Introduction

1In Webster’s New World College Dictionary (4^th Edition, 2010), ‘bricolage’ is defined as ‘1. the process of making or assembling something from materials at hand; 2. something so made’. The introduction of the term into English language scholarly literature is usually attributed to the French social anthropologist Claude Lévi-Strauss, who in his influential book The Savage Mind (Lévi-Strauss 1962) described how, within traditional cultural settings, mythologies might be created and re-created through the innovative recombination of conceptual elements already at hand, i.e. by drawing upon the culture’s heterogeneous ‘stock’ of material and intangible resources acquired from past experience – ‘oddments left over from human endeavours’ (Lévi-Strauss 1962: 12). Given the demonstrably wide applicability of the concept of bricolage (see below) and the obvious parallel notion of the artefact assemblage, it is surprising that it has so far been given little consideration in archaeology or allied material-based research (but see Liebmann 2013; Brilliant & Kinney 2011; Soar &Tremlett 2017).

2My intention here is to demonstrate the usefulness of adopting a bricolage heuristic by deploying a non-contextual or fractal reading toward an investigative unpacking of complex materialities characterizing two funerary assemblages of the Monte Claro culture/facies of southern Sardinia dating from the mid- to late 3^rd millennium BC. I will then briefly evaluate the relevance of the findings to existing agency-models of extra-insular associations and offer some novel interpretations of these data within their re-contextualized social setting (see Webster & Webster 2017 and Moravetti 2017 for reviews of the Monte Claro corpus).

A Bricolage-Collage Model

3The bricolage concept has been adapted to a staggeringly wide range of practical, conceptual, methodological and critical applications across most fields of inquiry and artistic expression (e.g. Duymedjian & Rüling 2010). In lieu of what would be a lengthy citation, a perusal of, say, Wikipedia makes abundantly clear the breadth of its diffusion into the arts (music, visual art, architecture, literature), as well as in linguistics, culture studies, anthropology, social psychology, philosophy, education, information systems, internet studies and computer science, business, popular culture and fashion, and history, religious studies (Altglas 2014), social enterprise research (Di Domenico et al. 2010), and ancient history (Terranato 1998; Sherratt 2010). For deconstructionists like Derrida (Derrida 1978: 285), every discourse was ‘bricoleur’ in so far as ‘concepts from the text of a heritage which is more or less coherent or ruined’ are borrowed, while Deleuze and Guattari claimed that ‘we are all bricoleurs’ (Deleuze & Guattari 1984: 1, in Altglas 2014: 476). Other recent applications may be seen in Fischer (Fischer 2012), Vanevenhoven (Vanevenhoven et al. 2011), Kincheloe (Kincheloe 2004), Muggleton (Muggleton 2000), Carstensen (Carstensen 2011) and Roberts (Roberts 2018), while a recent critique of the paradigm by an anthropologist has been offered by Melice (Melice 2009).

4In its subsequent applications within various fields of inquiry, Lévi-Strauss’s original analogy of the intellectual bricoleur operating in traditional settings has taken on diverse interpretive guises. Here, I propose a simple heuristic model drawn from the original formulation. While not denying the usefulness of subsequent permutations, the original version seems to me better suited to the constraints of the archaeologist, where inquiry perforce must proceed from and return to material evidence as both source and interrogator of its statements, and because the original formulation provides access points to relevant allied concepts like collage and its rhetoric of re-use. Following Lévi-Strauss too, I accept the dual definition of bricolage as both agency, i.e. in the bricoleur’s actions toward the innovative re-combining of resources at hand (a sort of ‘intellectual bricolage’ [Lévi-Strauss 1962: 11]), and as its outcome, i.e. the characteristically collage-like product that tends to result from bricolage practices (Lévi Strauss 1962: 20). The basic model proposed here addresses the implications of both senses of the concept for archaeological investigation.

Opportunism: making do with close-at-hand

5The defining logic of Levi-Strauss’s notion of bricolage is opportunism – the incidental acquisition of resources toward some possible but undefined future use, where the ‘rules of [the] game are always to make do with “whatever is at hand”’ (Lévi Strauss 1962: 11). The resulting stock, while convenient, is thus always limited. For Levi-Strauss’s intellectual bricoleur engaged in the conceptual work of innovating within a traditional mythological system, such limitations are set both by opportunities to ‘renew or enrich the stock’ of potentially useful conceptual tools and resources, as well as by the finite range of possibilities for recombining in novel ways ‘the remains of previous constructions or destructions’ (Lévi Strauss 1962: 11; see also Chao 1999). In subsequent adaptations of the model, the deceptively simple notion of making-do has been variously understood in terms of, for example, creating newness from existing fragments, of drawing on existing skills and competences, of hybridization via blending elements, of re-applying the prior and already existing, of improvisation, of creating from bric-a-brac; it has also been understood as refurbishment, resourcefulness, problem-solving, opportunity-seeking, ‘resource cooptation’ (Di Domenico et al. 2010: 683) or ‘tinkering through the combination of resources at hand’ (Ciborra 2002: 48-49, cf. Baker & Nelson 2005; Garud & Karnøe 2003).

6Similar interpretive work has been carried out on the notion of ‘whatever is at hand’. In Lévi-Strauss’s (Lévi-Strauss 1962: 11) original formulation, the implied geographical dimension of resources available to the bricoleur was metaphorical and, with few exceptions (such as Owusu & Janssen 2017: 7, Zahra et al. 2009, who relate ‘at hand’ to ‘local’), this sense has generally been maintained as meaning for example low-cost, readily accessible or “given”. In entrepreneurial research specifically, ‘at hand’ is taken to mean something internal to the enterprise or firm, or as referring to ‘an entrepreneur’s idiosyncratic, internal predicates – such as the prior knowledge of markets and ways to serve markets, customer problems, life experiences, educational attainments, professional knowledge, etc.’ (Vanevenhoven et al. 2011: 6; see also Duymedjian & Rüling 2010).

7Finally, the socially contingent and relative status of the ‘at hand’ concept of resource availability has been convincingly demonstrated for example in Altglas’s (Altglas 2014) insightful study of contemporary religious bricolage. Countering conventional rhetorics that view bricolage as a default strategy of the less-empowered subaltern operating under the resource constraints of a penurious environment (Fischer 2012: 1027), Altglas argues that bricolage is equally open to the hand of privilege. In her words, it is also a ‘stratifying social practice, which can be viewed as “empowering” for those who have the dispositions and entitlement to undertake it’ (Altglas 2014: 484). Some, we must remember, will always be relatively better positioned socially to determine what resources are at hand, and for whom. Bricolage then also ‘signifies and makes social distinction’ (Altglas 2014: 488). For the archaeologist, the task lies in identifying the materiality of such conduct as manifested in the evidence for selection, curation, re-combination and re-deployment of constituent parts or fragments toward novel creations within specifically defined social contexts.

The innovative collage

8This review of current thought on bricolage strongly suggests that a prime diagnostic of the bricoleur’s creation is its collage-like character. The term collage, as intended here, is generic and applicable to any work, whether material, conceptual or performative. Lévi-Strauss likened the bricoleur’s innovation to the artist’s collage as work born of creative reassembling, juxtaposing and blending of elements, regarding it as nothing less than ‘the transposition of “bricolage” into the realms of contemplation’ (Lévi-Strauss 1962: 20). Much earlier, the anthropologist Franz Boas had dwelt on the similarly collage-like constitution of native American myths, which to him seemed to ‘have been built up, only to be shattered again, and that new worlds were built from the fragments’ (Boas 1898: 18, emphasis added).

9The collage-to-bricolage analogy, I would argue, is useful to archaeology. Recent theoretical work on collage and related forms of assemblage-art (including theatrical and literary work) suggests fresh avenues for unpacking complex materialities – like archaeological assemblages – that may have arisen from bricolage-like agencies (Soar & Tremlett 2017, Smith & Protevi 2011, Phillips 2006, Venn 2006, Cambre 2013, Kilgard 2009, Heitkemper-Yates 2015, Hildner 1996, Dillon 2004, Vaughan 2005, Bell 2007). Of particular relevance to the present inquiry are approaches following Deleuze (Deleuze 2006, Deleuze & Guattari 1984) and other theorists that view the collage as an assemblage with potentially ‘emergent unities that nonetheless respect the heterogeneity of their components’ (Smith & Protevi 2011 in Cambre 2013: 72) and, following Walter Benjamin (Benjamin 1999), that privilege the critical analysis of constituent fragments over the system (see also Phillips 2006, Venn 2006, Dillon 2004). Studying the processes and procedures by which fragments are collected/recalled, compiled, curated and re-assembled into new objects ‘can reveal insightful biographical details about their creator’ (Kirkland-Robertson 2018; see also Flock 1981).

10In contrast to more traditional functional-contextual approaches, this invokes a deconstructing gaze on the object of study. As nicely expressed by Cambre, ‘collage gathers materials from different worlds into a single (yet multiple) composition demanding a fractally multiplying reading of its elements’ (Cambre 2013: 80). To be clear: in what follows, by fractal reading I mean an investigative unpacking of an archaeological assemblage as one might unpack a piece of collage art by dislodging each component from its pictorial context in hopes of revealing the multitude of microhistories among them (see for example Magnússon & Szijártó 2013). This is an alternative method of “looking” at archaeological evidence. It insists on the at least temporary relaxation of concerns about any functions or meanings emerging from the study of the object or assemblage as a whole. Instead, by disassemblement, it seeks to preserve the integrity of each constituent to allow the possibility of discovering clues to its biography and, in so doing, also an account of the hypothetical bricoleur’s project of discoveries, acquisitions, collections, curations, remembrances, interpretations and re-deployments. In what follows, I will apply this method toward unpacking some complex materialities of late 3^rd millennium BC Sardinia, specifically the tomb assemblages of Matutzu and Gannì.

Applying a bricolage-collage model to Monte Claro culture materialities

11Within what was probably a short period of time in the mid-3^rd millennium BC, traditional Sardinian material culture referred to as Abealzu-Filigosa was to a large extent replaced across much of the island with strongly contrasting expressions in pottery, metal work and architecture, as well as with different settlement patterns and burial practices, resulting in what is referred to as Monte Claro culture (southern facies; Atzeni 2010 see Webster & Webster 2017, Moravetti 2017 for recent reviews). The agricultural lowlands in the south became thickly settled (Webster & Webster 2017: figs. 32 and 55) and the ceramic industry produced a rich array of large jars with up to four handles suggesting bulk storage and transport (cf. Melis 2014: 15). Engagement with eastern cultural spheres was intensified, and within the southern facies of Monte Claro, traditional modes of burial were elaborated or replaced with unprecedented customs paralleling practices in Anatolia and Syria. Equally radical were the changes in the social landscape, judging by the appearance of large residential compounds (with house plans similar to examples in Anatolia), and relatively wealthy child burials, suggesting that some level of social differentiation with limited yet clear family-based elite status distinctions, perhaps ascribed, had attained (Bacarov 2008, Webster 1996: 60, Webster & Webster 2017: 111, see also Manunza 2013: 59, Lai 2008: 126). No confirmed examples of foreign imports have, however, been identified in any southern Monte Claro deposits, making conventional “foreign incursion-invasion-migration” models problematic (Moravetti 2002, 2009, 2017: 202, Manunza 2013: 58, Foschi Nieddu 2000: 809, see also Melis 2000: 142, Cocchi-Genick 2012).

12The clearest expressions of emergent elitism in southern Sardinia during the mid-late 3^rd millennium BC are found in the burial assemblages. The majority of Monte Claro burials appears to have followed more or less traditional modes of primary or secondary interment in hypogea, cists, pits and caves – modes with fairly widespread distribution in the Mediterranean and beyond at this time (Manunza 2013: 58, Cultraro 2011, Cassano & Manfredini 1975, Cammarota 2010, Massa & Şahoğlu 2011: 169, Yilmaz 2006). But with the southern facies of Monte Claro, there appeared also a new burial mode in the form of hypogeal chamber tombs with single access via a vertical shaft, here referred to simply as rock-cut shaft tombs. These have all been found in the lower Campidano plain, among the highest concentration of known Monte Claro settlements (Manunza & Ortu 2012, Manunza 2013: 45-55, Moravetti 2002: 161-66), and they present similarities with examples in Syria. Evidence from the dozen or so that have been investigated suggests that they accommodated the funerary needs of the socially more elevated (Webster & Webster 2017: 110-11). I will here investigate two such tombs through the lens of the bricolage-collage heuristic: the currently best preserved and recorded of the Monte Claro shaft tomb assemblages, Bau Su Matutzu and Gannì. Both are located in the lower Campidano plain near the Gulf of Cagliari (fig. 1).

Bau Su Matutzu

13The tomb of Bau su Matutzu is located near Serdiana, approximately 350 m north of the Riu Bonarda stream and the megalithic fortification at Brunku Balardi. It was excavated in the 1970s as a salvage operation, after agricultural work in an olive grove had caused the collapse of what was found to be a burial vault. The contents of this collective tomb – although partially fragmented by the collapse – were uncovered more or less in situ and documented in scrupulous detail by M.R. Manunza (Manunza 2010, 2012, 2013; fig. 2). Precise dating of the Matutzu burials is somewhat problematic: a radiocarbon (AMS) assay on bone samples from skeleton no. 2 (CEDAD - LTL4197A) returned a date of 4121+-45 years BP or 2880-2570 cal BC (95.4%) (Beckett et al. 2010: 250, Manunza 2010: 111-12, 2012: 614-18, 2013: 53), but the isotope analyst, Luca Lai, raised concerns that this date ‘is likely a few centuries older than the real date due to correction based on suspect d13C’ (Lai, pers. comm. Oct. 2013), and Manunza suggested that skeleton no. 2 had been introduced from another, pre-existing, grave (Manunza 2012: 617-18). On balance, I assume a later 3^rd millennium BC date.

14The tomb was entered via a sub-rectangular shaft with interior ledges/steps, roughly carved from the soft marl bedrock to a depth of 2.73 m, and accessed by a short rectangular tunnel with cramped dimensions of less than a square meter (found still sealed with dry masonry), through a portal with a carved threshold (Manunza 2010: 35-36). The chamber beyond is rectilinear (ca. 6 x 2.5 m with a vaulted ceiling ca 2.50 m high), off which two lateral burial niches/benches open – both ca. 2 x 1 x 1.75 m, raised and vaulted – along with a third in an axial position at the deep end of the chamber, separated by a short, rectangular passage. As in other collective Monte Claro shaft tombs, the central room contained offerings presumably meant as accompaniments both to the individual burials and to the tomb-group as whole. The former included single bowls near the heads of each lateral burial, and a copper dagger and amphora/tankard near the axial burial, the latter recovered below the niche but believed to have been displaced from above. Presumed group accompaniments were found standing and lying more or less in situ on the chamber floor and included two large four-handled jars, two smaller necked jars, two bowls and two tripod braziers (Manunza 2010: 108, fig. 62). All of the vessels were decorated and, as in other Monte Claro burials, the decoration coverage on the larger vessels favoured one exterior face only. Some vessels may have originally contained food or drink, but the absence of any animal bone, shell, etc. is probably significant.

15The contents of the burial niches/benches included the remains of three or possibly four skeletons (Beckett et al. 2010: 249). Although poorly preserved, all were clearly primary inhumations of adults, the bodies laid out semi-flexed on the left side on biers constituted by large sherds from three nearly identical, flute-decorated (on one face only), wide-mouthed pithoi, with original heights of about a meter. Manunza’s detailed records (2010: 43-49) suggest a sequence of interments within a single funerary episode: 1) a first body was installed in the end niche where approximately 1/2 of urn A (the wider of the three, see fig. 7b) was used as a bier and a copper dagger and amphora/tankard were placed nearby; 2) a second body was similarly positioned on a bier on the southeast bench for which a remaining fragment of the same urn (A) as well as part of another urn (C) were used, and a decorated bowl was placed upright by the feet; 3) a third body was placed on the northwest bench using the remaining parts of urns A and C, and a decorated bowl was placed upright near the head; 4) possibly a fourth body was added to the southeast bench where the remaining fragments of urns A and C were used along with sherds from a third urn (B), and a decorated jar was placed upright as accompaniment (alternatively, the last additions represent an extension of the bier for a fourth interment). Whether this procedure was a consequence of practical or symbolic concerns, the sharing of sherds from urn A from the first, rear-niche burial among all the other interments seems significant. This arrangement, in conjunction with the apparent juxtaposition of the amphora/tankard and copper blade (and tripod) to the central rear-niche, led Manunza to the reasonable conclusion that this location held the remains of the highest ranking individual among those interred (Manunza 2010: 108, fig. 62).

Gannì

16This site was discovered in 2009, at Quartucciu some 13 km south of Matutzu near the gulf of Cagliari and within a close radius of several Monte Claro settlements. Two shaft tombs were revealed in a good state of preservation and were excavated and recorded – by Manunza – with assiduous care. Tomb I has been the most informative, and my description here follows closely that in Webster & Webster (2017: 100-102). Based on ceramic comparisons and a single radiocarbon date on human bone of 2469-2293 cal. BC (Manunza 2013: 61, Melis 2013: tab. II), this tomb appears to have been in use slightly later than that of Matutzu. The plan it is also different, and features two vaulted hypogeal chambers with elevated burial niches (fig. 4a). Both chambers were accessed by a common central shaft with quadrangular section and rounded corners (not stepped, as at Matutzu), roughly carved in the soft bedrock to a depth of about 1.5 m below the original ground level. From the somewhat sloping floor of the shaft, two short corridors (ca. 1 m long, 50 x 60 cm w/h), found walled-up with dry-laid stones, lead to north and east chambers.

17The north chamber (T1 in Manunza 2013) is irregularly elliptical with a vaulted ceiling (3.50 x 2.45 x 1.50 m), and around it are four raised burial niches, radially oriented. The three to the south are continuous but demarked by faceted front profiles; the northern niche is separated by a gap in the stone. Rock-cut benches held four inhumations, one of a female and three of males, deposited in crouched positions on their left sides directly on the bench without sherd-biers, in contrast to those at Matutzu. The female (est. age 18-25 years, height 1.55 m) was laid out on the independently cut bench along with a burnish-decorated bowl, while the three males were placed head-to-feet without accompaniments on the continuous bench (est. ages and statures, ind. 1: 18-25 years, height 1.70 m; ind. 2: 30-40 years, height 1.75 m; ind. 3: 20-30 years, height unknown). As at Matutzu, the floor of the chamber also held offerings, here a modest collection comprising a plain plate and a decorated bowl, a ground-stone hand-axe, and four small worked stones, probably tools for applying ceramic decoration.

18The east chamber (T2 in Manunza 2013) has an oval plan with a vaulted ceiling measuring (2.30 x 1.80 m, height 1.40 m). Unlike the north chamber, there are no benches here and the deceased were deposited on the floor and in a small niche cut into the rear of the chamber, again without sherd-biers. On the floor, the skeletons of two adults where uncovered in flexed positions on their left sides with backs against opposite walls – a male of 18-25 years (1.70 m) and a female of 20-30 years (1.55 m). Next to the latter were the remains of a child of some 9-10 years of age, very tightly flexed, in the opinion of the investigator perhaps originally shrouded (Manunza 2013: 41). Also on the floor was a relatively rich assemblage of offerings discovered in situ. These included two fluted, burnish-decorated bowls (the smaller placed near the feet of the male and the larger between the deceased), two ornately decorated situlae (one placed near the feet of the male, the other by the head of the female), in addition to a handled cup, a snail shell, several unidentified stone implements, and a simple copper awl. The small niche (85 cm long, 45 cm high) was cut into the west wall 55 cm above the floor, and contained disarticulated skeletal remains representing the secondary burial of two children, one 3-4 years of age, the other less than 11 years, accompanied by a miniature cup and fragments of a marine mussel shell. In Manunza’s estimation, these remains represent a nuclear family, overcome by a calamity of some sort and buried together. As there were no osteological signs of violence, an epidemic is suggested (Manunza 2013: 59).

19Taken together, the material accompaniments at Gannì suggest that the individuals buried in the east chamber (Manunza’s nuclear family) were of higher social status than those in the north chamber (Manunza 2013: 59-60). This has been strongly borne out by the results of Lai’s stable isotope analyses of the bones (Lai et al. in press). Generally, both groups evinced periodic nutritional stress, tooth decay, physical stress from repeated work activities, skull porosity and vitamin deficiencies consistent with complex problems of food shortages. At the same time, it was found that the individuals buried in the east chamber (presumably a family) had consumed some 50% more animal protein than those buried in the north chamber – with the exception of the 3-year-old, who had depleted protein levels consistent with a newly-weaned status. Most notably, the male interred in the easternmost niche of the north chamber had apparently subsisted on an entirely vegetarian diet from childhood. The investigators interpret this as indicating a non-local origin (Lai et al. in press, Manunza 2013: 60): a person perhaps having come to occupy a dependent status in the household (or possibly married into it). Although uncertain, one conclusion to be drawn from the combined burial data at Gannì is that family affiliation was a greater determinant of status and its material and dietary prerogatives than were either sex or age.

Taking a non-contextual approach to the evidence

20Traditional approaches to investigating burial assemblages tend to foreground the interrogation of social context and function for clues to explaining observed variations in architecture, contents and implied observances. As noted, I will instead adopt a bricolage-collage heuristic that encourages a fractal reading of the evidence. By turning an alternative, de-contextualizing gaze on these assemblages, I have sought to trace the conceptual and stylistic inputs deployed in their creation back to their source- or reference venues – in the sense of object biographies or microhistories. A three-step methodology has been followed:

211. Disassembling the funerary repertoire to the constituent elements of tomb design, burial treatments (observances) and material accompaniments (with ceramic accompaniments further disassembled to decorative treatments and display).

222. Sourcing constituent elements. For this, a detailed comparative review of published records was carried out, proceeding from local to increasingly distant, extra-insular coverage, with the aim of documenting close contemporary parallels. Several hundred publications were consulted covering the Mediterranean and immediately adjacent regions north and south from Portugal to Asia Minor-Levant and beyond into Mesopotamia. To be clear, the interest was the sourcing of references, locating the venues from which concepts, imagery, stylistic templates, prototypes were likely gleaned, curated and eventually reproduced in novel re-combinations within each tomb assemblage. The geographically nearest instances of substantial bodies of closely comparable material were taken as the most likely reference sources of the funerary expressions in the tombs being examined. Instances of exceptions were also noted but not considered indicative of likely collection venues.

233. Mapping probable source venues for each tomb assemblage and evaluating their relevance for informing agency-based models.

Disassembling and Sourcing

Tomb design (figs. 2-4)

24Shaft tombs were widespread at this time across the Mediterranean and in Mesopotamia, and their appearance in Sardinia (as a generic type) might be read as denoting participation in a wider mortuary koiné of sorts (see Contu 2000 for a review). At the same time, the Monte Claro shaft tombs generally, and Matutzu and Gannì in particular, differ from the typical Mediterranean versions in several ways: 1) the presence of a central chamber off which the burial loci were positioned, 2) the presence of carved benches and raised niches for depositing the dead, and 3) the not infrequent occurrence of carved interior appointments such as thresholds, mouldings, curbs, etc. All have local precedents in hypogea dating from the 4^th millennium if not earlier (Robin 2016), suggesting that the subsequent Monte Claro tomb designs represent at least in part the incorporation of traditional elements. For Matutzu and Gannì in particular, however, a close consideration of the published corpus of tomb plans across the surveyed area suggests design inputs from more distant venues as well, namely those in the Orontes and middle Euphrates valleys of northern Syria (figs. 3-4, see Carter & Parker 1995, Cooper 2007: 94-104, Deb 2013 for reviews).

3. Eastern comparanda

a: Halawa, Syria, Grab H-32; b: Tell Banat, Syria, tomb 2

25Rock-cut shaft tombs seem to have appeared in Syria at about the same time as in Sardinia – in the mid-3^rd millennium or just before (Carter & Parker 1995: table 14.1). Although the eastern examples vary considerably in design details, some are remarkably similar in interior layout and carved features to those of Matutzu and Gannì. These include raised burial benches/niches as well as interior appointments. For the more longitudinal-rectangular plan of the Matutzu tomb, remarkably close contemporary examples (mid- to late 3^rd millennium) can be found at Halawa (fig. 3a) and Tell Banat (fig. 3b). Gannì’s more radial plan has fewer parallels. Locally, the radial burial chamber is not uncommon in more ancient hypogea (fig. 4b), and the Gannì tomb may represent a local adaptation. At the same time, closely comparable radial-chambered shaft tombs are also known in western Syria (fig. 4c). That western Syria provided an additional reference source for the Matutzu and Gannì tomb designs thus seems likely (cf. Orthmann 1989: taf. 32 and 33). It is worth noting that similar tombs appeared also in Cyprus (Lapithos Vrysi tou Barba) by the end of the 3^rd millennium (ECIIIA-B), and are felt to represent similarly close ties with Syria (Keswani 2004).

4. Gannì and comparanda

a: Tomb I at Gannì, Sardinia; b: Tomb 2 at Cungiau su Tuttui, Sardinia; c: Tomb IV at Mishrifeh/Qatna, Syria

Burial treatment

26The custom of positioning the deceased in semi-flexed attitude on the left side as at Gannì and Matutzu is a norm for Monte Claro inhumations and has local precedents dating to the Neolithic (Manunza 2013: 58, Luglié 2017: 53). Semi-flexed and flexed inhumations are moreover common throughout the wider Mediterranean basin during the 3^rd millennium. In consequence, any intended reference to specific traditional or exotic venues was not detected.

27Far more diagnostic in this regard is the custom of placing human remains on fragments of large jars in the manner of biers. Apart from Matutzu, examples of this particular practice have been recorded in southern Sardinia in a relatively wealthy child interment in a shaft tomb at Via Basilica (IV) at Cagliari, in several cist tombs at Sa Duchessa-Cagliari and San Gemiliano-Sestu, and probably within the hypogeal tomb of Padru Jossu-Sanluri (see Manunza 2013: 45-55, Webster & Webster 2017: 96-105 for reviews). In the western Mediterranean, the practice appears to be exclusive to the southern facies of Monte Claro, lacking precedents or parallels either locally or in adjacent regions. Indeed, outside of southern Sardinia, sherd-burials would seem to be restricted to Anatolia and northern Syria (Wheeler 1974, Yıldırım 2006, Massa & Şahoğlu 2011: 164-73, Topbaş et al. 1998, Welton 2010, Lamb 1937, Alpaslan-Roodenberg 2002: 93-94, Cooper 2007: 57-61, Valentini 2011, Deb 2013). The specific mode of interment varied, however. In Anatolia, pithoi – whole or fragmented – tended to serve as independent tombs, sometimes within vast cemeteries. Further east in northern Syria and western Iraq, however, one finds close parallels to the Monte Claro custom of laying the deceased on sherd biers within built tombs. In both areas, in contrast to the Monte Claro pattern, sherd burials tended to be reserved for infants or juveniles (McGeorge 2011, McMahon et al. 2009: 121, Lamb 1937: 55, Keswani 2004: 348, n. 32). Such treatments are particularly well documented in the Syrian middle Euphrates area. At Shiyukh Tahtani south of Carchemish, at least half a dozen child interments were placed on large fragments of narrow-necked jars, with accompaniments of pots and pins (Sconzo 2014: 13). Notably, these include a relatively rich burial on four large sherds within its own, “miniature” shaft grave (T.83), dated to 2650-2400 BC (Sconzo 2006, 2007: 274 and pers. comm.). The custom of sherd-burial has otherwise typically been thought to reflect cost-saving efforts in the burial treatment of the young and/or of low-status individuals (Wheeler 1974, McGeorge 2011: n. 48). However, the Syrian records (as well as the Sardinian data) may challenge this reading. At Hama on the Orontes, for example, multiple cases of adults interred in large, split jars with accompaniments are known (Deb 2013: 69, 111-12).

28Rarer are eastern instances of adults buried on sherd-biers assembled from multiple vases, as at Matutzu. At Tell Al-Ansari, ‘pottery fragments of more than one jar were put under and above a skeleton of an adult male’, buried with accompaniments suggesting a relatively high status (Deb 2013: 147). At Chagar Bazar in the Upper Khabur Valley, a partially brick-lined pit in Tomb 9 held an adult skeleton flexed on its left side, without grave goods but placed on a ‘paving or platform for the dead’ composed of sherds from several ceramic vessels (McMahon et al. 2009: 119). It would thus seem likely that along with tomb architecture, eastern regions provided reference sources also for the Monte Claro custom of sherd burial.

Tomb accompaniments: metal and vase shapes (selected)

Tanged leaf-blade dagger

29The tanged leaf-blade dagger found in close association with the rear-niche burial at Matutzu is one of a number of similar blades from a dozen or so Monte Claro contexts, mainly but not exclusively southern burials (Manunza 2010: 76-77, Usai 2005: 263-64, Lo Schiavo et al. 2005: 189-92, Manunza 2010: 76, cat. no. 109), although apparently absent from Gannì. Like these, the Matutzu dagger (10.2 x 2.3 x 0.3 cm) has more or less rounded shoulders, tanged haft with quadrangular section, and overall foliate shape with rhombic or quasi-rhombic section. The style is distinctive, and a review of published forms revealed no close affinities among contemporary assemblages in the wider Mediterranean (e.g. Vaquer & Remicourt 2012, Carboni & Anzidei 2006, Branigan 1967, De Marinis 2005), the Balkans (Courcier 2014) or the Near East (Stronach 1957, Bianchi & Franke 2011, Rosińska-Balik et al. 2015). It is, however, similar to earlier local blades (Filigosa-Abealzu) which in turn have clear Anatolian affinities (fig. 5; Webster & Webster 2017: figs. 35, 62, 77; see also Arcan 2012, Efe & Fidan 2006, Webb et al. 2006). In consequence, there appears to be little reason to attribute the Monte Claro dagger style to other than local references.

5. Tanged copper dagger blades

a: from Matutzu ; b: from Yazilikaya, Turkey; c: from Serra Cannigas, Sardinia; d: from Cyprus, Philia culture

Amphora/tankard

30The Matutzu amphora/tankard is a distinctive vessel, perhaps even unique (fig. 6b). Probably intended for drinking, it stands ca. 21 cm tall with a slender, tapered profile, opposing vertical strap handles and irregular, pattern-burnished vertical strips around the entire body (Manunza 2010: cat. no. 108). It has neither any apparent precursors locally nor any close parallels in the west-central Mediterranean. Its nearest parallels would appear to come from the Troad in western Anatolia. It is very like several amphorae from contemporary levels at Troy (II and III, ca. 2500-2200 cal. BC, see Easton 1989, Weninger & Easton 2014:73-375 and 72-879, Schliemann 1881: 72 – Ilios nos. 1089 and 1092), in particular Ilios no. 1092, with vertical, pressed line-patterns resembling the Matutzu decorations, and a slightly asymmetrical rendering of the handles (fig. 6c). Although there was nothing comparable at Gannì, other Monte Claro tombs have revealed vases alluding to similarly specific reference sources. One is a Trojan-like amphora with fish-spine burnished decoration (fig. 6a); another is a locally unique burnish-decorated tankard/jug which would seem to be a remarkable local emulation of an equally distinctive vase from Kusura in western Turkey (fig. 6e) – both came from the shaft tomb of Via Basilicata IV at Cagliari, which also preserved the only known local example of a child/adolescent sherd burial (Webster & Webster 2017: fig. 60 d, g).

6. Amphorae/tankards

a: from Via Basilicata IV, Sardinia; b: from Matutzu, Sardinia; c: from Hissarlik-Troy, Turkey; d: from Via Basilicata IV, Sardinia; e: from Kusura (B), Turkey

Flange-rimmed pithoi/urns/situlae

31As described above, the sherd biers at Matutzu were composed from three nearly identical vases with tall, rigid profiles, flat bases, vertical strap handles (three) and slightly splayed flanged rims (orlo a tesa) (Manunza 2010: cat. nos. 95-97). Two stand just over a meter tall (104 and 105 cm) and the third, which is wider, measures 99.7 cm. All are nearly identically decorated with orthogonal flutings along with one or two small triangular lugs (bugne a lingua triangolare) placed below the rim and near the base. The two situlae at Gannì – one with four handles – are highly decorated with orthogonal fluting, punctations and enlarged decorative spurs and have pronouncedly splayed, flanged rims.

32While the large flange-rimmed vase is a common feature in southern Monte Claro burial- and domestic assemblages (see Depalmas 1989: Tav. I, Manunza 2013: 56-57), there is nothing even remotely comparable among local repertoires, either Filigosa-Abealzu or Bell Beaker (Webster & Webster 2017: 79). Outside the island, rare finds of flanged rims are recorded in Corsica (Bailloud 1968: figs. 5, 8) and the Balearics (Benejam & Plantalamor Massanet 2000: 539, Depalmas & Plantalamor Massanet 2003: 59, fig. 3.9). But it is only further east, in Anatolia, the Ionian islands e.g. Emporio-Chios, ca. 2650-2450 BC (Hood et al. 1982: figs. 196: 1404-05, 204: 1620, 1643-44); Poliochni-Lemnos, Green phase (Bernabò Brea 1957); Aphrodisias, ca. 2150 BC (Joukowsky 1986, I: fig. 425: 39), and in the Upper Euphrates valley (Tell Banat, 2600-2450 BC; Porter 2007) that these rims are at all common, although occurring on other vase types. Surprisingly, the closest parallels to the Monte Claro splayed, flanged rims would seem be among the handle-less jars and urns of the upper Kabur valley on the Euphrates (e.g. Tell Brak, 2400-2300 BC, Oates et al. 2001: fig. 460: 1565; Tell Hamoukar, Colantoni & Ur 2011: fig. 10; Tell Beydar, Gavagnin & Mas 2014: figs. 17-18). Although geographically diffuse, the likely venues referenced in the distinctive Monte Claro flange-rimmed vases would thus appear to lie in the Near East.

7. Flange-rimmed vases

a: pithos B from Matutzu, Sardinia; b: pithos A from Matutzu; c: jar from Tell Brak, Syria; d: urns from Tell Hamoukar, Syria; e: situla from Gannì, Sardinia; f: situla from Gannì

Carinated, biconical necked jar

33This vase type from Matutzu (fig. 8a) is not typical in Monte Claro assemblages. It may represent a re-interpretation of a local Filigosa-Abealzu form found f. ex. at Lochele-Sedilo (fig. 8b), with added orthogonal fluting and reticular burnished designs.

8. Jars

a: from Matutzu, Sardinia; b: Filigosa-Abealzu facies jar from Lochele-Sedilo, Sardinia; c: from Gannì; d: Filigosa-Abealzu facies jar from Mind ‘e Gureu-Gesturi, Sardinia

Small necked jar

34Like the vase above from Matutzu, a similarly atypical two-handled, globular necked jar came from Gannì (fig. 8c) although was not found in situ. It is entirely undecorated and similar to a Filigosa-Abealzu jar/ewer (fig. 8d) from the contemporary (2620-2287 cal. BC) hypogeal burial of Mind ‘e Gureu-Gesturi in the uplands to the north (Usai 1998, fig. 23.10: 6). Although similar forms can be found outside the island (e.g. at Yortan, Turkey; Kiamil 1980: fig. 70), it seems just as likely that its source is local, i.e. that it represents a Monte Claro interpretation of a Filigosa-Abealzu type.

Trunco-conical bowls with inverted rims

35Flat-bottomed trunco-conical bowls with inverted rims like those at Matutzu (fig. 9b-d) and Gannì (fig. 9a) would appear to be exclusive to the southern Monte Claro facies, where they occur in several funerary contexts, and often present pattern-burnished decoration (Manunza 2013, Depalmas 1989: Tav. 12-13).¹ Outside of Sardinia, the type is only found in any number in western Anatolia (fig. 9e), western Syria (Curvers 1991: Pl. 21) and, especially, the southern Levant, where they are also frequently pattern-burnished (fig. 9f).

9. Trunco-conical bowls with inverted rims

a: from Gannì, Tomb I; b-d: from Matutzu, Sardinia, with pattern-burnished bases; e: from Hissarlik-Troy, Turkey; f: from Beth Jarah, Israel; g: from Tel Dan, Israel

Platform-rimmed tripod braziers

36Two low, trunco-conical vessels from Matutzu (fig. 10a-b) have flanged platform-rims with concentric flutes/excisions, two vertical strap handles, and three so-called tongue feet. Irregular punctations occur around the bowls and on the feet, between irregular, oblique incise/excised lines. Both vases are fashioned from black to light brown paste and have red (slipped?) and burnished surfaces (Manunza 2010: 205-206). Vases such as these are a distinctive element in many southern Monte Claro burial assemblages (Manunza 2010: 66-69, Depalmas 1989, Tav. 6: 1-6, E. Atzeni 2001, Fig. 2B, Mascia 2012, fig. 2: 3-4), although not at Gannì, and if the tongue foot is indeed diagnostic, also within settlements. Parallels outside the island are rare. Elsewhere, two miniatures with similar rims can be noted on undecorated, hemispherical and legless bowls from Tell Shiyuhk Tahtani in the Syrian middle Euphrates dating to 2800/2700-2400 BC (Sconzo 2007: fig. 17.11.10-11) and a third (also legless) from Stratum IV (ca. 2200-2000 BC) at Tepe Gawra (Speiser 1935: Pl. LXXIII, 179). In terms of likely reference sources, however, Tanda & Paglietti (2011) have suggested the local ceramic floor hearths. Two examples – one fragmentary and with uncertain provenience from Monti-Lunamatrona (Tanda & Paglietti 2011: 354); a second uncovered in situ, set into the floor of hut 13 at San Gemiliano-Sestu and filled with charcoal and ashes (fig. 10c), both have distinctive, flat bases and concentrically fluted platform rims. The form may have a deeper local heritage still, in the carved and concentrically incised floor hearths found in more ancient hypogea (Tanda & Paglietti 2011: Tab. I, fig. 1).

10. Braziers and hearth

a-b: platform-rimmed tripod braziers from Matutzu, Sardinia; c. ceramic floor hearth from S. Gemiliano-Sestu, Sardinia

“Corrugated” bowls

37From Gannì came at least two bowls (or open dishes) with horizontal fluting (fig. 11a and 11c). Although absent from Matutzu, these are a common southern Monte Claro burial accompaniment. They are quite rare elsewhere in much of the Mediterranean, although not uncommon in southern Turkey and Syria-Iraq where they are known as “corrugated ware” (Delougaz 1952: Pl. 63, Luke et al. 2005: fig. 3, Orthmann 1963). Especially stringent parallels to the Monte Claro bowls are found at Kulluoba in western Turkey (Turkteki 2012: resim 14.2 a-b, levha 3: 3), and Tarsus on the southern Turkish coast (fig. 11d). One specimen from Gannì, with heavily beaded rim (fig. 11a), finds a near parallel at Tell Banat in Syria (fig. 11b).

11. Corrugated bowls

a, c: from Gannì, Sardinia; b: from Tell Banat, Syria; d: from Tarsus, Turkey

Tomb accompaniments: decorative treatments and display

Decorative treatments

Fluting

38Use was made of various forms of decorative channelling, corrugation or incision/excision on pottery throughout the Mediterranean and beyond during the period in question. As frequently referred to by local scholars (e.g. Castaldi 1999: 151-53, Lecca 2012: 37, Moravetti 2009: 102), some use was made of flute-decorations in Corsica and southern France (Goedert 1988: figs. 4-5, Arnal et al. 1967, Gutherz & Jallot 1999: figs. 5-6, Giligny et al. 1997: figs. 14-15), on the Italian mainland (Melis & Talamo 2012: figs. 1: 2,3,6), on Lipari (Melis & Talamo 2012: fig. 1: 2,3,6) and in Iberia (Moravetti 2009: 102). Moving beyond this western core area, fluting becomes relatively more common in western Anatolia and northwestern Syria (eg. Blegen et al. 1950: fig. 412, Lloyd & Mellaart 1962, Mellaart 1954, Lamb 1937: figs. 7 and 1938: fig. 10, Bernabò Brea 1957: Tav. CLXV: i, l, m and Tav. XXXVII:h, i, j, l, Braidwood & Braidwood 1960, figs. 281, 283, Pls. 33-34, Batiuk 2005). Rare occurrences of fluted sherds in Aegean and eastern Mediterranean contexts have generally been interpreted as Anatolian imports or copies (Immerwahr 1971: nos. 70-75, Phelps 1975: 356, see also the Philia-Cyprus Bryn Mawr Collection, D. Bolger, pers. comm.). But the deployment of extensive fields of deep, parallel linear flutings, typically producing orthogonal designs as seen on the Matutzu and Gannì vases and common in the southern Monte Claro repertoire, is far less common. For these vases, local precedents were probably drawn upon: earlier dates on flute-decorated vases have come from a least two other contexts in the island (Melis 2013: tab 2.).

Pattern-burnishing (see fig. 9)

39On a few large Monte Claro funerary vases, decorative fields otherwise usually reserved for vertical fluting were covered instead with wide, burnished strips of parallel verticals or obliques, as in the two Gannì situlae. Pattern-burnishing was otherwise reserved for smaller vessel forms such as bowls. Outside the island, apart from rare examples on so-called Piano Conte pans from Lipari (Bernabò Brea 1958), the technique is little known west of the Aegean, while pattern- and streak-burnishing have a wide distribution in the eastern Mediterranean (Hood et al. 1982: 1758), Anatolia (Lamb 1937: 24), Cyprus (Bolger 2007: 173; Peltenburg 2007: 150-51) and the Levant (Greenberg 2000, fig. 11.5; Yamafuji 2014).

Decorative spurs/lugs (see figs. 6d-e, 7a,b,e)

40The occurrence of decorative spurs on many Monte Claro vases including those from Matutzu and Gannì – in the latter much enlarged and flattened – suggests western Anatolia as a likely reference venue, and/or the eastern Aegean, where they are commonly classified as lugs or warts (as at Emporio, Hood et al. 1982: fig. 207, see also Lamb 1937: pl. VI.15, Morris 2009: fig. 32, Mellaart 1954: no. 554, Kiamil 1980: figs. 44: 26 and 47: 22, Warner 1994: pl. 178: h/KA 77, Lloyd & Mellaart 1962: fig. P.25.2/7, Orthmann 1963: taf. 24-5/50, Sari 2012: sekil  8:C:11, Garstang 1953: fig. 124: 7, Schliemann 1881: 72).

Decorative display

41A striking feature of the southern Monte Claro vases is decorative coverage favouring one exterior face only, with the opposite face left plain or only minimally embellished, giving the impression of a front and a back. As far as is known, this approach appears to be unique to the southern facies of Monte Claro. While it is very clearly represented in tomb assemblages like those of Matutzu and Gannì, fragmentary evidence has come from settlements too, although it is possible that those vases were also intended for funerary observances. Foregrounding the performative dimension of funerary rites, we have suggested that vases decorated on one face only might be viewed as stage props in rites enacted within the tomb chamber or elsewhere. If burial jars or accompaniments were intended for public display prior to interment, an orchestrated presentation that obscured the undecorated face would have provided economizing hosts the means to benefit from displaying what were surely status insignia while minimizing their decoration investment (Webster & Webster 2017: 104). In all events, the custom of decorating vases on one side only – whether intended for offerings or biers – is entirely unprecedented in the island, and seems to have no known extra-insular parallels.

Reconstructing and interpreting source catchments

42Mapping the comparanda for each of the tomb elements set out above has produced the contours of the hypothetical catchment area (fig. 12) from which diverse elements representing skills, techniques, and form- and style attributes were collected and re-deployed in novel mortuary assemblages at Matutzu and Gannì. Several observations follow:

431. For the two assemblages examined here, the catchments comprised both local-traditional and distant-exotic venues. The latter, although geographically extensive, are more or less contiguous, spanning the Levant through the Syrian middle Euphrates across western Anatolia (and adjacent Ionian islands). The individual Sardinian burial assemblages reflect an eclecticism comprising stylistic attributes drawn from diffuse venues among which no single “homeland” has been identified.

442. There is but limited evidence that the western and central Mediterranean were important reference venues in creating the tomb assemblages examined here. The often cited parallels of some diagnostic Monte Claro attributes, namely fluting, pattern-burnishing and, more rarely, flanged rims, in areas adjacent to Sardinia (southern France, Corsica, Balearics, Italy, Lipari) are neither numerous nor, with rare exceptions (Lipari), stringently comparable. It should be pointed out here that although an eastern catchment may have been the most common source of exotic references in the Sardinian shaft tomb assemblages, there is no basis to assume this was the case among all Monte Claro burial contexts.

453. Lipari may represent a special reference point for eastern networks. Finds of fluted (although horizontally, and frequently inside the vase) and pattern-burnished sherds – some with very close stylistic ties to western Anatolia (Lloyd & Mellaart 1962: fig. P.36: 7; cf. Castaldi 1999: 153) – may add to the evidence, suggesting that this small island off northern Sicily served as a contact point or way-station from an early date, along the most likely transmarine route between Sardinia and eastern venues (fig. 2; Moravetti 2017: 202, see also Cosmopoulos 1991: 166, Avino & Rosada 2014, Bernabò Brea 1988, Cazzella et al. 2011, Kouka 2009).

464. Local venues, i.e. surviving Filigosa-Abealzu material culture, contributed minor but detectable inputs, specifically in elements of tomb architecture and the design of several accompaniments (tanged dagger, medium-sized jars). The Sardinian Bell Beaker culture, on the other hand, appears to have had no observable influence on the creation of either Matutzu the Matutzu or Gannì tomb assemblages (contra Manunza 20 10: 109-10).

475. Although architecture, accompaniments and implied burial modes vary between the two Monte Claro tomb assemblages examined here, both comprise elements drawn from source venues within the same hypothetical eastern catchment area.

486. Despite clear examples of local emulations of foreign objects (fig. 6a, c, e), no confirmed examples of foreign imports have been identified at Matutzu or Gannì, nor in other southern Monte Claro deposits. This does not exclude, of course, the likely probability of perishable material or human labour imports (Webster & Webster 2017: 111).

De-contextualized and re-contextualized interpretations

49A de-contextualized reading of the Matutzu and Gannì tomb assemblages structured within a collage heuristic suggests that they constitute the creative works of local agents engaged in intellectual bricolage. Significantly, the resulting compositions of the funerary creations revealed by this approach to these complex materialities were not unique to Matutzu or Gannì. Fixing a fractal gaze also on the wider corpus of less well-preserved or recorded assemblages strongly suggests that the innovative collage with multiple-source referencing was indeed typical of Monte Claro funerary expression, at least of the socially more elevated shaft tombs (see Webster & Webster 2017: 97-102, 104). The findings outlined here invite novel re-interpretations of the assemblages within their social context. One might now with some confidence state for example that mortuary repertoires played a key role as set pieces for the performance of burial rites, as facilitators of the observances involved, and as objects with significant references to the status, relationships and activities that had defined the deceased in life.

50This new reading of the funerary evidence also accommodates more fully certain features of the Monte Claro record that have heretofore proven enigmatic for conventional “foreign incursion-invasion-migration” models (Moravetti 2002, 2009, 2017: 202, Manunza 2013: 58, Foschi Nieddu 2000: 809, see also Melis 2000: 142), while at the same time lending some support to alternative formulations featuring Monte Claro agents engaged in foreign trade ventures (Webster & Webster 2017: 110-111, cf. Melis 2014: 15). At the same time, as Hahn reminds us (Hahn 2004), things and concepts transferred inter-culturally (for example via a bricoleur’s de-contextualizations, curations, re-combinations and re-deployments) are unlikely to retain the meanings, significance or associations accorded them in their previous venues. Beyond what must have been legible allusions to the Other, the foreign or the exotic encoded in the motifs, practices and styles brought in from afar, we do not know what such “implants”, re-embedded within collage-like tomb assemblages such as Matutzu and Gannì, might have meant to local recipients (although we can be certain that it will have varied among audiences).

51Assuming that our hypothetical bricoleur(s) rendered selectively, from a wide inventory of styles and attributes gleaned from near and far, only those with potentially meaningful and readable associations, one might speculate on their social reception under the alternate gazes within a southern Monte Claro cultural setting. At the insular level, references to indigenous Filigosa-Abealzu material culture (architectural, ceramic, metal) in these assemblages, although never predominant, suggest that some degree of identification with local tradition was intended. At the same time, other references appear to strongly support notions of alterity and elite affiliation. The variety of overtly exotic and highly embellished vases including massive storage jars – all presumably beyond the means of the non-elite – clearly announces social distinction. Moreover, such accompaniments being typical of elite funerary assemblages in the south may well have served a common syntax – a lingua franca – among those who, by acquiring such works, authenticated their affiliations and underlined their alterity while at the same time linking them to presumably prestigious commercial networks as well as to their enabling influence in local economies of surplus and storage (see Webster & Webster 2017: 96-111).

52Regarding the elite insider, yet another reading is prompted. Subtle variations in architecture, in the ritualised treatment of burial remains, and in the specific inclusions and quality of material accompaniments doubtless came under the critical eye of status-sensitive peers. At Matutzu, the tanged dagger and the extensive use of sherd biers composed of conspicuously decorated pithoi may have earned its custodians an advantage in the status arena, and the Trojan-like amphora perhaps announced privileged access to trade. As elsewhere in southern Sardinia, it was in such coded gestures on display that social competitors, in the context of funerary participation, might decipher clues to their rivals’ successes or failures, to the breadth of their mercantile wanderings, to special knowledge acquired, to new routes discovered or new contacts made. It would have been on such material evidence, offered up as a visual documentary to those able to read it, that judgments were made and opinions moulded in matters of social merit and status. One might indeed liken such funerary projects to the literary documentary-collage, where meaning is extracted via historical citation, fragmentation and ‘radical recontextualization’ (Jaeger 1994: 14, see also Jones 1993).

53We are still at pains to know the social dimensions of such hypothetical bricoleurs, or how best to envision the means by which the repertoire of resources was gleaned – whether directly or via intermediaries. But subsequent work on the Monte Claro corpus and its comparanda should provide some detail to this still sketchy picture. At the moment, it seems reasonable to propose that intellectual resources were acquired incidentally under the guise of local tradition and the banner of foreign trade toward deployments in distinctive creations that set their patrons and their families apart in terms of status and its material insignia.

Conclusion

54The bricolage model as originally conceived by Claude Lévi-Strauss has encouraged a novel investigative angle for addressing complex materialities like those often characterizing funerary expression. By applying the allied concept of collage, and engaging a de-contextualizing fractal reading of material remains, such assemblages can be newly conceived as collections of constituents with specific cultural references and in so doing enable their investigative unpacking toward a potential exposure of the agencies involved in their creation. In the present illustrative application, complex Monte Claro funerary materialities of the 3^rd millennium BC in Sardinia seem to reveal the hand of intellectual bricoleurs working from an eclectic inventory of locally as well as exotically gleaned skills, customs and styles for selective re-deployments in the creation of novel mortuary facilities. It is hoped that the interpretive statements offered above show that the bricolage-collage heuristic can provide a valuable complement to more conventional contextual-functional based investigations.

55Acknowledgements. I thank Peter Van Dommelen, John Cherry, Bernard Knapp, Maxence Bailly and several anonymous readers for their valuable critiques of earlier versions of this paper.

Atzeni 1959-1961, ATZENI E., ”I villaggi preistorici di San Gemiliano di Sestu e di Monte Olladiri di Monastir presso Cagliari e le ceramiche della ‘facies’ di Monte Claro”, Studi Sardi 17 (1959-1961 [1962]), p. 3-216.