Architecture
Fig. 23: Bderi, Level 20-14, "allotment house", House XIV, reconstructed original plan, EJZ 2 to 3a
(Pfalzner 2001: PI. 20).
"allotment house" size category.113 However, there are no indications of room function or doors, so it cannot safely be attributed to this kind of house concept. Houses C3, C4, C5, and C6m were not fully exposed, but their widths along the adjoining alley of around 9 and 6m make it plausible that they were constructed on allotted plots.125 .,
A particularly spacious "allotment house"126 was excavated at Abu Hujaira I, Area A, Level 3, dating to Period EJZ 3a (Fig. 24).127 It was built on a rectangular block, 11.2m in width, in a prominent position on
12' Pfälzner 2001: 303, Taf. 28.
124 Lebeau 1993: 100-102, PI. 28-40.
125 Pfälzner 2001: 303.
m The excavators had originally interpreted these rooms as belonging to several small houses (Martin & Tietze 1992:267 f.), this was, however, rejected by Pfälzner (2001:316 f.) on the basis of regional comparisons.
127 Lutz Martin, pers. comm.
155
Fig. 24: Abu Hujaira I, Area A, Level 3, "allotment house", EJZ 3a (Martin & Tietze 1992, Fig. 3).
the Tell.m It possessed the usual characteristics: an entrance corridor with a drain, a small central courd with attached to««Mr-chambers, a grinding room (Room 4), several other smaller rooms, probably for age accessible from the courtyard, and a main room located at the front of the house, beside the entrance parallel to the alley.12' This room must have been divided off from the house in a secondary phase and r directly accessible from the alley.
An interesting subtype of EJZ 3a "allotment houses" was discovered at Abu Hufur, in Levels 3-4.130 j have been labelled "allotment houses of the Abu Hufur-type"m. An EJZ 3a date can tentatively be assil for these two levels.132 Three houses (Houses 2, 3 and 5), of which the total area was excavated, proved 1 approx. 8m wide each, measured on the front side along the alley133 (Fig. 25). This hints at a specific plot! for "allotment houses" at Abu Hufur differing from the plot sizes at Khuera or Bderi. The interior layout oj rectangular houses resembles those from other sites: there is an entrance corridor with a drain, and a main r| located beside the corridor at the front side of the house. House 2 illustrates this: Room 14 was the entrance Room 13 the main room, and Room 19 functioned as a grinding room (Fig. 26). There is, besides the plot, another specific feature of the Abu Hufur-type of allotment house: some of the rooms, including the main* were equipped with interior buttresses on up to four of the walls, possibly once forming arches or quarter-art' These might have served, as was the case in many EJZ 2 houses (see above), as corbelled supports for horizcj roof beams. This feature combines the older (EJZ 2) tradition of houses with buttresses with the newer (EJJ concept of "allotment houses".
The domestic quarter in the Lower Town South at Leilan is difficult to interpret, as no detailed plans j been published so far. Level 7 dates to Leilan Period IIA (= EJZ 3a) and was excavated on a larger scale.134 Base'
I;
128 Martin & Tietze 1992: 267 f., Abb. 3-4; Martin & Wartke 1993-94: 206-208, Fig. 9-10.
129 Pfalzner 2001: 317f.,Taf. 38.
130 Kolinski&Lawecka 1992: 181-189, Fig. 3-4; Kolinksi 1996a: 139, Fig. 4-5.
131 Pfalzner 2001: 316.
132 Based on the preliminary analysis of the Abu Hufur pottery Levels 1-2 were dated Late ED III/ Early Akkadian (=. 3b) - on analogy to Bderi (Area 2965) -, while the older layers 3-4 were dated late ED II/ ED III (= EJZ 3a, probably alsi of EJZ 2); see: Kolinski & Lawecka 1992: 206-20"
133 Pfalzner 2001: 314-316, Taf. 36-37
134 Weiss 1990 a: 1990 b, 201-209: 1990c: 1991a, "06.
156
1
v,ut:.
!
if:
iL
'V. ?
n H
Fig. 22: Bderi, Level 17, "allotment house" House XVII, EjZ 3a (Pfalzner 2001: PI. 21).
5 rn
10m
Fig. 23: Bderi, Level 20-14, "allotment house" House XIV, reconstructed original plan, EJZ 2 to 3a
(Pfälzner 2001: PI. 20).
"allotment house" size category.123 However, there are no indications of room function or doors, so it cannot safely be attributed to this kind of house concept. Houses C3, C4, C5, and C6124 were not fully exposed, but their widths along the adjoining alley of around 9 and 6m make it plausible that they were constructed on allotted plots.125
A particularly spacious "allotment house"126 was excavated at Abu Hujaira I, Area A, Level 3, dating to Period EJZ 3a (Fig. 24).127 It was built on a rectangular block, 11.2m in width, in a prominent position on
123 Pfalzner 2001: 303, Taf. 28.
124 Lebeau 1993: 100-102, PL 28-40.
125 Pfalzner 2001:303.
126 The excavators had originally interpreted these rooms as belonging to several small houses (Martin & Tietze 1992:26" f.), this was, however, rejected by Pfalzner (2001:316 f.) on the basis of regional comparisons.
12T Lutz Martin, pers. comm.
155
Fig. 25: Abu Hufur, Area A, Levels 3-4, reconstructed organisation of house plots for "allotment houses", EJZ 3a
(Pfalzner2001.pl. 36).
Fig. 26: Abu Hufur, Area A, Levels 4 and 3, "allotment house", House 2, EJZ 3a (Pfalzner 2001, PI. 37).
the aerial photograph135 it is possible to assume that the quarter was organised according to the "allotment house" principle.136 The domestic quarter had already been founded in Leilan Period Hid (= EJZ 2), corresponding to Levels 8-12,"' and shows some regular traits already in its first phase - a straight street and house blocks with straight outer walls.138 Thus it might well be that regular planning and the allotment of house plots was already in place in Period EJZ 2 in Leilan's lower city.13'
5.3.5 Period EJZ 3b
The development of the domestic quarter from Period EJZ 3a to EJZ 3b can be best observed in Area K ("Kleiner Antentempel") at Khuera. Levels 3-1, dating to period TCH ID (= EJZ 3b), developed without interruption from the houses of Levels 4-8 of the previous EJZ 3a period. However, there was a fundamental change in the function of this quarter in Period EJZ 3b. In Levels 3 to 1, a small temple, the so-called "Kleiner Antentempel" was situated in this area.1-10 This has recently been interpreted as an ancestor shrine.'41 It may have resulted because
135 Weiss 1990b: Fig. 3.
136 Pfalzner 2001: 320 f„ Taf. 41.
137 Weiss 1990b: 205-209.
138 Weiss 1990b: 201; 1991,706. 135 Pfalzner 2001:320.
140 Moortgat 1965: 11-17, Fig. 3,5.6, Plan V-VI; 1967, 8-25, Fig. 3-9, 1".
141 Pfalzner 2001: 336-337.
H XII / .-'
Fig. 27: Khuera, Area H, the organisation of plots for "allotment houses", EJZ 3b (Pfalzner 2001: PL 49).
of the long-standing existence in this quarter of earlier households belonging to Periods EJZ 3a and EJZ 2, which were possibly regarded as venerable ancestors of an extended descent group by this time. Parts of the older EJZ 3a houses were re-used and modified to serve as additional rooms in this small sanctuary.1*2 As a consequence, the original structure of the EJZ 3a domestic quarter, which had been based on a regular arrangement of "allotment houses", no longer existed by Period EJZ 3b. Thus, Area K is a good example of how considerably domestic patterns changed by Period EJZ 3b.
In Area H ("Hduserviertel") a different kind of development can be observed. Here, there is ample evidence for the existence of "allotment houses" in Levels 6-3143, which date to Period TCH ID (= EJZ 3b).144 The houses of the main Levels, 4-5, in Area H illustrate the remarkable regularity and standardisation of the "allotment houses" (Fig. 27). In the N part of the domestic quarter there are three houses with a standardised plot width of 9m (Houses H I, IV, V), and two houses measuring 15m in width (Houses II, III).145 These dimensions reflect the basic 6m measurement of many allotted plots, as attested at Bderi, Melebiya and Khuera Area K. This corresponds to the Mesoptamian nindan-sjstem (1 nindan = 6m).146 Thus Houses I, IV and Vhave house plots of 1.5 nindan, whilst Houses II and III have a plot width of 2.5 nindan. Interestingly enough an alternative system is used in the S part of the domestic quarter. Here, all excavated houses have a plot width of 8m (Houses VI, VII, VIII, IX). They are standardised as well, but follow a different measurement-system, which was, for example, also attested at Abu Hufur in Period EJZ 3a with its 8m wide "allotment houses" (see above), indicating that both systems existed in both EJZ 3a and 3b across a wider region.
With regard to internal social differentiation at Khuera it can be noted that the "allotment houses" in Area H (^Hduserviertel") are generally larger than those in Area K (^'Kleiner Antentempel") (see above for Period EJZ 3a), thus suggesting that the households in Area H were larger or wealthier than those in Area K. The interior layout of the "allotment houses" in Area H is similarly standardised. The 1.5 and 2.5 nindan houses all have an entrance corridor with a drain at one side of the house front and an elongated rectangular main room attached to the corridor on the front side of the house and parallel to the alley, such as Room 2 in House H I147 (Fig. 28). The main rooms of these houses are nearly equal in size, hinting at the use of standardised building principles'. A large courtyard is situated behind this front row of rooms. On the other hand, the southern, 8m wide "allotment houses" have their entrance corridor in the middle of the front, and a much smaller main room, nearly square shape, next to it on the house front, such as Room 41 in House H VIllS (Fig. 29). Thus, the two alternative systems of plot dimensions are associated with a different house-building concept.
12 Rooms 1, 2A/B, 3, 4, 5; see: Moortgat 1965, 15, Plan V. " Moortgat 1960a: Fig. 1-2; Orthmann, Klein & Lüth 1986: 6-33, plan II-III; Moortgat 1960a: 32-39; 1960b: 2-12, Fig. 1-2. 44 For the pottery dating these levels see: Klein 1995: 110-117, Fig. 63-64.
15 Pfälzner 2001: 325-334.
16 Pfälzner 2001: 396-397, Fig. 120.
17 Pfälzner 2001: 326-330, Taf. 50-54.
18 Pfälzner 2001: 330-334, Taf. 55-5".
158
Architecture
Fig. 28: Khuera, Area H, House H I as an example of a 1.5 nindan "allotment house' EJZ 3b
(Pfalzner 2001: PI. 50).
Fig. 29: Khuera, Area H, House H VI as an example of a 8m "allotment house" EJZ 3b (Pfalzner 2001: PI. 55).
The installations on the interior of the Khuera "allotment houses" display a notable homogeneity, independent of plot sizes.149 The main room nearly always contains a gypsum-plastered floor, an altar platform at the rear small side of the room, interpreted as a house altar,150 a hearth in the middle ot in the central axis of the room, and benches along several of the walls. These rooms are defined as "nuclear rooms".151 Besides the standardised main rooms, there are other regular installations in nearly all of the houses, such as tananir (bread ovens), plastered grinding tables installed in specific grinding rooms, or gypsum basins used for a diversity of household activities.
The well-preserved and widely exposed houses in Area B at Melebiya, Niveau 2, can be dated to Period EJZ 3b.152 Lebeau differentiated between two major house plans at Melebiya: small, square houses, and larger central room houses (maisons a espace central)}''3 The central rooms of the latter type often have lime-plastered floors and installations, indicating that they were roofed, but they could alternatively be regarded as courtyards when compared to other EJZ houses in the region.154 Further analysis has led to the conclusion that most of the houses
145 Pfalzner 2001:325-334. lso Pfalzner 2001: 169-176.
151 Pfalzner 2001: 149-153.
152 Lebeau 1993; 1996.
153 Lebeau 1993: 102 f.
154 For a discussion see Pfalzner 2001: 134-138.
159
P. Pfalzner
Fig. 30: Melebiye, Area B, the organisation of plots for "allotment houses", EJZ 3b (Pfalzner 2001, PI. 24).
can be identified as "allotment houses"155 (Fig. 30). They have plots of different size categories: a type with a front width of 9m along the alley (Houses B 1, B 2, B 3, B 4), corresponding to 1.5 nindan; and houses with a width of 6m (Houses B 6, B 8), equal to 1 nindan. The latter are square in shape and conform to Lebeau's type of "square houses". The plot of House B 1, with a size of 200m2 - the largest house in Area B, is not rectangular, but trapezoid in shape. This results from the irregular outline of the streets. Nevertheless, its house front conforms to the 1.5 nindan principle. In addition, there is one house within the domestic quarter of Area B with a different planning principle (House B 7): it cannot be classified as an "allotment house", but rather seems to have been composed of three individual house units.156
The layout of the houses at Melebiya varies, but there are still some general characteristics that can be observed.13' The entrance corridor, equipped with a drain, is mostly located in the middle of the front side. It gives access to a courtyard in the centre of the house (identified by Lebeau as a "central room"). The main room is generally located in the rear of the house. Thus, the "allotment houses" at Melebiya differ in layout considerably from those at Khuera, although they are comparable in plot dimensions (compare Fig. 28). These principles are exemplified by the 1.5 nindan House B 2158 (Fig. 31), which has an entrance corridor (957), a central courtyard (959) and a main room (1110), equipped with benches and a small altar podium. Rooms 1524 and 904 contain grinding tables. Thus, despite the different layout plans, the domestic installations bare close similarities with those at Khuera and Bderi.
In contrast to Melebiya, "allotment houses" are not attested at Bderi during Period EJZ 3b, although they did exist during Period EJZ 3a (see above). Instead, alternative house concepts were introduced at Bderi, replacing the "allotment house" types. This was probably a consequence of an overall destruction that occurred at the site in Level 14, at the end of Period EJZ 3a.
When the settlement at Bderi was rebuilt at the beginning of Period EJZ 3b, in Levels 13 and 12, a house-type was used, which has been labelled "line houses" {"Zeilenhduser"). House XII (Level 12) consists of a row of three rooms, individually accessible159 (Fig. 32). The rooms seem to have been used exclusively for storage. Attached to the structure are two courtyards, each with a large, built storage pit. There are no living rooms associated with this structure, thus it probably belonged to a semi-sedentary household, not present at the site throughout the year. The open area in front of the built storage structures could have been used to erect a tent when the owners were present at the settlement. A rectangular stone pavement in the courtyard might indicate the former site of a tent.
155 Pfalzner 2001: 295-305.
156 Pfalzner 2001: 300 f„ Taf. 27.
157 Ibid.: 303-305.
138 Lebeau 1993: 92 £, PI. 53-60, Tab. 3; Pfalzner 2001: 29" f., Taf. 25. 159 Pfalzner 2001: 291 £, 373, Taf. 17-18.
160
Architecture
Fig. 32: Bderi, Level 12, "line house", House XII, EJZ3b (Pfälzner2001: PI. 17-18).
House XIII (Level 13) is a combination of a "line house" (Rooms CT, CR) used for storage, and an "arch house" (Room DC) used for living160 (Fig. 33). Subrerranean storage pits, spread over a wide area, surround the house. As with the case of House XII, this demonstrates that duringrhese phases, there was sufficient space available in the settlement.
A "dome house" ("Kuppelbaus") is attested in Level 11 at Bderi (House XI)1S1 (Fig. 34). The house includes two units of a double square room, each of the squares once having been covered by a corbelled mud-brick dome. The debris of one of the domes was preserved and serves as a basis for the architectural reconstruction. The dome houses very much resemble modern examples existing nowadays in N and NW Syria. Courtyards surrounded the dome-house: a large, enclosed courtyard, where food processing took place, and two more courtyards, which could have been used as open stables. There are other indications that the household practiced animal husbandry.
Level 11 at Bderi, to the W of the "dome house", also yielded four examples of "double arch houses' (Houses III, V, VI, VII)162 (Fig. 34). These possess two pairs of opposed buttresses inside the rooms. They are distributed at regular intervals, thus sub-dividing the room into three equally sized units. Each pair of buttresses can be reconsrructed as an arch coveting the room. Tie two arches supported a flat roof and served to facilitate roofing
160 Ibid. 292f.,Ta£ 18-19.
":i Ibid.: 129 f., 291,374, Fig. 68,Taf. 15-16.
162 Ibid.: 118-120,285-291, 375 f., Fig. 63,114, Taf. 6, 12-14.
161
162
Architecture
by reducing the unsupported area to be covered. This principle is especially suitable for areas where longer wooden beams for roofing were scarce or expensive.
The "double arch houses" were primarily single-roomed in their original state. However, as can be seen in Levels 10 to 9 at Bderi, the "double arch houses" formed the core of complex, multi-room houses which subsequently developed. This is best exemplified by House III, which began as a "double arch house" in Level 11 but developed into an complex, extended house in Level 9c2, dating within Period EJZ 3b163 (Fig. 35). The former double arch room in Level 11 was still structurally present in Rooms AC/ AD of Level 9c2. In this phase House III was the homestead of a latge, extended household."5'*
One futther house concept attested in Bderi Period EJZ 3b is the "negative house". This does not have specific planning principles, but was inserted into open spaces left between previously existing houses. This kind of house was only used when building ateas within an existing densely populated settlement became scarce due to as increasing population. Therefore, a "negative house" (House II)lfo is only attested in the final EJZ 3b phase at Bderi (Level 8), shortly before the abandonment of the town at the beginning of Period EJZ 4.
In conclusion, Period EJZ 3b at Bderi is characterised by a variety of different house concepts. They are a clear indication of the more or less contemporary co-existence of different models of socio-economic household organisation at this time.
House concepts similar to those found at Bderi were also detected at othet sites in NE Syria, especially the Khabur region, during Period EJZ 3b. At Gudeda, Niveau I, dating to Period EJZ 3b,"56 a house was excavated,167 which could probably be classified as an "allotment house".168 On the secondary hill at Atij, the Canadian mission
163 Ibid.: 285-287, Taf. 15-28. " Ibid.: 384 f., Fig. 117-118.
65 Ibid.: 284 f., Taf. 11-12.
66 According to Quenet, this volume.
67 Fortin 1990b: Fig. 33.
M Pfalzner2001: 312 £, Taf. 34.
163
P. Pfdlzner
uncovered a house (Room 501/502),169 attributable to the "arch house" :;. rc " which rooms were later added, as at Bderi. At Abu Hujaira I, Area B, the excavated remains of Houses L.OS :o L.ll belonging to Phase C-l,171 dating to Period EJZ 3b,172 and interpreted by the excavators as storage buncines. can be regarded as "double arch houses" or even "triple arch houses". Some have additional quarter-arches, perpendicular to the full arches, to further support a horizontal roof consttuction.
Brak has an example of a large residence dating to Period EJZ 3b. It was exposed in Level 5 of Area ER, dating to Brak Period L (= EJZ 3a - 3b).173 Only parts of the house have been uncovered, including storage, workrooms and private areas. Due to the limited nature of the exposure, no functional reconstruction can presently be undertaken of this important residence. Part of a domestic quarter, dating to Period EJZ 3b at Brak was uncovered in Area CH, in the so-called Late ED III destruction level (Level 6).174 A "nuclear room" (Room 66) attached to a courtyard can be identified. In addition, there was a storage room (Room 63), a tannur-room (Room 62), and a grinding room (Room 61).175 If all rooms belonged to the same house, it would also have been a very substantial construction. However due to the incomplete exposure of the house it is not possible to properly assess its place in the typology.
An extended dwelling area dating to Period EJZ 3b was excavated at Beydar in Area B.m It lies to the NW of the central palatial complex and is comprised of small-scale houses. Within this domestic area, the "Tablet Building" was found. This contained an archive of cuneiform tablets, leading to the assumption that this building, perhaps formerly a house, was used for official administrative purposes in Period EJZ 3b.177
5.3.6 Period EJZ 4
Private houses dating to Period EJZ 4 were excavated in the central part of Mozan's upper city, Area C2, where their development can be followed over several phases from Level C 11 to C 8 (all EJZ 4).1/8 The houses are small with thin walls, tiny rooms and do not show a regular, well-planned layout. They were built in a flexible agglomerative manner, according to the changing requirements of the households. A good example is Level 8b (Fig. 36), in which two houses (House X and XI) were constructed next to one another. House X has a possible courtyard (AP) in the centre surrounded on two sides by a series of small rectangular rooms, some of which have a gypsum-plastered floor. House XI, accessible horn an alley though a vety small entrance room (AF), has a similar arrangement with a series of very small rooms surrounding a central space (AD). These houses do not conform to the model of EJZ 3 "allotment houses" indicating that this type was no longer the genetal house model in Period EJZ 4. Instead, rather irregular multi-roomed houses seem to have replaced the allotment house type, at least as far as it can be seen in Mozan's city centre.
The picture at Khuera is different. Khuera's extended EJZ 4 house complexes contrast strongly with the modest houses found at Mozan. Even at Khuera however, allotment houses fell out of use in Period EJZ 4. The main domestic areas (H and K), where allotment houses had prevailed during the previous EJZ 3 period, were completely abandoned during EJZ 4. Domestic occupation only continued in Area E, Levels 2 and 1, both dated to Period TCH I E (= EJZ 4).179 Here, House E I, built in Level 3 (EJZ 3) and most probably an allotment house, was used continuously throughout Level 2 (EJZ 4). Its character was unusual however. The house was large, being comprised of 11 rooms. It was equipped with installations related to metal production and also incorporated a pottery production area.180 These features suggest the existence of an economically active and prosperous household.181 At the same time it maintained some of the features typical of an allotment house including the entrance corridor and the nuclear room with benches and a platform (Fig. 37a).
169 Fortin 1988b.- 165 ff., Fig. 32.
170 Pfalzner 2001: 311 f., Taf. 33.
171 Suleiman & Quenet 2003; 2004; 2006; and in press.
1/2 The date was proposed by Quenet (ARCANE database and Quenet, this volume), but cannot be regarded as certain.
173 Oates & Oates 2001a: 34-36, Fig. 35.
174 Oates 1982: Fig. 2; Oates & Oates 2001a: 28-32, Fig. 28.
175 Pfälzner 2001: 319, Taf. 40.
176 Lebeau 1996b: 3-4, Fig. 2, 5-6; Van der Stede 2007: 10-16, Fig. 3-24, Plan I, II.
177 Lebeau 1996b: 4-11, Fig. 2-3
m Dohmann-Pfälzner & Pfälzner 2000: 219-227; 2001: 115-121; 2002a: 159-162; Bianchi et.al., in press.
179 Both Phases 2 and 1 in Area E are characterised by the typical TCH IE (EJZ 4) pottery repertoire, with the very distinctive type of beaded rim beakers, whilst Phase 3 of Area E is related to Khuera ID (EJZ 3) pottery (Moortgat & Moortgat-Correns 1978: 55-57,65-67, 69, Fig. 25,34-35).
180 See: Pfälzner 2001: 345 f., PI. 65, 67 (= House A in: Moortgat & Moortgat-Correns 1978: 63-68, Plan IV). *s> Pfälzner 2001: 224, 230, 346 f.
164
Architecture
a? i
f i
r
f.... ? p' i v
.  ..' r Ac / ■ ' v
- //J   Ä ,'
''/)--.!/ //:{ßt
■ h Slim.
TALL MOZAN 1998-2001
Bereich C2 Nutzungsphase CSb
■ Begehungsfläche/Fuflboden
I    j Lehmziegel
I    1 Lehminslalialion
I     I Stampfehmfundafnent/Terrassierung
Fr'l Asche
j.'- Kalk.
fi&S% verschroauchte Oberfläche
I     I Grube
EZ3 Störung
O Stein
0 Basalt Keramik
Fig. 36: Mozan, Area C2, Level C 8b, Houses X and XI, EJZ 4 (University of Tubingen Tell Mozan-Project).
In Level 1 (EJZ 4) the so-called "Steinbau V" was erected,182 classified by Orthmann as a palace,183 but recently re-interpreted as an elite residence (House E II).18'' It includes a large, typical main room with a buttressed wall lacing the courtyard (Fig. 37b). The outer form of the house is irregular. This is the result of combining three independent house plots belonging to former allotment houses (Houses Ha, lib, lie) built in Levels 3-2 (EJZ 3-4) in order to erect this extended elite residence during the EJZ 4 period (Fig. 37c).185 Thus, the "Steinbau V" is a good example of the abandonment of former allotment houses for the sake of constructing large house complexes during Period EJZ 4.
Another substantial EJZ 4 house was excavated at Chagar Bazar. It is Building 1 in Area D, Stratum B4, Phases IIa-c,1M dating to the "Post-Akkadian" period.18. The large house is has a rectangular outline and has been reconstructed with arched doorways and two storeys (Fig. 38). The building is solidly consttucted with lm thick walls. On the ground floor is an entrance-way with four small rooms in a square arrangement, featuring domestic installations. This front unit of the house gave access to an exceptionally large reception room (6m x an
182 Moortgat & Moortgat-Cotrens 1975:40-53, Plan IV; 1978: 55-70, Plan IV; Orthmann 1990b: 27, Fig, 20.
183 Orthmann 1990b: 27.
184 Pfalzner 2001: 346-348, PL 66, 68.
185 For the older allotment houses of Levels 3-2 and the "re-organisation" of Area E in Level 1 see: Pfalzner 2001: 347.
186 Tunca & Miftah 2007: 15-35, PI. 2.1, 2.2, 2.8, - 2.10, 2.15, 2.16.
187 The excavators define "Post-Akkadian" as corresponding to very late EJZ 4 and EJZ 5 (Tunca 2007a: 7; MeMahon & Quenet 2007: 69); Li the Arcane database Chagar Bazar Phase Ila-c is dated to EJZ 4b to 4c; this dating is further specified by Quenet, this volume, in ascribing Phase Ilb-c to EJZ 4b''c and Phase Ha to EJZ4c/5; for the problematic meaning of the term „Post-Akkadian" see the comments below, footnote 180.
P. Pßzner
Fig. 37a: Khuera, Area E, Levels 3T, the organisation IEJZ 3) and reorganisation (EJZ 4) of house plots, EJZ 3-4 (Pfalzner 2001, PI. 67).
Fig. 37b: Khuera, Area E, Level 2, House E I, EJZ 4 (Pfalzner2001.pl. 67).
Fig. 37c: Khuera, Area E, Level 1, House EII ("Steinbau V"), EJZ 4 (Pfalzner 2001: PL 68).
estimated 12m) with its heart along the central axis, for which a "communitary function"188 has been suggested. The building should, however, be regarded as an elite-residence. An upper storey has been reconstructed, but only on top of the front unit of the house. This is the only EJZ domestic building, for which two storeys can plausibly be reconstructed.189
In summary, the excavated domestic architecture of Period EJZ 4 includes some very substantial buildings, which can be regarded as elite-residences. Fairly wealthy, extended households must have inhabited these. This stands in contrast to the small, moderate EJZ 4 houses in the central upper city of Mozan, hinting at low-class
;,s Tunca & Miftah 2007: 34.
KS9 On the general problem of reconstructing upper storeys in EJZ houses see: Pfalzner 2001: 130-134.
I j
■ P -
i m s
II
HI i
III I i 41 i
I
166
Architecture
Fig. 38; Chagar Bazar, Area D, Phase Ik-a, Building 1, EJZ 4 (Tunca & Miftah 2007: Fig. 2.15 - 2.16).
inhabitants. Taken together, this evidence suggests the existence in EJZ 4 cities of a complex and heterogeneous social structure.
5.3.7 Period EJZ 5
Mozan and Brak show the clearest settlement continuity during Period EJZ 5. It is unsurprising therefore that the remains of domestic architecture for Period EJZ 5 have been found primarily at these two sites.
The Pusham House in Level C 7 (EJHZ 5) in Area C2 of Mozan is a large rectangulat construction with an obviously regular plan, although the S half of the building was destroyed by erosion (Fig. 39).190 The outer walls are extraordinarily thick and very well built. Several oblong rooms, parallel to the outer walls of the house, are arranged around a huge central courtyard. The latter is paved with baked bricks in a Southern Mesopotamian manner. A pottery drain allowed the removal of rainwater from the courtyard. The rooms were for storage as evidenced by the hundreds of door sealings found in them. Based on this evidence, the building can be interpreted as the house of a long-distance trader. Whether the house also included living areas, or whether it was exclusively used for storage of trade goods- like a Medieval khan - is unclear based on present evidence. Based on the seal impressions and tablet, the inventory of the Pusham House can unequivocally be dated to the 21st century BC and is, thus, a major chronological benchmark for Period EJZ 5.
At Brak domestic structures of Period EJZ 5 came to light in Areas FS and CH. In Level 2 of Area FS (Brak Period N = Ur III)'91 a substantial residential unit was excavated, labelled the "Gtey Libn Building (GLB)" (Fig. 40).192 The house was accessible from the street (32/ 7) through an entrance corridor (11), which gave access to a courtyard (13/ 14). From here, a tannur-room (12) was accessible. Rooms 16,17,18,19 are interpreted as a residential unit within the house, whilst Rooms 2, 3 and 4 are regarded as storerooms. The rooms to the W of them could have been additional working rooms fot domestic activities. S of Room 5 was a small pebble-paved chamber (Room 6) with a drain connected to a hotizontal soakaway. This room can be interpreted with certainty as a well-equipped bathroom. Room 27 contained a grinding installation and three horizontally positioned storage jars, with their upper halves carefully cut away. This led to the assumption that the room functioned as a shop,193 however this structure could also be interpreted as a grinding room. In conclusion, the GLB was a complex and extended residence accommodating various domestic functions. N of it are the remains of other well-built houses
150 Volk 2004: 88-95; Dohmann-Pfalzner & Pfalzner 2001: 121-127, Abb. 18-19; 2002a, 163-168; Pfalzner & Dohmann-Pfalzner, in press.
1"' It should be noted that Area FS Level 2 - based on the few pottery specimens published for this level (in: Uates & Oatcs 2001a) - might be oider than Mozan Level C 7 with the Pusham House. Therefore, FS Level 2 might eventually also date to the end of Period EJZ 4 (EJZ 4c = post-Akkad). However, this is a badly understood period and should not be regarded as a phase of its own, especially because, in historical terms, a "post-Akkad" period before the Ur Ill-period does not exist, as the Ur III dynasty follows very closely onto the late Akkadian imperial period.
192 Oates&Oates 2001a: 63-71, Fig. 79.
153 Ibid. 69.
16"
P. Pfälzner
Fig. 39: Mozan, Area C2, Level C 7, "Pushain House", EJZ 5 (Universityof Tübingen Teil Mozan-Project).
Fig. 40: Brak, Area FS, Level 2, domestic houses, the "Grey Libn Building" EJZ 5 (Oates & Oates 200 la: Fig. 79).
of clearly domestic character, and a second street, demonstrating that the GLB was integrated into a densely populated and vivid domestic quarter in the EJZ 5 city of Nagar.
The succeeding and very last level of the 3'd millennium sequence in Area FS at Brak, Level 1 (Brak Period N = Ur III), again yielded domestic architecture.194 Interestingly, the houses of this level follow a new outline and plan, contrasting with the preceding Level 2. This indicates that there was another shift in urban layout towards the very end of the 3"' millennium. In the central area of the FS exposure are the remains of two houses built next to each other, one of them labelled the "Red Libn Building (RLB)" (Fig. 41). To the F. was a large oval structure, probably a granary connected to one of the two buildings. The houses, although incomplete in plan and so not teconstructed functionally, clearly illustrate the careful planning and high quality nature of
'* Oates & Oates 2001a: 71-73, Fig. 89; see also Pralzner 2001: 318, Ta£ 39.
168
Architecture
Fig. 41: Brak, Area FS, Level 1, domestic houses, the "Red Libn Building", EJZ 5 (Oates & Oates 2001a: Fig. 89).
construction techniques. They are large in size, have a regular, rectangular outline, and broad, solid walls.195 These are traits in common with the contemporary Pusham House at Mozan. One of the rooms (Room 1), interpreted as a bathroom, has a drain and is paved with baked bricks.196 This is the only example of a baked brick floor in EJZ architecture, besides the one in the courtyard of the Pusham House at Mozan. This further supports the notion that a common architectural tradition in domestic structures existed during Period EJZ 5 at Brak and Mozan.
Another settlement area dating to Period EJZ 5 at Brak had already been excavated by Mallowan (and restud-ied by Oates & Oates) in Area CH, Level 1 (Brak Period N = Ur III), to the east of the Naram-Sin palace.19" It contained a dense network of rooms, featuring domestic installations, such as a well, storage facilities, benches, and a house altar.19S Two gold and silver hoards were found within these rooms,199 hinting at the substantial prosperity of this household during Period EJZ 5.200
To this evidence from the main sites of Mozan and Brak we can add the scanty remains of EJZ 5 domestic architecture from Mohammed Diyab. In Chantier 1, Level 9, parts of an EJZ 5 domestic area have been excavated, including an outer space with an oven.201 In Chantier 5a, Level 13b, EJZ 5 domestic structures including a hearth also came to light.202 They indicate that Mohammed Diyab was occupied continuously during Period EJZ 5. The remains of EJZ 5 houses at Chagar Bazar (Area D, Phase Ic-Ib) are very badly preserved, due to erosion on top of
195 Pfalzner 2001: 318.
196 Oates & Oates 200la, 71.
m Oates & Oates 2001a: 15-19, Fig. 14-16. 158 Pfalzner 2001: 319, Taf, 40 199 Mallowan 1947: 71. ™ Pfalzner 2001: 319.
201 Nicolie 2006: 35, Fig. 2-2.
202 Nicolie 2006: 62, Fig. 4-4.
169
P. Pfihner
the mound. Two rooms and tananir could be identified, at least demonstrating that there was domestic occupation at the site during Period FJZ 5.2W
In summary, the few substantial examples of FJZ 5 dome-nc architecture at Mozan and Brak represent large, well-built and prosperous houses. This contradicts the general assumption of impoverished urban living conditions during this last phase of the 3"1 millennium. It rather hints at the existence of wealthy households involved in supra-regional activities during this time.204
5.4. Palaces
In line with the genera! development of political structures in the EJZ period, no palaces are attested in periods EJZ 0, FJZ 1 or for most of EJZ 2. The earliest palatial building attested in the region dates to the end of Period EJZ 2.
5-4,1 Period EJZ 2
The oldest palatial building dating to the EJZ period is the so-called "Leilan 11 Id Palace", situated on the city's Acropolis (Fig. 42). The part of the building excavated corresponds to Levels 17-16, dating to the end or Period EJZ 2.m The walls are up to lm wide and it has two groups of storage rooms, one comprising nine rooms, and the other three. More than SO ED Ilia style seal impressions were found. These were associated with the building itself and an ash layer on top of it. Most of the storage rooms are very small, the largest one (Room 1) measuring 2.6 x 2,3m. In view of rhis and the tact that sealings arc also widely present in EJZ period domestic contexts,20" it is possible that this complex of rooms was not actually palatial in nature. It could equally have been the storage area of a non-palatial communal, religious or private context. Thus, EJZ 2 palatial architecture remains obscure.
5.4.2 Period EJZ 3
Aside from the ambiguous Leilan (Period IHd) evidence, JZ palatial architecture is first fully attested in Period EJZ 3. Palace F at Khuera has been documented for three phases: Building Levels 2b and 2a date to period "Khuera ID late", whilst Building Level 3 dates to period "Khuera ID early".20" Both of these local phases correlate with Period EJZ 3b. The largest exposures pertain to Level 2, whilst Level 3 possibly represents an earlier phase of the building, only reached by soundings in a few places.203 The palace is not centrally located, but is positioned with its back attached to the inner city wall. In the main building phase, Level 2b, the palace is strictly rectangular in outline, although it is terraced (Fig. 43). The central residential area is located in the upper terrace
Fig. 42: Leilan, Acropolis, Levels 17-16, so-called "Leilan IHd palace", EJZ 2 (Calderone & Weiss 2003: Fig. 3).
203 Tunca&Mifcali 200": 13 f., Fig. 2.18a.
See: Pfalzner 2010: 4-10; in press. 205 Weiss 1990b: 209-213; Calderone & WVi-s 2003: 194 197. m See: Pfalzner 2001: 232-239.
207 Orthmann. & Pruss 1995; 121 f., 139 f.
208 Ibid. 122-124, PI. 17-20, Beilage 15: Pruss 1998a; 2       1     b; Orthmann 1990b: 25. Fig. 19.
Architecture
of the palace. To the N and E are two lower terraces, each of which is enclosed by a long corridor. The main access route to the palace led to the lower eastern terrace, with the main courtyard at its centre (3). The courtyard gave access through a recessed door ro a large, hall (2) to the N. This was the largest room in the palace (10 x 12m), and may have held official functions such as assemblies and receptions. To the E of the recessed door of hall (3), an.: attached to its outer wall, is a podium of large stone slabs, sitting within a courtvard.209 This is characteristic for EJZ official buildings (see below), thus the lower eastern terrace can be defined as the official wing of the palace.
The upper terrace of the palace was accessible from the central courtyard via an eye-catching semicircular stone-paved staircase. The staircase led to a three-winged room unit that formed the architectural core of the palace, serving both private and official functions. The central room or inner courtyard (8) gave access to two similar rooms: a large N room (6) with an altar platform on its E wall and a grinding-table on the N wall;210 and a slightly larger S room (12) with a low clay podium on the E wall.211 The interior arrangements of both rooms are comparable to the core rooms of private houses, suggesting that this was the residential unit of the palace. This assumption is supported by the existence of a small room unit comprising a toilet (22) and bathroom (21) located to the S or Room (12) from which it was directly accessible.
In the Wpart of the upper terrace were two irregularly arranged courtyards (16 and 44) forming the core of a series of rooms related to storage (room 54) and food preparation (rooms 47, 31). Thus, the W part of the upper terrace can be defined as the service wing of the palace. The N lower terrace of the palace seems to have been used for official purposes (rooms 53, 41). In summary, Palace F at Khuera was clearly organised according to various functions: two lower terraces with official roles surrounded the upper terrace housing the residential and service units.
The second fully-exposed EJZ 3 palace is the so called "official block" in Area F at Beydar. In contrast to the palace of Khuera, it is located in the city-centre. Hie palace is surrounded by temples and other public buildings and thus forms part of an extended "palace temple complex" in the middle of the upper town. Four phases (1-2-3a-3b) of the palace can be distinguished ail belonging to Period EJZ 3b. In contrast to the palace at Khuera, the outline of the Beydar palace is irregularly trapezoidal. The original plan of the building and the principles behind its design are clear in Phase i (Fig. 44).212 A vestibule gave access to a large courtyard (6233), which dominated
209 Orthmann & Pruss 1995: 122, PI. ISa.
210 Mobrtgat-Correns 1988: 21-28, Fig. 4-6 (Room 6 had in earlier excavations erroneously been regarded as a temple, the so-called ,,\\"est-Temple"s). For a general re-interpretation of the so-called "libation installations" of Khuera - as the one on the N wall of Room 6 ("West-Temple") - see: Pfalzner 2001: 139-146.
211 Orthmann & Pruss 1995: 123, PI. 19. This room has been tentatively labelled "throne-room' by the excavators.
212 \ an der Stede 2003: 29-36; l.ebeau 2003b: 21 I
P. Puilzner
RECONSTRUCTION
Fig. 44: Beydar, palace in Area F, The "official block" Phase 1, EJZ 3b (Bretschneider 2003: Plan 2).
the E half or the palace. The courtyard was bordered to the E and N by open halls, each with a large pillar on rhe nearest side. This architecture can be classified as "iwan-type".213 Atypical buttressed wall on the W side separates the courtyard from the largest room (6984) of the palace, interpreted as the central reception room. This central unit comprising a main courtyard and reception room is closely comparable to Courtyard 3 and Hall 2 on the lower E terrace of the Khuera palace, for which a similar function has been proposed (see above).
A third major room, situated along the palace's main axis with the courtyard and reception room, is the so-called triangular room (6732). A large plastered podium situated at the rear leads to its interpretation as a throne room.21'' Alternatively, it could have had a more private character and comprised part of the residential unit; the rooms in the upper storey were directly accessible via an adjoining staircase (6775). In this scenario the room would be comparable to the residential unit on the upper terrace of Khuera's Palace F, whose two large rooms (6 and 12) also contained platforms.
Hie central unit of the palace was bordered on both the S and N side by a row of service rooms. Two staircases indicate the existence of an upper storey, where additional private, residential rooms may have been located, comparable to those on the upper terrace of Khuera's Palace F. This would explain the small size of the Beydar palace compared to Khuera's Palace F, and was obviously a necessary consequence of the restricted available space due to pre-existing streets in the Beydar city centre.213 Thus, although the architectural layout of both palaces is, at first sight, strongly different, this can be attributed to practical reasons related to spatial limitation. The functional division of the Khuera and Beydar palaces is closely comparable.
During later phases (2, 3a, 3b) architectural modifications of the palace plan were carried out (Fig. 45). In Phase 2 an additional row of service rooms was added to the N and the E side of the palace.210 The most significant modification was made in Phase 3 when the triangular room was replaced with a larger rectangular room (6518) and connecting toilet (6610).21 This new layout - interpreted bv the excavators as a newly created residential unit218 - would support the idea that this room also served a residential purpose in Phases 1 and 2 (see above). The functional organisation of the Beydar palace remained unaltered through these architectural phases.
In close proximity to the Palace in Field F there is another palace on Beydar's Acropolis in Field P, labelled the Eastern Palace (Fig. 46).z15 It belongs to Level 5c of Field P and dates to Period EJZ 3b. It contains a iarge
213 Aft Jman" (a term borrowed from Islamic Architecture) is a room with a very wide opening or frontal wall opening towards a courtyard or a large hall, as it is typical for Arabic houses or mosques.
2H Lebeau 2003b: 22.
215 Lebeau 2003b: 22.
216 Debruyne 2003: 50-54; Lebeau 2003b: 23.
217 Bretschneider 2003: 86-91: Lebeau 2003b: 24-26. :iS Lebeau 2003b: 24.
m Pruss 2006a: 17-18, Fig. p. 21; Pruss 2006b. PL 3.
Fig. 46: Beydar, Eastern Palace. Area P, Level 5c, EJZ 3b (with kind permission of A. Pruss).