CORE Metadata, citation and similar papers at core.ac.uk

Provided by OAR@UM

Cremation burials in early Bronze Age : evidence from Tarxien • and Ggantija

George Azzopardi

A common way of disposing of the dead where a number of similar-looking natural across the entire Mediterranean (and even knolls are to be found. Leading on from this beyond) from prehistoric down to late Roman point, he further argues that the physical times was to bury the corpse - or the burnt constitution of the Maeshowe mound mirrors remains, if cremated - in a built chamber or that of the earth itself. As it is constructed of in a floor cavity and then cover it up under stone but is then covered by a mound of nc.tural a mound of stones and/or earth, usually clay, the chamber resembles the geological encircled by a kerb or retaining stone wall. formation of the earth whose bed-rock is Sizes varied and larger ones might even covered by natural till. According to Richstrds, have a passage providing access to the burial this structure to house the dead might have chamber. In many cases, these burial mounds been intended to be visible as a monument, or tumuli used to be circular, having an overall yet, at the same time, it positioned the dead appearance of a cone, hut, or hill. below the surface of the world inhabited by It is worth investigating whether the humans. 3 Richards' argument in respe.::t of resemblance of burial mounds to huts was Maeshowe's burial mound may, per:J.aps, intentional or merely accidental (something also be applied to other similar burials. Quite which I do not intend to do in this contribution). plausibly too, Chris Scarre suggests that If it was intentional, was it meant to convey the low rounded form of the artificial burial an idea of the tomb as a house of the dead? mounds may replicate the shape of the m:.tural For instance, Richard Bradley supposes that hills in which previous burials took place. thus cairns were regarded as houses of the dead. 1 manifesting a shift from a tradition of burial Cited by Robert Layton and Peter J. Ucko, 2 in natural hills to a tradition of artificial burial Colin Richards argues that as the living mounds. 4 This theory may also complement resided in houses where the central hearth Leif Sahlqvist's suggestion that a barrow can contained the life-maintaining fire, the dead be viewed as a sacred mountain in miniature resided in tombs whose perpetual darkness and/or a metaphor for a sacred mountain. 5 signified their role as residences of the dead. Nonetheless, other forms of burial different Richards continues to elaborate on the late from the above were also encountered. 6 Neolithic passage grave of Maeshowe in In this contribution, I shall be dealing west Mainland, Orkney, in northern Britain with cremation burials on the Maltese islands by describing it as a place being conceptually in the early Bronze Age (Tarxien Cemetery below but physically above ground, thus phase: 2400-1500 B.C.), focusing, towards placing the dead in an ambiguous position the end, on a type of clay figurine associated located between two worlds. He sustains his - as yet, exclusively - with this early Bronze argument first by highlighting the resemblance Age practice. Some new interpretations of the mound to the surrounding topography are attempted in respect of both the br..rials

Malta Archaeological Review • Issue 8 2006<2007 9 themselves as well as the mentioned bones and gathered in double-layered fat inside figurines. To this end, I shall be resorting a golden urn covered with a linen shroud. to evidence from two temple sites: those at Then, the circle of the barrow was marked off, Tarxien and Ggantija, with the richer corpus the stone revetments set around the pyre, and a of e\idence coming from the former. 7 To begin mound of earth was immediately heaped up. 10 with, I derive insights for my interpretations Finally, towards the end of the Iliad, we by initially resorting to the earliest literary are given another account: that of Hector's evidence we have and which concerns the burial following his death in retaliation for Aegean region where it comes from. But I also the death of Patroklos. Book XXIV gives us a draw on comparisons with some other sites similar, albeit shorter, description of Hector's and/or materials outside the Maltese islands. It funeral: great heaps of wood were gathered, is mainly this comparative approach that, for Hector's corpse was laid on top of the high the greater part, provides the backbone of my pyre and set on fire. On the following day, the argu:::nents. fire was quenched with wine and his bones were gathered, wrapped in soft purple cloths, and placed in a golden urn. The urn with his Literary evidence bones was then laid in a hollow grave, covered over with large stones closely set together, and The earliest literary description we seem to immediately after by a barrow. 11 have of cremation followed by burial under mounds or barrows is provided by Homer. Though he probably lived and wrote in the The cremation cemetery at Tarxien 91h or 8th century BC, Homer displays good knov;ledge of the funerary practices of Bronze Like much of the rest of the Mediterranean, Age Greece with which he vividly colours his the Maltese islands may not have been alien narrative. to this type of funerary scenario, nor are The narratives which give us the relative they expected to have been. Furthermore, the descriptions are to be found in his Iliad. The possible existence of this kind of burial in first concerns the death and burial of Patroklos. Malta and has never, to my knowledge, We are first given a brief account of the funeral been seriously explored or possibly even preparations in Book XVIII: after having been thought of. washed and treated with olive oil, Patroklos' The first season of Temi Zammit's corpse was laid on a bier and covered with a excavations at Tarxien in 1915 was to thin Jght cloth from head to feet, over which provide us with the first and, as yet, the most a white linen shroud was finally laid.8 Book substantial evidence of Bronze Age cremation XXIII gives us a fuller and more detailed burials. 12 Just about 1.20 m below the field description of what went on in the funeral: surface, Temi Zammit came across a deposit c. wood was heaped up and a pyre was built, on 30 cm thick consisting of dark soil and ashes. top of which the corpse was laid. Fat sheep It contained cremated bones, burnt fabrics, and oxen were skinned, dressed, and made pottery, and figurine fragments amongst other ready for the pyre. The corpse was covered material. A large number of in situ crushed jars from head to feet with fat extracted from the were also found. On the basis of their relation dead animals, while their flayed carcasses to the cremated bones and other material, were heaped around the corpse. Twelve slain they appeared to have been urns holding the Troja::1s and more animals were cast upon cremated remains and accompanying grave the pyre, along with jars of honey, oil, and goods. In fact, in addition to bones and unguents which were leaned against the bier, pottery, they contained carbonised seeds and and a:..1 was set on fire. 9 plant remains, bronze implements/weapons, Th next day, upon instructions of Achilles, personal ornaments, and lumps of both fine the fire was quenched with gleaming wine, the and coarse tissues from burnt dyed fabrics bones of Patroklos were singled out (as he lay with which the cremated bodies were evidently in the middle of the pyre) from the rest of the dressed or wrapped.

10 Malta Archaeological Review • Issue 8 2006/2007 This 'cemetery' /cremation deposit, which investigative works at this site. The clearance covered a relatively small part of the area of an amount of stones and earth in front of the occupied by the whole Neolithic temple north temple entrance exposed an undisturbed complex (in fact, it was only found within the deposit of 'dark grey earth and ashes'. This Tarxien South temple), rested on a fine sandy deposit contained many sherds described as soil deposit devoid of stones. This last deposit 'neolithic', flint and obsidian flakes, stone itself lay over the temple floor and over any objects, bone points and, very interestingly, stone blocks and other debris from the temple small fragments of carbonised matter. The ruins. According to Evans, it may have been an deposit extended along the entire length of artificial fill intended to produce a level floor both temple facades, apparently reaching the for the 'cemetery'. Moreover, the area covered western wall too. 14 by the 'cemetery' /cremation deposit seems to More recently, in 1999, a discoid figurine have been roughly circular, with a diameter of fragment was discovered by Mr John B::jada about 12.2 m. while on duty as a Museum Officer at Ggantija From the outset, it appears that the stone (Fig. 2). This important fragment was found enclosures of the already ruined Tarxien embedded yet exposed in the same grey ashy temples quite conveniently accommodated layer in front of the North Temple facade the cremation burials of the early Bronze Age and to the right. On close inspection and people. It is quite difficult to ascertain whether documentation of the fragment's find-spot the cremation of the corpses was undertaken and associated context, the layer was found to on the site ofthe 'cemetery' itself. The amount contain not only pottery sherds but also p:eces of ash in the 'cemetery' /cremation deposit and of bone (Fig. 1)_IS the evidence of burning on the stones in and To a lesser extent and not quite clearly, around the area convinced Zammit that it was. this situation at Ggantija somewhat mi:::rors But the burning does not seem to have been that at Tarxien, the main difference being restricted to the 'cemetery' area. It was evident that, at Ggantija, what appears to be a also in other areas which must have had cremation deposit is situated outside and not nothing to do with the 'cemetery'. Therefore, within the Neolithic temple compound :::s at it is likewise possible that, while the cremation Tarxien. However, the picture at Gganti~a is took place elsewhere, the cremated bones enhanced and enlightened by a further piece 1 along with the other grave goods were placed of evidence provided by the l8 h cer.lury in urns and buried in the 'cemetery' area. French traveller Jean Houel. In his published It is also unclear whether any mounds (or, plan of the Ggantija temples, Houel depicts perhaps, a communal mound?) were raised a circular yet smaller structure next to the above the burials. The roughly circular form north temple facade and to the right (Plate of the 'cemetery' area may be indicative, but 1). 16 Coupled with the presence of the grey any remaining traces might have been cleared ashy deposit - even if this is reported to away as a result of agricultural activity, to the have extended further to the South Temple extent that Zammit found no such traces at facade and west wall too - and its contents, all. On the other hand, the stone enclosures of this evidence leads me to seriously consider the earlier temple structure might have been the one-time presence there of a cremz.tion found to accommodate the burials so well that burial under a mound. In this respect, what no mound or mounds were needed. This may Jean Houel saw and documented pictor::ally explain, in this case, the choice of the site for when visiting Ggantija might have been the the 'cemetery' _13 remains of the mound, which is pretty \\'ell outlined (even if schematically) in the plan he produced. A slight heap is still visible on this The ashy deposit at Ggantija spot to this day and to my knowledge it has never been investigated. On the other hand, Less obvious, but quite indicative, is an the grey ashy deposit with its contents might interesting piece of evidence from Ggantija. be spilt material from the mound itself as a In 1936, the Museums Department resumed result of its decay.

Malta Archaeological Review • Issue 8 200612007 11 /" 0 10

Metres

LEGEND

D Documented area

x Find-spot of discoid figurine fragment

LEGEND

//Ill/ Megaliths I big stones

Limit of documentarion

"~ Bedrock

'Torba' naor with small stone chippings

Grey deposit containing pieces of bone and pottery sherds "0'" Q ~" ''!) 'I':. X Find-spot of discoid figurine ~ fragment embedded in 3 0. ' E E .c t 0 t 0 ~'" E 1' A I 2 w~ ' (,) ~l?_. -~-----·-· -·-·~. -·-·-·-·- ·-·-·-· -·-.- -·-·- ·-· -·-.- ·-·-·

PLAN

1 L------~--~---~ Metres

Fig. 1. Plan ofGgantija temple complex with documented area on the right (top) and detailed plan of the documented area showi'lg the discoid figurine fragment's find-spot (marked 'x') and associated context (bottom). (Drawing: John Bajada and George Azzopardi)

12 Malta Archaeological Review • Issue 8 2006/2007 Plate 1. Plate CCLI (bottom) in Jean Houel's ·Voyage Pittoresque des isles de Sicile, de Lipari et de Malte', IV, (Paris. ! 787 ). The circular structure (a) on the left is rhe Xaghra Stone Circle while the complex structure (b and g) on the right is Ggantija. Another circular, yet smaller, structure (h)­ possibly, the remains of an early Bronze Age cremation burial mound - on the extreme right stands next 10 the facade of the smaller northern temple.

Discussion Tarxien?). This indicates a tendency whereby newer monuments cluster around much older But one may ask: why build mounds or barrows ones. 20 next to or within an earlier temple structure Referring to developments of earlier (as might have been the case at Tarxien)? Neolithic enclosures in Britain, Mark Edm::mds Monuments could be adapted or altered to suit plausibly suggests that building barrows next new demands or changing circumstances. Thus, to or on the perimeter of an enclosure (or, in while the physical monuments themselves tend my view, any other monument, for the sake of to endure, the ideas and symbolism associated this argument) meant inserting 'new people with them may change as the monuments get into the stories and associations of the phce'. appropriated by successive generations who From a phenomenological perspective, he might see them from different perspectives. 17 adds that the burial of individuals under such Monuments (or monumental enclosures) mounds (next to or on the perimeter of an originally used for public ceremonies could, enclosure) asserted a claim to the enclosure/ therefore, be appropriated by new people for monument and all that it stood for by the the burial of their dead. 18 During the early people responsible for the buriai.2 1 Bronze Age (end of 3rd_beginning of 2"d In a similar perspective, Richard Bn:dley millennium BC) in Britain, for example, many suggests that the appropriation of certain burials were located near (not within) the large monuments by particular people marks their communal monuments (as seems to be the case close identification with these places.22 If so, at Ggantija) and finally covered by earthwork this may suggest that by burying their dead mounds. 19 In fact, a major feature of the znd - presumably, their ancestors - next to a millennium BC in Western and Northern temple (as in the case of Ggantija), the Tarxien Europe was the location of burial mounds on Cemetery people were likewise appropriating, sites with a long tradition of ritual activity identifying themselves with, and laying claim extending back into the Neolithic period (as to the earlier temple and to all that the same seems evident at Ggantija and, perhaps, at temple stood for. And as with the monument

Malta Archaeological Review • Issue 8 2006/2007 13 Front view Side view Rear view

Fig. 2. Drawing of a discoid figurine found in the 'cemetery'!cremation deposit at Tarxien. The drawing also shows the Ggan:ija discoid figurine fragment superimposed on the Tarxienfigurine. (Drawing: Joseph Calleja)

at Flagstones in Dorset (Britain) where, contents of the cinerary urns in the 'cemetery' I with time, it also became a focus for burial cremation deposit at Tarxien. When complete, grou:J.ds, 23 the proximity of the dead to the the figurines consisted of a relatively large flat earlier temples at both Tarxien and Ggantija disk, incised on both sides with an elaborate might have suggested a 'genealogical depth' to linear 'design'. The disk stood on a pair of the Les binding the Tarxien Cemetery people legs balanced by an elongated projection on to these temples and all that they represented. the rear. The figurines lacked arms. They were Sustaining such a link, the Tarxien Cemetery also faceless while their tapering heads were people could perhaps better manage the ties plain and undecorated (Fig. 2).24 between earth (themselves) and sky (their Ceramic figurines specific to the prehistoric gods/ancestors) to their own advantage. Cucuteni-Tripolye culture (Romania) but with Tierefore, the re-use of the previously a similar 'decorative' pattern were recently exisLng monuments of Tarxien and Ggantija studied by Dragos Gheorghiu (Plate 2). Like the has probably involved a change in their ones at the Tarxien 'cemetery', these figurines mea:r.ing while maintaining their sacred were, in the main, fragmented and related to charc..cter. This re-use might have reflected incomplete or fragmented skeletal material. At a new socio-ideological order based on the the Tarxien 'cemetery', the skeletal remains appropriation of the past. Therefore, the re-use were cremated. At Ggantija too, the figurine of the temples involved discontinuity but, to a fragment was found in association with skeletal certain degree, permanence too. fragments within the grey ashy deposit. In all instances, fragmentation of both figurines and skeletal material might have been deliberate The Tarxien cemetery figurines and there might have been a correspondence in meaning between the fragmented figurines A number of clay discoid figurine fragments and the fragmented or incomplete skeletal - some of which could be reconstructed into materiaJ.25 As the fragments of the broken complete figurines- were also found among the figurines at the Tarxien 'cemetery' were

14 MaltG Archaeological Review • Issue 8 2006/2007 Plate 2. Ceramic figurines of the prehistoric Cucuteni-Tripolye culture (Romania) with their chevron-like 'decomtion' (left) and the experimental modelling carried out by Drag os Gheorghiu suggesting representation ojajunerary wrapping (right). (After Gheorghiu 2001, 77, Figs 2 and 3; 78, Figs4 and 5). Note the striking similarity ofthe 'decorative'pattern on both these figurines and our discoid ones from Tarxien and Ggantija on Fig. 2.

Malta Archaeological Review • Issue 8 2006'2007 15 found together (a characteristic attributed funerary wrapping of the body. 28 Therefore, our to the 'killing' of artefacts), could it be that figurines from Tarxien and Ggantija might have these figurines were deliberately broken or been representations of completely wrapped ritually 'killed' prior to their deposition in deceased persons or ancestors. The burnt fabrics the urns as part of a mortuary ritual practice, found at Tarxien and, possibly, the carbonised as shown by examples from Minoan Crete?26 matter from Ggantija (though the true nature of Could this have taken place so that the ritually the latter is not made known) seem to lend further fragmented figurines would correspond to the support to this hypothesis. And as the Tarxien bodies fragmented through cremation? and Ggantija figurines are likewise faceless and Through experimental modelling, their tapering heads lack any decoration, the Gheorghiu sought to demonstrate that the heads of the deceased individuals or ancestors chevron-like 'decoration' incised on some of they presumably represented may also have his figurines could possibly be a representation been covered with a veil. of an interwoven funerary wrapping of The breaking of the presumed ancestors' the corpse and, thus, these figurines could figurines might have been seen as a represent deceased individuals or ancestors. 'consumption' of the figurines similar to the The ~omplete wrapping of the body is further 'consumption' of the bodies through cremation highlighted by the figurines' absence of arms (in the case of Tarxien and, apparently, also of and by their cone-shaped legs. On the other Ggantija) or through skeletal dismemberment hand, the absence of any decoration on their (in the case of Cucuteni-Tripolye). This would heads - typical of the ancestor figurines in denote a sense of recycling the past or the Cucuteni-Tripolye culture - might suggest powers associated with the past. This presumed that the heads of the deceased were not practice might, therefore, infer a belief in wrapped like the rest of the body. Taken some sort of 'power' residing in the ancestors' together with the absence of any decoration on substance which might then explain the special their heads, the facelessness of the figurines protection afforded to the bodies of the deceased would seem to point towards the possibility by wrapping them. Even any decoration of the head of the deceased having been reproducing the interweaving of bands (like covered instead with a veil. The wrapping of that seen on the above-mentioned figurines) or the corpse (and the covering of the head with of plants or mats could have been a symbol of a veil) is corroborated by osteological data. protection not least for the ancestors.29 The total wrapping of the deceased would The dead individuals or ancestors which the symbolise the 'ancestors' on the one hand, figurines from the cinerary urns (at Tarxien) and 'protection' on the otherY would seem to represent might have been Ir: terms of 'decorative' pattern, of the those whose cremated bones and ashes were absence of any arms and face, the modelling of held inside the same urns. This hypothesis the head and, consequently, representation, the is lent weight by the fact that both figurines above figurines seem to provide fairly good and cremated remains shared the same urns. parallels to the ones from the Tarxien 'cemetery' Assuming that, like those of Cucuteni­ and the one from Ggantija represented by the Tripolye, our figurines represented deceased recer:tly discovered fragment. Thus, like the individuals or ancestors, they might have 'chevron-decoration' on the Cucuteni-Tripolye been simultaneously an image of 'the world ceraoic figurines, the similar 'decorative' of the living' and of 'the world of the dead', pattern on the Tarxien and Ggantija figurines the binary symbolic structure found in every might have likewise represented interwoven traditional society.30

16 Malta Archaeological Review • Issue 8 2006/2007 Notes Antiquities of the Maltese Islands: A ~urvey, (London. The Athlone Press, 1971), 41-43). But in R. Bradley, 'The land, the sky and the Scottish the absence of any further details, the picture is less stone circle', in C. Scarre, (ed.), Monuments and clear and reliable than that at Tarxien. No m~ntion Landscape in Atlantic Europe: Perception and of the grey, ashy deposit is made in the Ta:. -Qadi Society during the Neolithic and Early Bronze Age, excavation report carried in Museum Annual Report (London. Routledge, 2002), 130. 1927-28, 11-III. 2 R. Layton and P.J. Ucko, 'Introduction: gazing on 13 T. Zammit, Annual Report of the Mf.,.cseum: the landscape and encountering the environment', I916, (Malta. Government Printing Office, 19~7), 1, in P.J. Ucko and R. Layton, (eds.), The Archaeology 3-7 and Evans, 149-151. and Anthropology of Landscape: Shaping your 14 Museum Annual Report 1936-37, VI-IX, and Evans, landscape, (London. Routledge, 1999), 14. 173. 3 Richards C., 'Monuments as Landscape: Creating 15 Author's personal fieldnotes. the Centre of the World in Late Neolithic Orkney', 16 J. Houel, Voyage Pittoresque des Isles de Sic!le, de World Archaeology, vol. 28 no. 2, (1996), 202. Lipari et de Malte, IV, (Paris. 1787), 79, pi. CCLI 4 C. Scarre, 'Introduction: situating monuments. The b,g,h (bottom). dialogue between built form and landform in Atlantic 17 W. O'Brien, 'Megaliths in a mythologised landscape. Europe' in Scarre, 9. South-west Ireland in the Iron Age' in Scarre, !58 5 Sahlqvist L., 'Territorial Behaviour and 18 As happened, for example, in Britain (see: Bradley Communication in a Ritual Landscape', Geografiska (2002), 131. Annaler: Series B, Human Geography, vol. 83 no. 2, 19 J.C. Barrett, 'The Mythical Landscapes of the British (2001), 87. Iron Age', in W. Ashmore and A. Bernard Knapp, 6 For other forms (even in evident resemblance (eds.), Archaeologies of Landscape: Contemporary to houses), see: R. Bradley, The Significance of Perspectives, (Oxford. Blackwell Publishers Ltd, Monuments: On the shaping of human experience 1999), 253-254. in Neolithic and Bronze Age Europe, (London. 20 R. Bradley, The Passage of Arms, 2"d edn., (Oxford Routledge, 1998), 15-19, 56-65. See also: C. Tilley, and Oakville. Oxbow Books, 1998), 130. 'Art, Architecture, Landscape', in B. Bender, (ed.), 21 M. Edmonds, Ancestral Geographies of the Landscape: Politics and Perspectives, (Oxford. Berg Neolithic: Landscape, monuments and me.'1lory, Publishers, 1993), 57 (Fig. 2.2). (London. Routledge, 1999), 141. 7 What appears to have been a cremation burial of 22 Bradley, The Significance of Monuments,18. the Tarxien Cemetery phase was found at Kappara 23 Edmonds, 148. Hill, Msierah in 1964 while clearing a new building 24 Zammit, 5, and Evans, 160-162, pis. 56-58. site. But the site was disturbed to such an extent, that 25 D. Gheorghiu, 'The Cult of Ancestors in the East any remaining evidence is deemed inadequate (see: European Chalcolithic. A Holographic Approach', Museum Annual Report 1964, 3, pi. 1). in P.F. Biehl and F. Bertemes with H. Meller, leds.), 8 Homer, Iliad XVIII, 408-412. The Archaeology of Cult and Religion, (Budapest. 9 Ibid., XXIII, 187-208. Archaeolingua, 2001), 76-79 (including Figs.1-6). 10 Ibid., 270-295. 26 J. Chapman, 'Object Fragmentation in the Neclithic 11 Ibid., XXIV, 921-938. and Copper Age of Southeast Europe' in Biehl and 12 Making reference to Zammit's field-notebook, J.D. Bertemes with Meller, 90. Evans mentions the presence of early Bronze Age 27 Gheorghiu, 76-79 (including Figs.1-6). (Tarxien Cemetery phase) material in grey, ashy soil 28 The 'decorative' pattern on our exa:nples at Tal-Qadi temple ruins on the way to Salina Bay, generally runs diagonally and, in fewer instances, suggesting a later use of the building in the Bronze perpendicularly. However, the wrapping bands could Age, possibly to house a cremation burial like those be 'interwoven' both diagonally and perpendicularly. at Tarxien, even if no mention of any cremated bone 29 Gheorghiu, 83. is made by Zammit (see: J.D. Evans, The Prehistoric 30 Ibid., 76.

Malta Archaeological Review • Issue 8 2006/2007 17