Chapter IX
CONTINUITY AND CHANGES IN THE ELITE FOOD MANAGEMENT DURING THE 4™ MILLENNIUM BC. ARSLANTEPE PERIODS VII AND VI A: A COMPARISON*
Maria Bianca D'Anna, Paolo Guarino
The aim of this chapter is to define the changes in the nature of the centralised economy of food at Arslantepe in the 4th millennium BC through the comparison of the archaeological evidence from the public buildings of periods VII (LC 3-4) and VIA (LC 5). Thanks to several years of dedicated excavation, Arslantepe offers the opportunity to follow, through fine-grained archaeological evidence, the development of complex socio-economic processes from their formative phases to their more elaborated configuration. This is the case for the formation process of a centralised economy system during the 4th millennium BC.
Excavations of the early phases of period VII, in the western part of the mound of Arslantepe, unearthed some residential structures that, due to their scale and internal features, stood out from the other buildings of more modest nature discovered on the north-eastern part of the site. This variability in size and architectural layout of the domestic structures had already suggested the presence of some forms of social stratification and possibly the emergence of elite groups, but it was with the discovery of Building XXIX or Temple C (end of period VII-LC 4) that the discourse of social complexity in Arslantepe was elevated to an entirely different scale. It provides the evidence for redistributive activities directly associated with an administrative system through which the economic transactions were controlled and recorded. Building XXIX was an imposing and monumental structure whose large central room (A900) had a central platform and decorated walls. Its floors and fills yielded a large quantity of mass produced bowls, many clay seahngs and few fragmentary storage jars'. These finds, interpreted as evidence of redistribution act.vit.es under admimstrati ve control, suggest that these practices were somehow embedded in a ceremonial setting where large numbers of people were brought together in an ideologically charged environment. At the end of period VII a centralised economy seems to be already established at Arslantepe. the elites of these periods might have been able to monopolise econom.c resources and control the labour force.
* We would like to thank Susan Pollock for having read the final draft of this paper also suggesting future research directions.
' Guarino 2008; Frangipane 2000c; 2001a.
193
Economic Centralisation in Formative States. Chapter IX
Studí di Preistoria Orientale 3
In order to substantiate these interpretations it is useful to look at Earle's2 theory of ideology which he describes (also in line with neo-Marxist theories) as an apparatus of ideas and beliefs fashioned and manipulated by elites in order to reinforce and legitimise their control over economy and therefore their power; its translation into physical events like ceremonies, rituals, symbols and monuments is the 'materialisation' of ideologies. Among the materialisation processes indicated by Earle, which allows those ideologies to be controlled, one that is particularly relevant for our case is the organisation of ceremonial events that can be controlled by restricting participation to certain groups of people, for example, thereby demonstrating to the community that they are able to afford the costs of ceremonies and legitimising their right to perform such ceremonies.
Ceremonies and public events are powerful means of social power negotiation, thanks to their immediacy and their ability to involve large portions of the community3. The frequent association of feasts and ceremonies with public consumption of food and/or drinks provides an important connection with the material remains that archaeologists may expect to find in sites where these kind of ceremonial practices were performed. Archaeological indicators of collective consumption of food are, for example, monumental spaces, large concentrations of serving vessels as well as cooking and storing areas4. These thoughts are important when it comes to the interpretation of the archaeological data offered by Arslantepe VII and VI A.
In anthropological and archaeological research, a growing interest has been recently paid to feasts, as ritualised communal practises in which food and drinks are shared. Feasts can create, reproduce and manipulate social relations of inequalities. They are also a powerful means of converting economic capital into symbolic capital and may stimulate and help processes of goods and prestige accumulation. These approaches appear particularly fruitful when looking at political and economic systems based on the centralisations of primary goods and their redistributions in the form of food rations.
Following these lines of thoughts, it can be argued that what we see at the end of Arslantepe VII is the picture of an elite group that is powerful enough to build a monumental structure, wh.ch Building XXIX must have been, able to control and monopolise a certain amount of staple products (hence the redistribution activities at least partly controlled through the use of clay seal.ngs). but also still in need of the ideological legitimisation of its power, wh.ch would explain the ceremonial setting provided by Building XXIX. In favour of this hypothesis is also the tripartite layout of the plan of this building. In 2003 Frangipane5 discussed the importance ot the tripartite plan tradition that traces back to the Ubaid domestic structures m Mesopotamia and is then transferred to public buildings in northern Mesopotamia towards tne end ot Lba.d period. Frangipane argued that the transition of the tripartite plan from domestic to public buildings is charged with important symbolic meanings as it was probably underlain by the emergence of leading groups within the community. Frangipane also remarks ecTnomf ^ma{ ^ to occur during the 4* millennium B.C ,the emergence of I Ire    7 Cha0ged fUrther the function of th^e public buildings, which became
new cín Í fV aSS°Ciated the administrative system that developed in support of the new centralised economies.
: Earle 1997.
' De Marrais et al. 1996.
' Dietler. Hajden (eds.) 2001.
" Frangipane 2003.
Period VII Building XXIX (Temple C)
In this chapter we present the archaeological evidence from period VII public structure which, as mentioned earlier, provided the first indication of centralised economy in Arslantepe. Building XXIX (fig. II.6) found on the western slope of the mound, just to the north of period VI A palace is an imposing and monumental structure built on the highest spot of the Chalcolithic mound raised by an imposing foundation made with large stone slabs and layers of mud. The architectural layout reflects the tripartite module, with a large rectangular main room connected to smaller rooms located along both its longer sides. This type of plan is clearly similar to the monumental buildings so typical of the Ubaid period in Mesopotamia and little known in the upper Euphrates area6. The scale of Building XXIX has no comparison in any other contemporary building excavated in the prehistoric levels of the site: the central room (A900) is about 18m long and 7m wide, its walls are 1.6m thick. The thickness of the walls could have supported either a very high ceiling or a second floor. The internal structure is also indicative of a building of exceptional quality and characteristics. A plastered rectangular platform of about 3m by 5m stands in the middle of the main room. Traces of three hearths were found, one of them on top of the central platform, the other two north of the platform along the long axis of the room; next to these, fragments of andirons were found. Wall niches were located in direct association with the doors leading to the side rooms. In the north-eastern corner of the main room the walls were decorated with figurative paintings. Only the two rooms on the east side of the central room were preserved, the western side of the building was badly damaged by the natural erosion of the mound. Of these two side rooms the northernmost (A950) was very well preserved. The walls were still standing to more than one meter and the internal deposit was well sealed and virtually intact, only the south wall of the room was partly ruined by a large modern trench. A door on the west wall of the room connected this area with A900; a brick bench was found along the corner between the east and the south wall of the room. The southernmost of the two side rooms (A932). partly damaged by the same modern trench, also had direct access to A900.
The materials that were found on the floors of Building XXIX have provided us with a wider understanding of the kind of activities that were carried out within (fig. IX.1).
Scattered on the floor of the central room (A900) and in the layer that covered the floor, almost 600 mass produced bowls made of Chaff Coarse Ware were found (tig. IX. 1: c) , along with the fragments of six storage jars and two small beakers (fig. IX.2: a, b /). Apart from two fragments of andiron found next to one of the two small hearths to thenorth ot the central platform and two flakes of flint, no other artefacts lay on the floor of A900.
Figures for the A950 (the northern small room) ceramic distribution show a very similar trend to the one already seen in A900 (fig. IX.i: b). About 300 mass produced bowls were found between the floor and the fill8 of the room together with six smaII cylindrical beakers one pithos and 6 carinated beakers. Virtually no other artefacts were found on the floor of the room.
-__ f the . !baid slur of Deg.rmentepe in the Malatva pla.n see
* Fransipane 2001a: 9; Frangipane 2003. For the case of the Loa also Esin, Harmankava 1991; Stein. Ozbal 2006. ca,culation of an 'estimated vessel equo alem « >rt„! ,r
Guarino 2008: the number of the bowls is the result of a calcu.a
'993: 21). f.   m o{ the room suggests that son* bowls -re stored or, , ^„„
8 The concentration of bowls in some areas of the till oi storey of the room or at least on shelves above the floor.
194
195
Economic Centralisation in Formative States. Chapter IX
Studí di Preistoria Orientale 3
Economic Centralisation in Formative States. Chapter IX
Studí di Preistoria Orientale 3
The materials found on the floor of A932 (the southern small room, fig. IX.1: a) are quite consistent with those found in the rest of Building XXIX. Again the quantity of the mass produced bowls (roughly 220 in total) is impressive; these together with six small storage jars and a funnel are important for the possible interpretation of the activities carried out in the room. Two cylindrical beakers and one carinated beaker were also found. The bowls, the cylindrical beakers and the funnel were produced from the chaff coarse ware while the other objects had a red slipped and burnished surface. Together with the bowls and the other pots, a number of clay sealing fragments were grouped on the floor of A9329.
Assuming that Building XXIX was abandoned to a slow decay rather than suddenly destroyed10, we have to keep in mind that the portable materials found in it might represent only part of the overall assemblage that was in use during the last phase of life of the building. Attempting nonetheless an interpretation of the archaeological evidence, the presence of only few functional classes might enforce the idea that very specific activities were carried out in Building XXIX. Also the trend in ceramic distribution seems to be quite consistent for the whole building. The association of several hundred bowls with some jars and beakers would seem to suggest that the activities carried out in Building XXIX consisted of large-scale communal events during which foods or drinks (or possibly both) were distributed and consumed. The fact that in the side rooms (A950 and A932) the bowls were found stacked upside-down, and that in A932 they were associated with a fairly large number of cretulae (clay sealings) suggested that the two rooms were used as store rooms where both the bowls ready for use and cretulae already removed from containers were kept. The cretulae had been used during the transactions involved in the distribution of meals.
Although the absence of cooking pots and the rarity of food storage facilities in Building XXIX could be due to the reuse of these objects in other contexts after the abandonment of the structure, it may also suggest that storage and preparation of the foods and drinks consumed during the ceremonies took place somewhere nearby and that only the required amount of foodstuff would be brought into the building.
Period VI A Temple A and Temple B
In the last centuries of the 4th millennium BC, the large area exposed so far was mainly occupied by public buildings. These, built next to each-other, progressively expanded, filling up the south-western flank of the Chalcolithic mound11.
The public buildings built during period VI A on several terraces constituted a multifunctional complex of highly inter-related structures, characterised by a high degree of internal specialisation: areas with prevailing religious and ceremonial functions, others intended for more representative, economic or administrative functions. Redistributive and administrative activities as well as commensal consumption of meals were carried out at different scales and torms in most of the spaces of the VI A public complex12
' Me/zasalma 2008.
1 Building XXIX did not show any signs of sudden destruction, in fact the traces of bat bones retrieved from the fill of the central room. A900. suggested that the building was abandoned and left to decay slowly. This evidence pro\ides some important implications regarding the formation of the archaeological deposit that created the remains of the building as it was recovered during the excavation.
1 Alvaro in this volume.
: D'Anna in this volume: Frangipane 1997; Frangipane (ed.) 2004; Frangipane et al. 2007.
198
The location of the public buildings within the Chalcolithic tell and in relation to the surrounding territory as well as the orientation of the long corridor place the complex as an interface between the plain and the higher area of the VI A mound occupied by the residential buildings13.
Temple B is highly integrated in the palatial complex. Its only access faces the entrance of what seems to be a massive structure (A830)14. At the moment of its destruction. Temple B was still in use as a ceremonial building. The materials were concentrated in the main room A450 (fig. VIII.5): pithoi, sizeable cooking pots, large necked jars, and two bottles of distinct Late Uruk type lay in the north-western and south-eastern comers. Four high-stemmed bowls and fewer than 20 mass-produced bowls were also found on the floor of the room. It is possible that Temple B hosted ceremonies during which high quantities of food and drinks were consumed by a restricted number of people. Evidence for limited administrative control over circulation of foods is offered by a small amount of cretulae set aside on the upper floor of the small entrance rooms. At any rate, as suggested in chapter 8, the levels of exclusions and inclusions of other members of the VI A society from these apparently diacritical feasts appear to be complex15.
Temple A "was built at a slightly later stage on a terrace on the western edge of the tell, even though in view of the horizontal stratigraphic situation it may also have been in use at the same time as at least part of the other structures"16. The two small entrance rooms were so full of large and medium-sized vessels that it seems unlikely that people could walk in and through these rooms in the final phase of use of the building (fig. VI1I.4). In the main room A42 some mass-produced bowls and high-stemmed bowls lay on the floor, while in the former stairwell of the building (A77) animal bones, ceramics and cretulae were discarded. These clay sealings suggest that redistribution activities were carried out in the temple before its destruction.
In the final phase of its life, before its destruction, Temple A was probably used to store goods and possibly also to redistribute them on a rather meagre scale, as suggested by the small number of mass produced bowls. It is possible that in the main room A42 ritual and or feasts could have been hosted, as the presence in A42 of high-stemmed bowls and large mass-produced bowls suggests. Other open shaped vessels lay in the entrance room A46. Cooking activities are witnessed by the large number of cooking pots, whose capacity does not reach beyond 11 litres17. Unique is the presence of a human skull found nearby the basin laid against the end wall of the room, together with a small cluster of animal bones18. Activities of a different nature were performed in Temple A, possibly in different moments.
Comparisons between VII and VIA
a the outline of the public buildings, the most Besides the general structure of the site ana me ^ ^ the LC 4 and
meaningful changes which took place at Arslantepe dunnc
_ .        Arslantepe-s surroundings is noi clear yet. Dunng the
» Actually our knowledge of the Chalcolithic occupation_ot Ana ^   ^ ^ . ^ lhc liccumn«, ot
three-year survey project on the Malum plain, only live si      ^ ^ ^ ^ a/ 2(W8)
LC 5 sherds in these sites are often extremely scanty <L» • 14 See Alvaro present volume. 13 d"Anna present volume.
'* Frangipane 1997; 49. k     [(pm of the Temp|e A cooking pot assemblage resembles thai
" As for mean capacity and standard deviation the pattern of"the        ^ ^ R w,kmg ^
of the containers intended for cooking found in the resia lx Frangipane, Palmieri 1983a: 321.
199
Economic Centralisation in Formative States. Chapter IX
Studí di Preistoria Orientale 3
LC 5 periods relate to a stronger economic centralisation and the pervasiveness of administrative control. A further comparison of the socio-economic systems of Arslantepe VII and VI A societies can be attempted through the analysis of the functional and organisational aspects of pottery production. The organisation of pottery production19, fabrics, shapes and surface treatments of pottery changed markedly from period VII to VI A. Probably these changes were linked to different social organisations and to the new relationships that VI A communities experienced with the Late Uruk communities and the Anatolian world. Nonetheless it is also evident that one element of continuity between the two periods is the presence of the mass production of bowls, which is deeply and structurally linked to the political and economic organisation of these Late Chalcolithic communities.
Although this particular class of pottery is very common in most of Arslantepe VII contexts, the massive concentration of bowls and their distribution in Building XXIX is unique and represents a statistically meaningful assemblage. All the bowls found in Building XXIX share similar characteristics. They are all made with chaff-tempered ware, the surfaces are normally coarse or roughly smoothed and generally the core of the vessels is not fully oxidised. These characteristics and a great degree of formal variability seem a direct consequence of the production system, which possibly focused more on speed and quantity than quality. Apparently in contradiction to the commonly accepted idea that increased specialisation produces higher standardisation of the product20, we argue that the bowls could have been the product of specialised workshops where the production targets were more related to quantities and speed of production than formal standardisation of the products. For the purpose of this analysis the bowls can be divided in three groups21 on the basis of differences in the manufacture processes visible on their surfaces: some of them had the external surface 'regularised' around the base with a scraper (fig. IX.2: e, g); others had 'string cut'22 marks on their base; and finally some others only had a simple flat base (fig. IX.2: c,f, h). The presence of these differences among vessels that were made roughly at the same time and used contemporaneously in the same building is evidence of the coexistence of different manufacturing techniques23. This idea seems to be supported by the presence of at least five bowls, found in room A900, which presented both string-cut and flint-scraping marks (fig. IX.2: df\ It is important to note, though, that the use of the 'flint scraping' technique is already recorded in the earliest layers ot period V II, whilst the appearance of those signs that are usually associated with the use of a rotating device (string cut) is documented towards the end of period VII. The use of rotating dev ices to manufacture ceramic vessels in Arslantepe seems therefore to be inversely proportional to the use ot the scraping technique. Possibly the increasing skills of the potters in the use of rotating devices progressively reduced the need to regularise the profiles of the vessels with the help ot scrapers. This could also be argued on the evidence that the mass-produced bowls ot the subsequent Arslantepe period VI A were all wheel thrown and none of them showed signs of scraping (see D Anna in this volume, fig. VHI.la). The occurrence of the typical
mould/hand-made Late Uruk Bevelled Rim Bowls is very scanty in all period VI A structures. The VI A wheel-thrown truncated conical bowls appear to be a possible outcome of the manufacturing and shaping processes visible in various period VII levels and contexts.
Large concentrations of mass-produced bowls are a trademark of many Chalcolithic sites in Mesopotamia. Their abundance, during the 4th and 3rd millennia B.C.. has inspired an ongoing debate in which several scholars have argued about the possible interpretations of the function and use of these artefacts. Many different uses have been proposed so far: ration containers25, bread moulds26, taxation bowls, presentation bowls27.
In the case of Arslantepe VII and VI A the hypothesis that the bowls could have been used to distribute and consume rations of food is backed up by convincing evidence provided by the concentration of the bowls in a building with specific ceremonial features and their contextual association with clay sealings.
Another piece of evidence in support of the idea that Arslantepe VII bowls in Building XXIX were used to distribute food/drink rations rests on the fact that their volumes seem to cluster around two main peaks (between 400-600 ml and 800-900 ml)28. Drawing on Johnson 1973, this could be explained with the intention of providing simultaneously different quantities of food for different 'types' of labourers, possibly distinguished by gender, age. and status. Period VIA mass-produced bowls have a more closed shape and a lower degree of variability: they are smaller than period VII ones, nonetheless their volumes also tend to cluster around two peaks (much as those of period VII), 500-600 and 650-800 ml29. Frangipane has compared VI A bowls' trend with the rations recorded in texts dated to the Jemdet Nasr period, linking the differences in absolute volumes of Arslantepe bowls (also seen at Habuba Kabira) with Uruk III texts. She also stresses the relatively consistent occurrences of small-medium-large bowls and rations30.
Comparing the architectural layout of the monumental areas of the two periods, it is immediately apparent that Building XXIX was built as an isolated structure in a high spot ot the mound and raised by large foundations which increased its monumentality and visibility. The activities performed in its rooms were probably supported by auxiliary structures like storage areas or cooking areas separated from the main building. Conversely, the public buildings of period VI A form a complex system of rooms and open spaces complementary to each other that met specific functional needs. Frangipane argued31 that the transition from the tripartite layout of Building XXIX to the bipartite plan of the two temples in the period VI A palace is the material evidence for an important shift in the socio-political organisation ot Arslantepe's society. The monumentality and the elevated position of Building XX X seem to suggest that the main function of the building was to host large numbers of people during public ceremonies and represent a physical landmark for the -c0^mun,ty-7h^f!trln ! the ideas, discussed earlier, on the importance of monumental buddings and public ceremonie, in the process of ideological legitimisation of power.
" The use of potters" marks (see for example in fig. IX.2: c.f. h), distinctive of period VII and indicative of t^e presence of different workshops (see Trufelli 1994 for more detailed discussion), disappears in period VI A. At e ■~ame time the use of the potter's wheel became more widespread.
: Costtn. Hagstrum 2005: Rice 1991.
:l For more details see Guarino 2008.
:: These signs are left b\ the wire or string used to detach the pot from the wheel or tournette on which it was
shaped (Rye 1981: 75 and figs. 63-64).
These techniques v^ere possibly applied at different stages of the production process. 11 Frangipane 2000a: 442-443.
25 Gelb 1965; Johnson 1973.
:6 Millard 1988: Chazan. Lehner 1990.
- Beale 1978; Nicholas 1987. See also Btackman et al. IW,
1996; Pollock 2003 for a debate on siandarJi«t.oii
issues.
28 Guarino 2008.
29
VJUOIIHU £OVIO. .    . IQ02.1.
Frangipane 1989: 53; see also: Frangipane. Palmien w. storerooms and residential buildings: D'Anna 2010. 3" Frangipane 1989: 56-59. 31 Frangipane 2003.
, 343.5; and for the solume of the MBPs found ,n ihc
200
201
Economic Centralisation in Formative States. Chapter IX
Studí di Preistoria Orientale 3
The temples of period VI A were part of a much larger administrative and ceremonial complex with a series of other structures. They were much smaller than Building XXIX and access to their main rooms was restricted to a small side door. They were not built to stand out in the "urban landscape' of Arslantepe. Rather the mound itself possibly represented a landmark in the surrounding landscape32. Hence they do not seem to be designed to host ceremonies for large numbers of people; indeed the whole public complex of period VI A seems to reflect new needs of storage for centralised goods and frequent practices of administration as well as redistribution. A450 is ca. 60 square meters and A900 126 square meters. Assuming that 1.9 people can sit in one square meter33, up to 114 people could have been seated in A450 at the same time and about 240 in A900. In fact, unmovable and movable features as well as portable artefacts strongly reduced A450's free surface: a large table and a shorter structure stood against the south-western wall; another short platform was located by the long north-western wall, facing the large central fireplace; two windows took the whole central part of the long south-eastern wall; finally two large group of vessels occupied two corners of the room, and the door to the passage room A800 was just next to the north-eastern corner. It seems that the whole eastern part of the room was an area of passage from the room's entrance toward the end wall. Temple B could host a maximum number of 35-40 seated people. The number of bowls found in situ in Temple B further corrects the calculated data: less than 20 lay on the A450 floor. Conversely the estimated number of people circulating, sitting down, and eating in the Temple C main room, if compared to 600 bowls found only in A900, is possibly underestimated.
The presence of painted decoration in domestic and public architecture is an original tradition of Arslantepe and other south-eastern Anatolian local Late Chalcolithic communities34. During period VII, walls in all kind of buildings (domestic and non domestic) are decorated with geometrical and figurative patterns painted in black and red on the white plasters. Wall paintings were rather common and characteristically recovered in living spaces. In period VIA the residential buildings probably also had painted decorations. Both painted and impressed decorations in public buildings, were mainly located in corridors or passageways. These were probably the most visible places, accessible to high numbers of people, and also the passageways might have had a special decorative treatment because they could have been perceived symbolically as 'special' places. The shift in the location of wall decorations from main rooms to passageways seems to confirm also the possible exclusion of high numbers of people from the mam rooms of the LC 5 buildings.
Dietler links the concept of redistribution as traditionally meant in economic anthropology to n.s patron-role feast', which "involves the formalized use of commensal hospitality to symbolically reiterate and legitimize institutionalized relations of asymmetrical social power"35
be rial    '1z h n '\   NCar EaStem iandsCaPes (Steadman 2005)- As for the case of Arslantepe. it can
1 ,tu,n £ h Tlf JT %C°nr °f     VI A ~eS: at the <* *e «Hv Bronze age, the
pic e       l!nr       R°-31 T°mb "earb> the ruins <* ^e late Chalcolithic palatial complex appears "an act
T^^Z^j^ 15 sure -the fact that Arslantepe continued to a »9*^ p'a£even ■**
^r^ZVt^T    7* 1" h,St°riCal,-V <and also ^tegicallv) import enough to be chosen
^M^^T^ I* 3 fUnCti°n °f ltS area" and 1 Calcula,e ma*im»m occupancv based on modern
4 Frangipane 1992. ' :-i Dietler 21)01: 82-83.
The process of "institutionalization of authorities relies on this binding asymmetrical commensal link between unequal partners in a patron/client relation"36. Another interesting case is that of the so-called 'work feast'37. In this case, hospitality and meal offers are linked to collective work events. Work feasts may vary from a totally volunteer and communal event to a completely obligatory occasion. The latter overlaps corvee labour in communities with an "institutionalized central authority" and in this case "a ritual of commensal hospitality subtly serves to euphemize the nature of the transaction" of work and food. These 'hosting' practices may become a catalyst for social differentiation: the sectors of the society that is able to ask work and refunding it with food may acquire more economic and symbolic power thanks just to this work and feasting events, while other sectors become a real labour reserve.
Focusing again on the Arslantepe VII case, the frequent use of the mass-produced coarse bowls to redistribute and consume meals in a ceremonial context of public commensality seems to be evidence for a materialisation of asymmetric relations through an ostensible emphasis on equality. In the period VII case this is much more marked than in the subsequent period VI A: Building XXIX, characterised by high accessibility from the outside and high potential hosting capacity, is not only a 'topographically high' place, but also a place that belonged to everybody, the place where everybody, in an exchange of labour, received their share of food, which, in turn, was the result of a process of monopolisation of resources. The ritualised practice of commensality overlaps with and, at the same time, is redistribution. Commensality and redistributions might have been extraordinary practices, possibly related to occasional collective works which did not take place on a daily basis.
During period VI A, a more consolidated elite group that controlled a more structured centralised system did not rely on ideologies as much as before in order to legitimise its power38. Centralisation is much higher in period VI A, as shown by the large amount of goods stored in the public buildings. At the same time, redistribution activities are also more complex and frequent. Different practices of formal commensality have been hypothesised for different spaces of the palatial complex39: what was taking place in Temple B is very different from what was taking place in A340 and the back courtyard adjacent to it as in Temple A. The thorough study of the administrative materials has also displayed a large number of officials involved at different levels in a variety of transactions.
Organisation and management of a centralised economy in Arslantepe achieves its most complex and articulated shape during period VI A. The distribution of portable artefacts w.thin the archaeological contexts and their architectural nature shows that all the activities related to the management of centralised resources were carried out in an organic complex where every element had a specific and indispensable function. A highly developed administrate control over food transactions was a key element in the life of the complex Food storage, preparation and redistribution were performed within an organ.sed and controlled env,mnment where ceremonial activities were restricted to a relatively small number of from period VII of Arslantepe suggests that earlierfonms«, ce^~^Z on the ceremonial aspects of a r—tion Jj^^^ of administrative
occasions and not on a regular daily basis. And also, retrospei.ii very, i activity was smaller than in the following period.
36 Ibid.: 83.
37 Dietler, Herbich 2001.
38 Frangipane 2001a; Helwing 2003.
39 D'Anna present volume.
202
203
Economic Centralisation in Formative States. Chapter IX
Overall we believe that the remarkable differences between period VI A and VII should not be regarded as evidence for two different economic systems but rather as the result of the development of the centralisation process from its earliest shapes, visible in period VII, to the more defined characteristics of period VI A.
204