayots of Mallorca”, Archaeoastronomy, no. 27 (2002), S67–74. See also the relevant chapters in BelmonteJ. A.HoskinM., Reflejo del cosmos: Atlas de arqueoastronomía en el Mediterráneo antiguo (Madrid, 2002).
21.
For data on southeastern Iberian dolmens, see Hoskin, op. cit. (ref. 20).
22.
For data on megalithic monuments and hawanat in the north of Africa, see: BelmonteJ. A.EstebanC.JiménezJ. J., “Mediterranean archaeoastronomy and archaeotopography: Pre-Roman tombs of Africa Procunsularis”, Archaeoastronomy, no. 23 (1998), S7–24; BelmonteJ. A.EstebanC.CuestaL.Perera BetancortM. A.Jímenez GonzálezJ. J., “Pre-islamic burial monuments in Northern and Saharan Morocco”, Archaeoastronomy, no. 24 (1999), S21–34; BelmonteJ. A.PereraM. A.MarreroR.Tejera GasparA., “The dolmens and hawanat of Africa proconsularis revisited”, Journal for the history of astronomy, xxxiv (2003), 305–20; and SantucciJ-F.KhoumeriE. H., “Orientations of megalithic tombs in Algeria (1): Djebel Mazela and Rocknia Necropolises, and the Kabylian ‘Allées couvertes’”, Journal for the history of astronomy, xxxix (2008), 65–76. For data on the Thracian dolmens, see: González-GarcíaA. C.KolevD. Z.BelmonteJ. A.KolevaV. P.TsonevL.V., “On the orientation of Thracian dolmens”, Archaeoastronomy, the journal of astronomy in culture, xxii (2009), 21–33.
23.
As noted by González-GarcíaBelmonte, op. cit. (ref. 4).
24.
Something already suggested by González-García, op. cit. (ref. 4).
25.
Broadly speaking, we could be talking of a sun-decending/sunsetting, sun-ascending/sunrising and sun-at-the-meridian — or ‘Southern Cross’ — customs, respectively, in agreement with Hoskin’s earlier proposals, op. cit. (ref. 20). Indeed, the actual situation will certainly be much more complicated with the moon possibly playing a role. See González-García., op. cit. (ref. 2).
26.
It has been reasonably argued that the moon played a relevant role in the planning — and the related water cult — of certain important sacred wells. See LebeufA., Il pozzo di Santa Cristina, un osservatorio lunare (Cracow, 2011).
