Nuraghi as Ritual Monuments in the Sardinian Bronze and Iron Ages (circa 1700–700 BC)

1 Introduction

The great number and diversity of well-preserved buildings of the nuragic period in Sardinia (circa eighteenth–eighth centuries BC) provide an unusually clear view of Bronze and Iron Age (IA) architecture in a western European context. The most prominent are the nuraghi (singular: nuraghe) themselves: large stone towers that were sometimes enclosed in walled bastions and compounds. Notwithstanding a long history of study and discussion, the purpose and significance of these striking buildings have always been contentious. Religious or funerary functions, sometimes advocated in antiquarian works, receded to the fringes of academic discussion for most of the twentieth century as residential and defensive explanations became dominant. While the earliest, Middle Bronze Age (MBA), nuraghi (circa seventeenth–fourteenth centuries BC) are sometimes regarded as the dwellings of “an egalitarian society of farmer-pastoralists” (Blake, 2015, p. 96), the larger towers are central to an influential model of social organization and change that identifies them with the elites of an increasingly hierarchical or ranked society in the later Bronze and Iron Ages (e.g. Lilliu, 1982; Webster, 1996, 2015). Nonhierarchical reconstructions of later nuragic social structure have also been ventured (Araque Gonzalez, 2014, 2019), as well as the hypothesis of a sociopolitical crisis towards the end of the Bronze Age involving a challenge to the old order, which had been centred on the former dominance and power of the towers (Perra, 2020b).

While most authors subscribe to the proposition that nuraghi were essentially dwellings, a certain willingness to diverge from traditional explanations about their defensive and elite nature has been accompanied by new fieldwork and excavations, synthesized in recent monographs (e.g. Cossu, Perra, & Usai, 2018; Moravetti, Melis, Foddai, & Alba, 2017). Most notably, it is generally believed that some towers began to incorporate cult activities towards the end of the Late (or Final) Bronze Age (circa tenth century BC) and, more obviously, in the Iron Age (tenth–eighth centuries BC) or later. In contrast with cult buildings such as well shrines, however, few scholars have proposed religious or ritual uses for the towers ab origine or would go so far as to call them temples. Burgess (2001), Mulas (2012), and Pittau (2013 [1st ed. 1977]) are notable exceptions, though not often cited. Somewhat detached from the mainstream of archaeological literature, explanations for nuraghi have also been sought in the field of archaeoastronomy (e.g. Atzeni, Garau, & Mura, 2016b; Zedda, 2015). Some of these works appear to have a wider following among nonacademic readers, who are most likely to ask about the function of the towers or “what they were for” in disarmingly direct, albeit sometimes reductive, terms.

This study reviews these old controversies with reference to the materiality and conceptual place of “ritual,” an ill-defined term in archaeology, but one with an established tradition of usage (e.g. Bradley, 2005; Insoll, 2004, 2011; Stephenson, 2015). It serves here as a catch-all for a range of activities connected with cults and religious beliefs or other forms of symbolic behaviour, inferred from artifacts and buildings. Using only archaeological evidence to distinguish between these and more or less prosaic activities, which might also be ritualized, is often difficult, however, especially in prehistory.

This study begins with a brief review of three prominent and often interconnected hypotheses about the main uses of nuraghi: those of ordinary dwelling, elite residence, and defensive or fortified building. I devote more attention to a fourth proposition, however, which does not exclude the other three, but places more emphasis on a ritual dimension. I argue that this dates back to the earliest phase of tower building in the Bronze Age, rather than to a late, essentially Iron Age, transformation of nuraghi and nuragic society. While this view contrasts with that of most authorities, it has occasionally been advanced in various but not always convincing ways by others (e.g. Atzeni et al., 2016b; Mulas, 2012; Pittau, 2013). One aim of this study is to challenge interpretations of nuraghi as secular monuments, rather than to disavow them or claim a universal explanation for such a large number and variety of towers in every period. Indeed, separated from other functions, a “ritual” explanation would not be very enlightening. In order to illustrate the main points, I refer to a selection of better-known sites (Figure 1, for most of their locations).

Figure 1

Map of Sardinia with main sites mentioned in the text.

2 Chronology and Typology

Several studies estimate that there are currently circa 7,000 extant nuraghi, out of an original total of 8,000 or 9,000, and that they first appeared in the MBA (Bronzo Medio in Italian), probably between 1700 and 1600 BC (e.g. Usai, 2015, p. 60; Vanzetti et al., 2013). Construction continued in at least the earlier part of the Late Bronze Age (henceforth LBA), also known as the Recent Bronze Age (or Bronzo Recente, henceforth RBA), circa 1350–1200 BC. It is possible that very few towers or none at all were built in the Final Bronze Age (or Bronzo Finale, henceforth FBA), circa 1200–950 BC (Vanzetti et al., 2013, p. 97), although many contain material of the Early Iron Age (circa 950–750 BC). While the persistent importance of nuraghi at this time can hardly be doubted (e.g. Bernardini, 2017, p. 212), their solidity and prominence also increased the propensity for ongoing or intermittent use up until the present day. Their significance to societies of the Phoenician-Punic, Roman and Byzantine periods, when they sometimes hosted cults, burials, and various domestic or agricultural activities, is a complex topic (e.g. Blake, 1998; Stiglitz, 2021; Trudu, Paglietti, & Muresu, 2016), which exceeds the remit of this study.

According to recent estimates, there may have been as many as 6,500–7,000 tholos and 1,000–2,000 complex nuraghi (Depalmas, 2018b), although not more than one or two percent have been excavated. While they vary in details and combinations of formal features, the usual “tholos” tower resembles a truncated cone with a single chamber and a corbelled vault or cupola on each of two or three stories, often provided with recessed alcoves, resembling cupboards, and linked by intramural stairs. Some evidently had a protruding balcony at the summit (Cappai, 2012), though none survive in situ. Complex nuraghi have additional towers and features, often contained within a bastion and a compound surrounded by an external wall. Usually said to result from a process of expansion over several centuries, as in the classic exposition of Su Nuraxi by Lilliu (1955) (Lilliu & Zucca, 1988), it has been suggested that certain multitowered nuraghi, including Arrubiu and Santu Antine, were designed outright as part of a monumental complex (Campus, 2019, p. 16; Manca & De Rinaldis, 2017). Less is known about the smallest towers or “nuraghe-huts” (nuraghi-capanne), which may lack alcoves, stairs, corbelled vaults, and even upper storeys (Usai, 2006, p. 560; 2020; Vanzetti et al., 2013, p. 89; Webster, 2015, p. 47).

Around 400 more divergent structures, commonly termed “corridor nuraghi” or archaic, proto- and pseudo-nuraghi, are characterized by elongated or subcircular ground plans enclosing a narrow passage. They seem to represent an older (MBA) type of monument (Depalmas, 2018b; Moravetti, 2017a) and are not discussed in this study. The term “nuraghe” is used rather loosely, therefore, and does not always designate a tall or massive tower. In consequence, the typological variability of nuraghi increases the likelihood of functional variability as well.

3 Distribution

According to Melis (2017, pp. 48–50), the greatest density of towers by elevation is in areas up to about 400 m above sea level. Around 11% are between 400 and 1,000 m, and less than 2% (circa 68 nuraghi) are at over 1,000 m. The maximum concentration of all types (corridor, tholos, and hybrid forms) is about 1.5–1.7 per km² in certain upland zones of central-western Sardinia, but much lower in some north-eastern and southern areas and around 0.27 per km² on average.

The number, size, and distribution of nuragic sites, which include separate village settlements as well as towers, plus evidence of deforestation (Leonelli, 2017a), suggest population growth in the MBA-LBA in a widely farmed and inhabited landscape. Population density is hard to calculate, since we rarely know if tower utilization was continuous, intermittent, or perhaps even seasonal, which might apply to those in less hospitable locations. Contemporaneity between towers and other buildings is not always clear even in excavated sites (Cattani, 2017, p. 89). Nor can one be sure how many nuraghi were detached from villages, which are often less well-preserved or visible from field surveys. A map of the Sinis-Oristano area shows that only 17 of 72 nuraghi were part of a village, but 29 others were no more than 500 m away and most others within 2 km (Usai, 2006, p. 561, fig. 2, p. 564, fig. 4).

While the distribution of nuraghi suggests some form of territoriality, which could be said of most large monuments with such a conspicuous presence in the landscape, the numbers and spatial patterns vary (e.g. Namirski, 2020b; Usai, 2006, 2018a; Vanzetti et al., 2013). In areas with high densities, the majority often comprise single towers, followed by two-to-four-towered nuraghi and by a small minority of five-to-eight-towered buildings (e.g. Usai, 2006, figs. 2–4; Vanzetti et al., 2013, pp. 105–107, figs. 5–8). From the perspective, or presumption, of a social landscape structured by hierarchical power relations, the complex nuraghi would represent dominant families (e.g. Vanzetti et al., 2013, p. 107). Alternatives are conceivable, however. Some towers are no more than 2–3 km from various small settlements (e.g. Usai & Marras, 2005, p. 200, fig. 1) to which they might have been affiliated, perhaps in a shared arrangement. If complex nuraghi are mainly later in date, they could reflect a growing preference for bigger or enlarged towers. Together with larger villages, this may signify greater social complexity over time, although this could be envisaged in horizontal or heterarchical and vertical or hierarchical terms.

Regional analyses have confirmed the intervisibility and prominence of towers in some areas (e.g. Schirru, 2017) but have not produced a single explanation or consensus about the function of nuraghi or even about the extent to which settlement patterns reflect hierarchical principles of sociopolitical organization (Vanzetti et al., 2013). Nevertheless, interpretation usually hinges on a demarcation or defence of larger areas of land, such as upland plateaux or resource zones, marking routeways, and signalling “cantons” or smaller territorial units by means of a coordinated system of interrelated sites and monuments (Depalmas, 2015, p. 82; Usai, 2018a). Excavations have also revealed that some nuraghi were built on sites occupied in preceding or much earlier periods, such as the Copper Age. While this may be due to the recurrent attractions of certain localities and their natural resources as well as the reutilizable potential of older buildings, additional sociocultural factors and perceptions or constraints may have influenced their distribution.

4 Towers as Ordinary Dwellings

The proposition that nuraghi were dwellings, or places in which to live, work, and sleep, has various permutations. The hypothesis of an elite or chiefly residence has been advocated most consistently with reference to the larger monuments, which have been the main focus of attention and excavation. An alternative or complementary proposal is that the smaller and simpler towers were ordinary dwellings, more especially in the MBA (e.g. Depalmas, 2012, p. 143; 2017, p. 101; Webster, 2015, p. 46). The kinds of artifacts found inside them and the high numbers or density of towers in certain areas lend some support to this view.

Very few smaller nuraghi have been excavated, however. One is Bingia ’e Monti, which stands alone but within 2 km of a “Giant’s tomb” (a monumental chamber tomb with a curved façade), seven more nuraghi, and a little-known village settlement (Usai, 2020, p. 345). It presents a ground floor chamber of about 5 m in diameter, without alcoves or evidence of an upper storey, while the entrance faces northeast onto an antechamber, unlike the more common southerly orientation of classic tholos nuraghi (Figure 2a). The contents include pottery for food preparation and consumption, grinding stones, obsidian, and bones of sheep, goat, pig, and cattle (Palazzo, 2018), while the antechamber also has storage pottery and spindle whorls (Usai, 2020). These relatively ordinary materials, datable mostly to the LBA (RBA1), confirmed by 14 C dates in the fourteenth century BC, seem consistent with a residential function and not especially suggestive of elevated status.

Figure 2

Simplified plans of selected nuraghi. (a) Bingia ’e Monti, (b) Duos Nuraghes, (c) Funtana, (d) Arrubiu, (e) Santu Antine, (f) Nurdole, (g) Mulinu di Cabu Abbas, and (h) Su Mulinu. (Adapted from Fadda, 1990; Galli, 1989; Lilliu, 1982; Lo Schiavo & Perra, 2017; Moravetti, 1988; Saba, 2012; Usai, 2020; Webster, 2001).

Other small and apparently isolated or detached nuraghi, however, seem inconveniently located for residential purposes (e.g. Trump, 1990, pp. 28–32). Even if they contain artefacts consistent with everyday activities, this does not necessarily mean that they were permanent dwellings. A distinction may also be drawn between the towers that were or were not isolated or detached from a settlement, which we might think of as “rural” and “village” nuraghi. Standing alone, the former may have had an enhanced role as territorial or landscape markers, but in this case, instead of a permanent residence, it may have been sufficient to indicate some activity or presence on the land, whether social and economic, or even ritual, in nature.

Based on current evidence, the wider validity and applicability of the ordinary residence theory is hard to evaluate. In its favour, the large numbers and concentrations of nuraghi in several areas do not accord easily with the presumption of high-status dwellings and expose the limitations and relativity of the term “elite” for such a common type of building. As noted below in the case of “village nuraghi,” it is hard to explain why one family would have resided in the tower and all others in more conventional village houses. The apparent absence of nuraghi in some villages is also perplexing. These questions recur with reference to the elite residence theory, as summarized next.

5 Towers as Elite Residences

In many works on Sardinian prehistory, nuraghi are considered primarily to have been homes for those of variably elevated status, from moderately privileged families residing in open countryside or within small communities to powerful chiefly households in the case of the large complex monuments, especially those surrounded by a sizeable village. This interpretation rests on the premise that all nuraghi are residential buildings, regardless of scale or elaboration. Associating monumentality with elites, it appeals to the characterization of a society that was ranked to a limited degree in the MBA but became increasingly unequal and hierarchical in the course of the LBA. It also suggests that nuraghi were privately owned or occupied dwellings, a few of which grew into veritable fortresses over time. Like all theories about nuraghi, however, it presents some weaknesses.

Single towers and classic tholos nuraghi do not conform closely with a modern notion of an elite house insofar as their monumentality derives mainly from thick walls and height rather than opulent suites of rooms. While one must be wary of preconceptions about space requirements, the utilizable floor area relative to overall size is limited. One might compare this with other prehistoric ritual monuments in the Mediterranean, such as Maltese temples, Bronze Age sesi (burial towers) of Pantelleria or Aegean Bronze Age tholos tombs, although they obviously differ in several other ways. Nuraghi have occasionally been discussed with reference to these and other monuments from various ethnographically documented or geographically distant contexts, especially in older literature (e.g. Pinza, 1901). This has not lead to any consensus about function, however, largely because multistoried round towers are uncommon and tend either to have had divergent specialized functions in different periods and places (such as lighthouses, steeples, and windmills) or present analogies with other ancient buildings whose function is also debatable, such as Balearic talayots and Scotland’s Iron Age brochs. Trump (1990, p. 46) and Webster (1996, p. 97; 2001, p. 122) likened nuraghi to the tower houses of Medieval Tuscany or southeast Europe with particular reference to competitive display and warfare. Apart from structural differences, these later buildings provided much more living space, especially in their upper stories. Likewise, the larger Scottish brochs can exceed 100 m² of floor space and could easily have accommodated 20–30 people, if not more (Geddes, 2006, table 5). Most single-towered nuraghi offer less than 30 m² (below), suggesting no more than single-family occupancy.

In order to illustrate certain difficulties with residential explanations, I will briefly review some evidence from the site of Duos Nuraghes, so-called due to the slightly unusual, but not unique, presence of two adjacent towers (Figure 2b). These contained material dated from the MBA to the Middle Ages, while the surrounding village is ascribed primarily to the LBA-IA (Webster, 2001). This permits some comparisons between them, mainly with Tower A, which had the more informative deposits. The excavator likens both nuraghi to “a fortified agro-pastoral farmstead, incorporating the roles of tower house, livestock barn and repository of valuables,” inhabited by one or two families who, at least by the LBA, were of a more “elevated status” and “with better access to animals, metal tools and status insignia” than their fellow villagers in the surrounding houses (Webster, 2001, pp. 121, 125).

Tower A, the simpler of the two, lacked alcoves and intramural stairs. The northerly orientation of its entrance is also more comparable with Bingia ’e Monti (above) than a classic tholos nuraghe. The ground floor room, about 4.4 m in diameter or 14.5 m² internally, was probably the only one. The lower room of Tower B was around 26.5m², while the upper storey was perhaps half that size, giving a total of around 37 m² (Table 1). By comparison with the village houses, the lower room of Tower A is well below average size, while that of Tower B is slightly above. The complementary room functions (such as sleeping and cooking) postulated for the village, however, suggest that each residential unit comprised at least two rooms (e.g. 6 and 3, or 2 and 4; Webster, 2001, p. 126) (Table 1). Their combined floor spaces would range between 30 and 43 m², which is about the same as Tower B. The two nuraghi could only be regarded as more spacious than a village house if taken together as one twin-towered residence, with a combined floor space of about 50–55 m² (Webster, 1991, p. 179).

At several other sites, the lack of any size advantage in single towers also transpires from comparison with houses comprising several rooms arranged around an open space (e.g. Cattani, 2017; Depalmas, 2017, pp. 103–104). The contrast is especially striking with large multi-roomed residential units, which also differ in conception and layout from nuraghi, as exemplified at Serra Orrios (Moravetti, 1998b). Some single tholos towers, moreover, have extremely limited floor space, such as Biriai, where the lower chamber is a mere 2.5 m × 3 m in diameter or about 10 m² (Salis, 2020, p. 507).

The elevated status of the Duos Nuraghes towers is also questionable with reference to their contents. Tower A and the village produced similar pottery, stone tools, metal items, and animal bone, but the wider range of LBA-IA artefacts comes from village structures (e.g. 2, 3, and 5: Table 2). Even if the ceramic platters from the towers are of superior type (Webster, 2001, p. 125), they also occur in village buildings (3, 5, 9, 12, 14, and 15) with other decorated wares. A sign of differential status might come from finer exotic pottery, but this seems to comprise just two sherds from Tower A and one from the village (Webster, 2001, pp. 49, 67), while most metal and glass items are in the village (Table 2). Only the faunal remains, thought to reflect a superior meat diet (Webster, 1991, p. 181; 1996, p. 134; 2001, p. 125), might support the notion of tower dwellers with certain privileges insofar as a small bone sample from Tower B layer X (MBA) consists of younger sheep or goat than in Tower A, but the sample size hinders comparisons with the village. In short, artefact range and typology do not indicate that Tower A was at the apex of an intrasite hierarchy, although it is possible that meals consumed in Tower B differed slightly in type.

Combustion features also provide activity indicators. A MBA layer in Tower A had fire-cracked stones and possible metal slag, followed by a small hearth (“LBA1”), and two hearths, an oven, copper slag, and a crucible associated with a possible forge in “LBA2” layers (Webster, 2001, pp. 43, 46–47, 51–54, fig. 4.1). The latter leave only 3 m² of remaining floor space in the southeast zone and even less between the forge and the oven, which were subsequently replaced by hearths (Webster, 2001, p. 79, fig. 4.29), suggesting some changes in use. The unpleasantness of living in such a confined space between a forge and an oven all but exclude a normal dwelling, let alone an elite one. Instead, Tower A was surely a workshop for at least part of the MBA-LBA, which would be consistent with its architectural simplicity. Tower B is more enigmatic, although its artefacts and hearths could reflect domestic use. Even in this case, however, a hypothetical association with an emergent family rests mainly on the prominence of the building.

That nuragic towers could host a range of activities or fulfill a number of different functions transpires from various complex nuraghi, most notably from the well-excavated complex at Arrubiu, a massive five-towered bastion enclosing a central tower (A) within a walled and turreted compound (Figure 2d). While Tower A held accumulations of ash and carbon from repeated hearth activity as well as ordinary and storage pottery, Tower C had a slightly different assemblage: small vessels, obsidian blades, saddle querns, hand-stones, spindle whorls, hearths, clay cooking platforms, and traces of unleavened bread, suggesting a working chamber, perhaps incorporating a bakery, as might be consistent with its window slots (Fonzo, 2018; Marinval, 2018; Melosu & Lugliè, 2018; Orrù, 2018). Tower H, which was detached from the main bastion and subsequently incorporated into the outer wall, when several light slots were blocked up, contained ceramic hearths, quantities of pottery and ground stone tools (Orrù, 2020). Tower D also had evidence for cooking from hearths, ovens, stoves and remains of acorns, wheat, spelt, barley, olive, grape, sloes, land snails, marine mussels, sheep, goat, cattle, pig and deer bones (Carannante & Chilardi, 2020; Perra, 2020a). Unlike the central tower, discussed below, these towers seem consistent with workaday activities.

Although few sites permit a comparison or spatial analysis of the contents of towers and adjacent houses, one cannot readily agree with Russu (1999, p. 201) that no differences existed between them. Unusual items have been found inside certain nuraghi, especially in the central towers of larger monuments, which we consider below from a ritual perspective. The possibility of changes over time and contrasting accessibility to the towers add complexity to questions of function. For example, Tower D of Cuccurada, superimposed on an archaic nuraghe towards the end of the MBA, contained grinding stones, pottery, and some charred grain, which prompted a residential explanation (Atzeni, Cicilloni, Marini, Ragucci, & Usai, 2016a). Ragucci (2015) inferred a change to a ritual function in the Iron Age on the basis of a sword fragment and a bronze figurine in a later layer. More recently, however, the excavators have suggested a LBA community venue rather than an elite dwelling (Matta & Cicilloni, 2019, p. 17), a hypothesis that might equally apply to other sites, including Duos Nuraghes.

Likewise, the various manufacturing and storage functions evinced at Arrubiu could be interpreted as serving wider community rather than purely elite interests. This view avoids the assumption, also contested by Araque Gonzalez (2019), that nuragic society comprised large numbers of more or less elite families. Indeed, there is a dearth of clear evidence for high-ranking individuals from burials. While it is possible that some ordinary people were buried in rock fissures and caves (Skeates, 2012) or in archaeologically invisible ways, the most widespread and conspicuous graves are the Giants’ tombs: monumental long cairns or gallery graves, which were evidently for collective burials, sometimes numbering over a hundred individuals of all ages and sexes, while lacking any signs from grave goods of conspicuous individual wealth or distinction (e.g. Usai & Fonzo, 2018).

Presumptions about the restricted use of large nuraghi and about monumentality as an expression of elite status were not often challenged in the older literature. Built to impress, monumental architecture typically exceeds practical requirements, but this can also apply to public buildings (Trigger, 1990). What we might call the architectural rhetoric of complex nuraghi is also of limited or debatable relevance to a residential function. This is partly why they have often been regarded primarily as fortified residences or even Bronze Age castles inhabited by military leaders (Ugas, 2014).

6 Towers as Fortified Dwellings

The idea that nuraghi were built primarily for or with an eye to defence is deeply rooted in the literature, not just in the pioneering works of Giovanni Lilliu (e.g. Dyson & Rowland, 2007, pp. 58–60; Moravetti, 2017a, p. 26; Ugas, 2014; Usai, 1995, p. 256; Webster, 2015, pp. 110, 140). Many of the terms which have been inherited to describe nuraghi are tendentiously militaristic, as exemplified by: the “keep” (mastio) for a single or central tower; “embrasures” or arrow-slits (feritoi) for intramural slots or narrow windows; “guard-room” or “sentry-box” (garitta) for entrance-corridor alcoves; “machicolations” or slots through which to drop missiles onto attackers (piombatoi) for various shafts; “bastion” (bastione) for complex nuraghi; “antemural” (antemurale) for compound walls; and “parade grounds” (corte d’armi) for external courtyards or compounds. Some of these terms are more prominent in the older archaeological literature, however, as many archaeologists no longer take them literally.

The linguist Massimo Pittau (2013) was one of the first and most outspoken critics of the use of military terms for nuragic architectural features, which he interpreted differently: “embrasures” were light or ventilation slots, “machicolations” were observation holes, and “sentry-boxes” had ritual functions. His work, which is rarely cited, provided effective, if somewhat polemical, counter-arguments to the martial hypothesis. Writing in the 1970s, Pittau (2013, pp. 70–75) conceded that one or two towers may have been incorporated into large complexes whose defensive capability was enhanced by the addition of a perimeter wall, but defence, he maintained, was never the primary purpose of nuraghi. He saw the defensive theory as a projection into the past of anachronistic or naive views of military techniques and tactics, arguing that nuraghi, and especially the archaic corridor forms, would have been hopeless traps in which to seek refuge from attack. Most would have been too small to hold a village community even if it were prepared to abandon surrounding houses and possessions to besiegers, while their topographical locations and associations are inconsistent with effective networks of towers acting as forts or military bunkers.

A more nuanced view tends to redefine defense in terms of deterrence through an assertion of power or territorial dominance (e.g. Trump, 1990, p. 47; Usai, 1995, p. 257), which mitigates Pittau’s objections by emphasizing the symbolic significance of nuraghi, although it does not explain why so much effort should have been expended on a building of little real defensive value of this particular design. Already critiqued by Pinza (1901, pp. 242–244), some writers in the 1990s began to abandon or modify the defensive hypothesis and the implication of an island beset by warfare (e.g. Burgess, 2001; Depalmas, 2006, p. 568; 2015, p. 82; Russu, 1999, p. 201; Trump, 1992, p. 199). Lo Schiavo and Perra (2018) have also noted little direct evidence for conflict from recently excavated nuraghi.

Usually linked to notions of defense is the idea that nuraghi were watchtowers or signalling towers, akin to nineteenth-century semaphores. Plainly, any tall building might serve as a lookout or signalling point, as can natural elevations in the landscape. The roofs or projecting balconies would certainly facilitate observation and communication whether for defense or other reasons. The partiality of this explanation, however, stems from the fact that some nuraghi are not on high ground with the best views of surrounding land, and it does not account for the architectural elaboration and contents of the towers, which may be more relevant to other primary functions.

It is well attested that nuraghi could be used for storing and protecting goods (e.g. Usai, 1995, p. 258; Usai & Marras, 2005, p. 188). This downplays the need to grant them an active role in warfare or to explain their peculiarities as residential structures and emphasizes an economic role instead. Apart from alcoves and upper chambers, the towers of complex nuraghi occasionally have presumed silos built into them (e.g. Orolo: Moravetti, 1998a, p. 246) or located in adjacent courtyards (e.g. Arrubiu: Perra, 2017a, p. 92; Namirski, 2020a) (Figure 2d(s)). The insulation provided by thick walls might also have helped preserve some foods, although the range of artefacts found in towers does not point to a single or specialized type of storage. Rather than their main purpose, storage was more likely just one concern and does not solve the conundrum that buildings for domestic, defensive, or religious use might all incorporate storage facilities or functions for similar or different reasons. It also raises the issue of whether storage was oriented to collective or communal rather than restricted or elite usage.

Despite shortcomings and criticisms, the defensive hypothesis has proved resilient, especially in the case of complex nuraghi, bolstered no doubt by an impression of unwelcoming austerity and by an undeniable potential for defence, however limited, of any solid building surrounded by thick high walls. Martial analogies were perhaps also favored by the prominence of war, borne of experience, in the minds of many twentieth-century archaeologists and by the widespread perception of the European Bronze Age as a time of growing conflict and emergent warrior elites. Nuragic weaponry and figurines of armed men could be cited in support, even though they probably date mainly to the Iron Age. In recent years, however, several authors have begun to de-emphasize the bellicose connotations of the bronzetti (e.g. Salis, 2018, p. 252) and consider small-scale, short-lived, episodic, and localized raiding more likely than endemic warfare (e.g. Araque Gonzalez, 2018, pp. 28–29; Lo Schiavo & Perra, 2018). Even stalwart proponents of the fortress theory (e.g. Ugas, 2014) concede multifunctionality, noting that some large nuraghi incorporated cult activities in the Iron Age, a purported change of use, which is often associated with a sociopolitical transformation of late nuragic society.

7 Ritual and Cult Associations of Nuraghi

7.2 Architectural Perspectives

Discussions of ritual and cults usually pass over the towers in favour of other types of building (e.g. Cossu et al., 2018; Depalmas, 2014; Lo Schiavo, 2006), sometimes divided into functional or stylistic categories (e.g. Moravetti et al., 2017). This may encourage the notion that sacred and secular are separable and that nuraghi were “civil” buildings (e.g. Lo Schiavo, 2009b, p. 397). The Giants’ tombs (tombe dei Giganti), sacred wells (pozzi/fonti sacri), circular shrines or “rotundas” (rotonde) and megaron temples (templi a megaron) are more readily classified as ritual buildings, characterized by monumentality, elaborate or unusual design, and refined masonry, sometimes exploiting different properties or colors of stone, whereas domestic or secular architecture is identified with residential structures employing relatively ordinary materials and techniques.

Such distinctions are often blurred, however. While architectural refinement along with bronze figurines and offerings help us to identify the sacral aspect of water shrines, these structures also connote social or civic utility, which would have served to augment their credibility and significance as ritual monuments. Residential contexts may also include ritual activities and installations, as in the case of chambers containing large round stone basins at Su Nuraxi and elsewhere, often interpreted as sweat rooms (e.g. Burgess, 2001, p. 175; Paglietti, 2018, pp. 55–57; Salis, 2018, p. 257; Usai, 2013, p. 187), unusual buildings with niches and alcoves such as TC1 at Serucci (Santoni, 2010, p. 9), and the better-known “meeting houses” or “council chambers” (capanna di riunione, sala del Consiglio) discussed below.

If the presumed residential or defensive functions of nuraghi have encouraged a secular or civic categorization, their monumentality and sophistication should also link them with ritual. Circularity is common to both. The more compelling analogies are between the tholos nuraghi and the corbelled subterranean vaults of well shrines (Figure 3f) (e.g. Salis, 2017, 2018). Isodomic masonry is another feature shared by wells and some, generally more complex, nuraghi (Paglietti, 2015, p. 286). Amongst the Iron Age rotundas of undoubted cult use, Su Monte at Sorradile (Figure 3d) is especially like a nuraghe in plan, with its thick walls, triple alcoves, and a large central basin to which a model tower was attached (Santoni & Bacco, 2008). The allusion to nuraghi is equally clear from the sanctuary altar of S’Arcu ʼe is Forros, which resembles the protruding parapet of a tower, topped by a ritual hearth (Fadda, 2017, p. 228). The unity or convergence of secular and sacred is also exemplified by the so-called council chambers, which are considered “civic-ritual” structures with a political and community orientation on account of their large size and other unusual finds and features (e.g. Moravetti, 2017b). They present various analogies with nuraghi; wall niches and perimeter benches resemble those of, for example, nuraghi Nurdole (Fadda, 1990), Santu Antine (first floor), Santa Barbara, and Funtana (Figure 2c and f). A circular platform in the centre of Nurdole and Funtana is very like that of the council chamber at Palmavera, which supported a model nuraghe (Moravetti, 1988, p. 50; 1992). Another analogy is the ambulatory corridor encircling the lower rooms of nuraghi Santu Antine and Leortinas (Figure 2(e)), and also encountered in the circular cult building of Su Romanzesu (Fadda & Posi, 2006; Moravetti, 1988, pp. 47–48), which may evoke a ceremonial space.

Figure 3

Architectural features of nuraghi (in schematic form) found in different types of ritual buildings: (a) nuraghe, composite model; (b) megaron (Fadda, 2017, p. 242); (c) meeting house (Moravetti, 2017b, p. 151); (d) rotunda (Moravetti et al., 2017, p. 408); (e) Giant’s tomb (Melis, 2017, p. 75); and (f) well shrine (Salis, 2017, p. 269).

Megaron and well shrines elicit additional comparisons with complex nuraghi due to their walled compounds. A good example is Domu de Orgia at Esterzili (Figure 3b), which includes a circular chamber in the outer wall (Fadda, 2017, pp. 241–246). While these are easily explained as symbolic boundaries, like the periboloi or temenoi of classical antiquity, those around nuraghi are usually regarded as defensive (Salis, 2015, pp. 137–138), and yet some conceptual overlap or duality is perfectly possible insofar as the latter could be symbolically protective or exclusionary in nature.

Use of colour to lend distinction and attract attention also links nuraghi and cult buildings. Paler limestone or sedimentary rocks and darker volcanic rocks are sometimes juxtaposed in well shrines, cult buildings, and various nuraghi, such as Alvu, Su Nuraxi, and Genna Maria (Paglietti, 2015; 2018, p. 34; Salis, 2017, p. 261). Traces of a red pigment comprising ochre and animal glue on plaster were identified recently on the inner wall of the main chamber at Santu Antine (Campus, 2019, p. 67). Apart from the many ritual uses of ochre in prehistory, this has analogies with red plaster in older cult monuments, such as the prehistoric “pyramid” of Monte d’Accoddi in Sardinia (Contu, 2000, p. 45) and the walls of some Maltese temples (e.g. Skeates, 2010, pp. 161, 192).

Even Giants’ tombs, with their prominent façade and forecourt, present some analogies with nuraghi, exemplified by those with stone blocks in regular courses. The characteristic central monolith (or “stele centinata”) is often taken as a symbolic door to another world (e.g. Antona, 2018, p. 241; Melis, 2014, pp. 76–77), but other interpretations are possible. The small aperture at the base, the superimposed compartments carved in relief, and the domed top, could symbolize an entrance, chambers on different levels, and the summit cupola of a building, rather than the top of a door. The same design appears on those carved into rock (Figure 3e), perhaps evoking a concept of turriform monumentality, similar to that of a nuraghe or even a sacred well. The allusion could also be metaphorical or metaphysical if, for example, the various compartments refer to different spheres or stages of existence.

Some architectural elements incorporated into the walls of certain towers, which must have come from much older, probably Neolithic, ritual monuments, also deserve mention. Two large menhirs (monoliths) were prominently redeployed as lintels in nuraghi S’Ortali ʼe Su Monte and Rodas (Fadda, 2012; Melis, 2016). The unusual form of the chamber of Rodas lead Melis (2016, p. 47) to note that this was perhaps not just a case of casual or expedient reuse, but a way of emphasizing or asserting a connection with a much older ritual or cultic building at this location or elsewhere in the landscape.

To sum up, we can see similar architectural forms and techniques recurrently deployed in nuragic towers, meeting houses, cult buildings, and tombs. Apart from the tombs, which share some features with earlier types of burial, the towers have precedence in terms of their greater antiquity and numerical prominence. The other cult or civic-ritual structures were used intensively during the FBA and EIA, although ritual activities at certain wells are now back-dated to the RBA (Paglietti, 2015; Salis, 2018, p. 250). The tholos nuraghi, therefore, are a possible, or likely, source of inspiration for circular cult buildings; perhaps not simply by virtue of their monumentality or venerable age, but because they were also identified with ancient ritual activities.

Finally, a ritual use of towers has also been suggested from observations of their cardinal orientations and potential astronomical alignments, interpreted by various authors in different ways, which merit further discussion. To this reader, the idea that nuragic peoples were so preoccupied with detailed stellar observations that they decided, for reasons otherwise inexplicable and impractical, to locate certain nuraghi in the landscape around Santu Antine in imitation of a map of the Pleiades (Mulas, 2012, pp. 203–233) seems unlikely and only weakly grounded in selected data. The claim that a few fenestrated towers associated with complex nuraghi would have facilitated temporal or calendrical observations thanks to their window-slots, with reference primarily to solar movements or solstices (Atzeni et al., 2016b), is perhaps not easily refuted or proven. These authors also regard the light effects and shapes produced by solar rays passing through spaces above door lintels or apertures in the ceilings of some nuraghi (although it is not clear if these were originally open) as intentionally created. The statistically better-documented preference for southerly orientations of the entrance passages of nuraghi is explained by Zedda and Belmonte (2004) (Zedda, 2015) in terms of an alignment to the rising sun at the winter solstice or a lunar standstill. They are cautious in their claims, however:

We do not believe that these orientations had an actual astronomical use (i.e. the nuraghes were not “observatories” or markers), but rather they had a ritual character, including in the design of the monument the orientation to one, or various, phenomena that would have been significant for the builders… However, we do believe that this discovery argues against the view that the Sardinian nuraghes were fortresses, for one does not need to orientate a defensive building astronomically unless it hides at the same time a strong ritual component in its design. (Zedda & Belmonte, 2004, p. 104)

While concurring with this perspective, the reasons for a building’s orientation may well result from an entanglement of environmental factors, local traditions, and beliefs. In early Italic religion, for example, sectoral divisions of the sky and cardinal orientations are often concerned with augury, portents, and divination (e.g. Stevens, 2009). Recurrent orientation is not in itself particularly informative. The doorways of British Iron Age roundhouses, for example, are sometimes preferentially oriented to the east or southeast, but the reasons for this are contentious and not necessarily cosmological (Pope, 2007).

A brief description of the central tholos towers from an embodied perspective might go as follows. The smallish door and thick wall emphasize the distinction between exterior and interior, which is unusually well insulated from sound, smell, and light, enhancing the potential for privacy and sensory manipulation. The entrance alcove and symmetrical arrangement of alcoves in the main chamber, with its unexpectedly high vault, lend further distinction to a building that reveals little of its internal complexity from the outside. Progressively smaller cell-like chambers on the upper storeys increase the sense of seclusion, detachment, or restriction, further conditioned by a rather disorienting winding staircase and by changing levels and scale. Nevertheless, points of reconnection with the outside world, or with people in the courtyard below, are sometimes possible by means of an upper storey window and, more dramatically, by the open view from those towers with a roof terrace or a wraparound balcony, like a minaret’s şerefe or a crow’s nest. The compounds, courtyards, passageways, and extensions of complex nuraghi offer an enhanced experience, which entails crossing a succession of thresholds and spaces of generally diminishing size in order to access the main tower, thereby emphasizing its centrality and venerability. Secondary or peripheral towers, which are subordinate in height and prominence, were allowed numerous window slots, exempting them from the stricter rules governing the central nuraghe. With a nod to Gell (1992), we might call this an architecture of enchantment and ritual rather than just power and utility. A systematic study from a phenomenological standpoint, with an established methodology (e.g. Hamilton et al., 2006), has yet to be developed for nuraghi but could add much to our understanding.

8 Conclusion

One aim of this study was to review interpretations of nuraghi as essentially prosaic monuments within the context of a long history of discussion, including those older or minority views, which are rarely mentioned in recent mainstream literature. Another was to expose or reiterate (since I am not alone in doing so) weaknesses not only in narrowly utilitarian explanations of nuraghi but also in some claims about cult functions. Old controversies about the secular or ritual nature of nuraghi still resonate in current scholarship notwithstanding approaches that stress their inseparability in European prehistory (e.g. Bradley, 2005). They have some analogies with doubts about the secular and ritual nature of prehistoric buildings in the Near East (e.g. Banning, 2011), and in the Aegean where, for example, the ritual and ceremonial role of Minoan and Mycenaean “palaces” is now highlighted, and associations between Iron Age cults and elite residences or multifunctional buildings are debated (e.g. Bintliff, 2012, pp. 132, 228, with references). Like nuraghi, Scottish broch towers have often been assumed to represent a high security residence for a dominant family, although their defensive efficacy has also been questioned and alternative explanations admitted: communal buildings with economic and sociopolitical functions, venues for meetings, feasts, and celebrations, also providing storage space for goods held in common and, more generally, buildings to which a range of people from the surrounding village or area had access (e.g. Armit, 2003, p. 98). The ritual dimension of brochs may also merit more consideration than it has received in the past.

We do not know if or how Bronze Age Sardinians used the word “nuraghe,” which is nevertheless of pre-Roman origin. Turriform monumentality no doubt contributed to a sense of Sardinian cultural identity (Blake, 1998, 1999), although an architectural form lacking any powerful communitarian and, I would suggest, ritual significance, would lack the social relevance or agency that most likely underpins its longevity and importance. The cylindrical or conical design was adaptable, however, and subjected to considerable variation and experimentation. The remarkably large number of nuraghi and their typological variability, which resists an easy or clear-cut classification, would suggest variability of function and significance.

In the case of relatively modest single towers or single rooms, lacking alcoves or tholoi, especially those in rural or non-village settings, essentially utilitarian explanations, already current in the literature, are credible. They would include primarily residential uses and places for storage, closely linked with local agropastoral practices, and a probable workshop in the case of Duos Nuraghes A. Their symbolic significance, especially in rural contexts, is undoubtedly relevant to forms of territoriality concerning rights of land use and access. A ritual or religious dimension to these buildings is not obvious on present evidence, which is very limited however, due to the bias of research towards larger monuments.

The classic tholos nuraghi with additional storeys and a subsidiary tower or two would augment the scale of these facilities and functions, although the significance of those at the centre of a village probably differed from those in detached rural settings. By comparison with the surrounding houses, they have usually been considered special dwellings for elites, although this term is not well defined and might, for example, include ritual specialists. The theory has certain attractions insofar as any aspiring community leaders may well have wanted or needed to be associated with such prominent monumental buildings and the organization of labour required to construct them. Nevertheless, doubts emerge about the adequacy of a purely residential hypothesis in that, for example, some of their contents as well as their architectural and design peculiarities align them with other cult buildings. Nor can one rule out the proposition of a venue for wider community use or benefit.

These considerations also apply to the largest complex nuraghi, which are much less common, except that their accommodation of multiple functions was massively enhanced by forecourts, compounds, and additional or subsidiary towers and structures in which several specific and spatially differentiated activities (artefact production, storage, food processing and consumption) took place. The more prominent and architecturally distinctive central towers placed at the heart of these elaborate complexes are also those in which ritual aspects are more frequently detectable or inferable in multiple forms, including (a) a stricter adherence to a recurrent design and use of architectural features and embellishments, many of which have analogies in other cult-like buildings (notably wells and circular shrines); (b) design features, such as a roof parapet, that seem redundant from a strictly residential perspective; (c) the storage or hoarding of unusual items, especially metals, some of which (e.g. figurines and long swords) are also found in other cult contexts; (d) the presence of unusual metal artefacts inserted into the fabric of the masonry; (e) structured subfloor depositions of vases with ritual connotations; (f) items of ritual furniture (altars, stools) paralleled in other specialized or “civic-ritual” buildings; (g) signs of ritualized feasting and fragmentation of artefacts; and (h) close physical proximity to, or incorporation of, water wells or cisterns, which can have cultic associations.

It is possible, nevertheless, that even the simplest nuraghi possessed an intrinsic ritual quality by virtue of, or in proportion to, their similarity to more elaborate monuments. Models of nuraghi include both single and multitowered forms. Without excluding an identification with deities or concepts of the divine, at least by the end of the Bronze Age, I prefer to avoid designating them as temples due to this term’s particular and often exclusive association with culturally specific religious practices and am wary of assuming a theistic orientation (however common in classical antiquity) rather than other religious forms. In any case, they would have provided oversight for different activities and concerns, from the prosaic to the spiritual. The central towers would represent the main focus of authority and ideological support for activities at least superficially materialist or economic in nature, which are more closely associated with the adjacent towers and surrounding structures. From this perspective, several nuraghi of even intermediate size are no less deserving of interpretation and designation as “civic-ritual” buildings than the “meeting houses” or “council chambers,” with which they share some features.

Having drawn attention to the earliest items to which a ritual function could be ascribed, I cannot deny that the evidence is more abundant for the late nuragic period (late FBA-IA), when sanctuary sites with a multiplicity of ritual monuments were major centres of ceremonial activity (e.g. Ialongo, 2013). Nevertheless, I find the theory of old nuraghi changing from originally “secular” functions, such as a residence or fortress, into monuments of cultic or religious significance, unsatisfactory or at best unproven. It rests partly on arguments from silence or an absence of more eye-catching signs of ritual in earlier phases, which are often less well documented. A reliance on bronze figurines and metal items can also mislead, since they are everywhere more abundant in FBA-IA contexts. Equally constraining is the assumption, or false expectation, that “ordinary” materials imply domestic, as opposed to ritual, buildings, or spaces, and that the two are separable. MBA rituals may have made more use of everyday items, even cooking pottery, which occurs in tombs. Like bread and wine consumed in church, their significance and performative role would have changed according to the context and occasion.

The idea that nuraghi were suitable monuments for conversion due to their age or identification with venerable ancestors, a form of explanation more generally critiqued by Whitley (2002), is debatable. A continuing sense of their sacral associations is no less likely. It is legitimate to argue, therefore, that the ritual dimension of nuraghi goes back to their origins, that it was an intrinsic feature of at least the more architecturally elaborate examples, designed to elicit feelings of awe and enchantment, which enhanced their power and authority. This interpretation would favour long-term continuity in nuragic culture, at least in its beliefs and rituals, at the core of which were the ever-present nuraghi.

From a broader theoretical perspective, even if one emphasizes the ritual significance of nuraghi, divergent frameworks of interpretation or explanation will undoubtedly persist, from a Marxist perspective or premise of domination, control, exploitation, and manipulation, to more functionalist views of social utility, including mechanisms for reducing tension, promoting group solidarity, cohesion, and motivation (e.g. Renfrew, 1994). Ritual may well provide a useful or necessary means of legitimation, regulation, and mediation, required in most societies regardless of the extent to which it is grounded in or reinforces structured inequality. While not disputing the common linkage between ritual and the role of ideology in establishing and reinforcing social positions and political authority, one might also consider its power to promote change rather than merely preserve traditions or be passively reflective. In the western European Bronze and Iron Ages, the Sardinian archaeological record offers perhaps the greatest potential for this branch of research.