Husn Salut and the Iron Age of South East Arabia: Excavations of the Italian Mission to Oman 2004-2014 [1 ed.] 9788891316424, 9788891316363

The ancient oasis of Salut, located on the deserts fringes in the heart of the Oman Peninsula, not far from the modern c

177 23 18MB

English Pages 425 Year 2018

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Husn Salut and the Iron Age of South East Arabia: Excavations of the Italian Mission to Oman 2004-2014 [1 ed.]
 9788891316424, 9788891316363

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

A rabia A ntica 15 Archaeological Studies

collana diretta da Alessandra Avanzini

A rabia A ntica

1 - A. Avanzini (ed.), Khor Rori Report 1, 2002, pp. 388. 2 - A. Avanzini, Corpus of South Arabian Inscriptions I - III. Qatabanic, Marginal Qatabanic, Awsanite Inscriptions, 2004, pp. 606. 3 - A.V. Sedov, Temples of Ancient Ḥaḍramawt, 2005, pp. 328. 4 - D. Mascitelli, L’arabo in epoca preislamica: formazione di una lingua, 2006, pp. 316 + 19 figure fuori testo. 5 - A. Avanzini (ed.), A port in Arabia between Rome and the Indian Ocean (3rd C. BC - 5th C. AD). Khor Rori Report 2, 2008, pp. 742 + 6 tavole fuori testo. 6 - A. Avanzini (ed.), Eastern Arabia in the first millennium BC, 2010, pp. 250. 7 - J.-F. Breton (ed.), Le sanctuaire de ʿAthtar dhû-Riṣâf d’as-Sawdâʾ, 2011, pp. 224. 8 - A. Prioletta, Inscriptions from the southern highlands of Yemen. The epigraphic collections of the museums of Baynūn and Dhamār, 2013, pp. 408. 9 - M. Mouton & J. Schiettecatte, In the desert margins. The settlement process in ancient South and East Arabia, 2014, pp. 336. 10 - A. Avanzini, By land and by sea. A history of South Arabia before Islam recounted from the inscriptions, 2016, pp. 376 + 3 carte fuori testo. 11 - A. Lombardi, South Arabian funerary stelae from the British Museum collection, 2016, pp. 204. 12 - A. Pavan, A cosmopolitan city on the Arabian coast. The imported and local pottery from Khor Rori, 2016, pp. 398. 13 - D. Amaldi, The origins of the history of Oman. The Kitāb al-ansāb by al-ʿAwtabī, 2017, pp. 264. 14 - D. Mascitelli, I testamenti dei re e dei principi della stirpe di Qaḥṭān ibn Hūd, 2018, pp. 256. 15 - A. Avanzini, M. Degli Esposti (eds), Husn Salut and the Iron Age of South East Arabia. Excavations of the Italian Mission to Oman 2004-2014, 2018, pp. 424.

Husn Salut and the Iron Age of South East Arabia Excavations of the Italian Mission to Oman 2004-2014

edited by

alessandra avanzini, michele degli esposti

«L’Erma» di Bretschneider

Alessandra Avanzini, Michele Degli Esposti (eds) Husn Salut and the Iron Age of South East Arabia. Excavations of the Italian Mission to Oman 2004-2014 Redazione e impaginazione a cura di Alessandra Lombardi Revisione del testo inglese a cura di Geoffrey Phillips La trascrizione dei toponimi arabi segue un criterio semplificato che ignora i segni diacritici e le vocali lunghe Crediti fotografici: © IMTO e autori Disegni: Sergio Martelli e autori eccetto pls 6/4, 9; 10/2; 83/15: Alain Rosa, e pl. 6/8: Simona Spano © Copyright 2018 «L’Erma» di Bretschneider Via Cassiodoro, 11 - 00193 Roma www.lerma.it Tutti i diritti riservati. È vietata la riproduzione di testi ed illustrazioni senza il permesso scritto dell’Editore

Avanzini Alessandra, Degli Esposti Michele (a cura di) Husn Salut and the Iron Age of South East Arabia. Excavations of the Italian Mission to Oman 2004-2014 - Roma : «L’erma» di bretschneider, 2018. - 424 p. : ill ; 28 cm. (Arabia Antica ; 15) (Archaeological studies) ISBN: 978-88-913-1636-3 (brossura) ISBN: 978-88-913-1642-4 (pdf) CDD 930.1074439 1. Archeologia - Arabia

Index

A lessandra A vanzini , M ichele D egli E sposti Introduction Acknowledgements

9 17

M ichele D egli E sposti , M auro C remaschi , A lessandro P erego , A ndrea Z erboni 1. Geographical and environmental setting Geo-environmental zones of the Oman peninsula Regional climate and rainfall The plain of Salut: potentials for human occupation The geoarchaeology of the Salut plain Irrigation systems at Salut The ancient oasis of Salut. The age of the irrigation system and its relation to climatic change

19 19 22 24 25 28 29

M ichele D egli E sposti , C hiara C ondoluci 2. Husn Salut in the Iron Age Architectural and stratigraphic phasing Husn Salut I Husn Salut II Husn Salut III

The Iron Age material culture: general remarks Pottery and other finds from Building I (HSII and HSIII phases) PLATES C hiara C ondoluci , M ichele D egli E sposti , C arl P hillips 3. A complete stratigraphic sequence for the Early Iron Age of Central Oman. Excavations in the Basement and in the Burnt Building The excavation Foundation substructures

Original layout Floor raising and a large fire The basement’s obliteration and late modificatios to the Burnt Building’s layout

33 33 34 43 47 49 59 61

93 93 94 96 99 101

6

husn salut and the iron age of south east arabia

Pottery and other objects from the Basement and the Burnt Building Foundation substructures Original layout Floor raising and a large fire The basement’s obliteration Burnt Building’s re-arrangement Later modifications to the Burnt Building’s layout Other finds

PLATES

102 102 103 103 104 106 107 107 109

E nrica T agliamonte , M ichele D egli E sposti , C arl P hillips 4. Other areas of excavation The central part of the site (Area 4) The outer walls and the main tower PLATES

181 181 187 193

C hiara C ondoluci 5. The Iron Age architecture of Husn Salut A sophisticated building technique: compartments foundations The compartment technique: carrying on a local tradition

Early phase foundation deposits and associated materials (Husn Salut I) Compartment substructures (and superstructures) in Husn Salut II phase Late foundation deposits and associated materials (Husn Salut II) Beyond materials and techniques. General remarks PLATES

201 202 208 210 225 226 233 239

E nrica T agliamonte 6. Stone vessels Bronze Age stone vessels Iron Age stone vessels Lathe-turned stone vessels: post-Early Iron Age occupation or early lathe introduction? Catalogue PLATES

279 279 281 286 287 295

M arzia S asso 7. Metal objects Ceremonial or cultic objects Weapons and tools Personal adornments and decorations Raw material: ingots Iron objects Gold ornaments Catalogue PLATES

301 301 306 310 311 312 313 314 323

index

7

L aura S trolin (with a contribution by J acqueline S tuder ) 8. Faunal remains from Iron Age Husn Salut: an archaeozoological study Bone sample Methods Fauna at Husn Salut during the Iron Age Skeletal pathologies Final remarks TABLES

331 331 332 333 350 352 353

C hiara C ondoluci , M ichele D egli E sposti 9. The function of the site

363

M ichele D egli E sposti , C hiara C ondoluci , C arl P hillips , E nrica T agliamonte , M arzia S asso 10. The Early Iron Age chronology of South East Arabia: a reassessment on the basis of Husn Salut excavations Chronological phasing and terminology for South East Arabian Iron Age Phases and absolute chronology at Husn Salut A different chronological model for the Early Iron Age in central Oman Preliminary remarks on the end of the Iron Age occupation at Salut

371 371 373 377 380

M ichele D egli E sposti , C hiara C ondoluci , C arl P hillips 11. Salut during the Islamic period PLATES

383 397

Bibliography

401

Introduction alessandra avanzini , michele degli esposti

In late March 2004 the bases for a demanding but greatly rewarding and stimulating enterprise were set, when the original members of the Italian Mission to Oman (hereinafter, IMTO) carried out a preliminary survey in the close surroundings of the site of Salut, near Bisya in central Oman (fig. 1).1 Although it had entered the archaeological literature not less than 40 years earlier, from this moment on the role of Salut in the understanding of South East Arabia’s2 ancient history has been radically transformed, thanks to the program of extensive excavation that has been since conducted by the IMTO. In fact, it seems that, following a preliminary reconnaissance by J.C. Wilkinson in February 1973,3 the members of the Harvard Archaeological Survey in the Sultanate of Oman, who recorded it as site BB-15, first surveyed the site of Salut in winter 1973. Based on the surface material, two major periods of occupation were determined: “one in the latter half of the first millennium B.C. and one in the 13th-14th centuries A.D.”.4 Later, D.S. Whitcomb published the surface collection of medieval pottery.5 In winter 1974-1975 the site was also visited by the archaeologists of the British Archaeological Expedition and was listed in their ‘Gazetteer’ as site 38.6 However, none of these surveys included excavation, and the site presented itself as archaeologically pristine to the IMTO’s trowels and shovels. The excavation of what is now referred to as ‘Husn Salut’7 was to last for more than ten years, finally revealing a site that has undoubtedly to be listed as one of the most impressive archaeological monuments of South East Arabia. However, new tasks await, and the project is now finding its natural continuation in the excavation of the associated, substantial Iron Age settlement of ‘Qaryat Salut’. The archaeological complex of Salut comprises sites and isolated remains spanning a wide chronology, from the Early Bronze Age to the Late Islamic period,8 and occupies quite a large portion of the wide valley that flanks Wadi

1 Over more than twenty years of activity, the IMTO has been directed by its founder, Prof. A. Avanzini (Dep. of Civilizations and Forms of Knowledge, University of Pisa). During the 2004 and 2005 seasons, fieldwork at Husn Salut was directed by Prof. A. V. Sedov (Museum of Oriental Arts, Moscow). Carl Phillips was appointed as field director in 2006, until 2014. Two of the present authors (MDE and ET) supervised the final activities at the site during the late 2014 and 2015 campaigns, mainly aimed at assisting restoration works. The whole IMTO project at Salut was carried out thanks to the invaluable support of the Office of His Excellency the Advisor to His Majesty the Sultan for Cultural Affairs, Muscat. 2 South East Arabia is used throughout to indicate the territory of modern United Arab Emirates and northern Oman, from the Musandam peninsula to Masira island. The expression ‘Oman peninsula’ is synonimous. 3 Wilkinson 1977: 129. See also Costa 1988: 15. 4 Humphries 1974: 51. 5 Whitcomb 1975. 6 De Cardi, Collier and Doe 1976: 145, 164. 7 The site became initially known simply as Salut in the archaeological literature, a name that the IMTO has kept using until late 2015, when a new programme of excavation started revealing the associated settlement. This settlement has been distinguished as Qaryat Salut in order to avoid confusion with Husn Salut during the recording of contexts and finds, as well as mirroring the association of a ‘village’ with its ‘fortress’ that it seems to bear witness to. 8 Not to mention the presence of earlier, although more ephemeral, evidence for a Neolithic occupation of the plain (Cremaschi 2007).

10

alessandra avanzini , michele degli esposti

F igure 1 – The location of Salut in central Oman and the main Iron Age sites mentioned in the text. Squares indicate modern towns.

introduction

11

F igure 2 – General view of Husn Salut in 2004, at the beginning of IMTO’s work.

Saifam, east of its course and near the point where it joins the larger Wadi Bahla, roughly 2.5 km to the northeast of the city centre of Bisya. The closest main centre is Bahla, some 30 km to the north, while the small village of ad-Dhabi stands two kilometres from the site as the crow flies, along the Wadi Bahla’s course. In the early 1970s,9 as well as at the time of the IMTO’s first visit to the area, the visual focus of such a dense archaeological landscape were the massive ruins of what appeared to be a fortress set on a rocky outcrop standing rather isolated away from the nearby hills, thus providing a privileged position for overlooking the surrounding plain (‘BB15’ itself, fig. 2). Remains of structures such as stone defence walls or mud-brick features were clearly visible on the top and slopes of the hill, and it came almost naturally that the IMTO activity concentrated exclusively on this fortress during the first years of field work. Such a choice was also suggested by the overall situation of South East Arabian archaeology at the time of the project’s outset. In fact, the Iron Age period had not been studied in central Oman as extensively as the earlier prehistoric and Bronze Age periods; a situation that has not substantially changed today. It is difficult to think of any Iron Age settlements in Oman that have been studied in such detail and extension. The fortress at Lizq for example, a site that indeed shows many similar traits to Husn Salut and is possibly even larger than it,10 was only partially investigated. Though the distinctive, often painted, pottery discovered at Lizq has represented a key assemblage for the South East Arabian Iron Age, only a very specific part of the site was investigated, and many aspects of its layout remain unveiled.11 The excavation of Iron Age houses at Maisar (now Moyassar) and Raqi is also owed to the great efforts

Whitcomb 1975: pl. IB provides a general view of the site, almost unchanged until the IMTO’s arrival. Not including the surrounding settlement of Qaryat Salut, now under excavation; see Kroll 2013: 169. 11 Most recently, Kroll 2013. 9 10

12

alessandra avanzini , michele degli esposti

of the German Archaeological Oman Expedition that has been working in Oman at length.12 To these can be added the partial excavation of an Iron Age site at Manal, in the Wadi Samail area, along the modern Nizwa-Muscat road.13 Other less extended excavations have been conducted, and several settlements have been reported, but only as a result of surveys (among them, one can recall the remarkable extension of al-Zahra 2, near Sohar).14 In many instances the Iron Age can be represented only by burials. These can be chronologically restricted to the period itself, such as those excavated in Wadi Baushar,15 or, more often, represent the re-use of earlier tombs, with consequent mixing of the burial goods (as for example the recently fully published al-Wasit and Nizwa graves).16 A greater wealth of data on Iron Age settlements was available from the United Arab Emirates, specifically from the pioneering research at Rumeila and Hili17 and later excavations at Muweila.18 A remarkably comprehensive program of archaeological investigation was also carried out between 2000 and 2004 in the Bithna area (Emirate of Fujaira) by the French Archaeological Mission to the United Arab Emirates, where several sites testify to a complex organization of human occupation. Although several papers have appeared over the years,19 the final publication of this project appeared in 2013.20 A similar program was developed by the same team in the Masafi area, at the border between Fujaira and Ras al-Khaima Emirates,21 and is proceeding towards its final publication. Also in this case the results outlined the modality in the occupation of a well-watered area nested among small hills; the latter were preferentially, but not exclusively, chosen for the construction of buildings, in order to preserve cultivable low lands. Dated to the central and final part of the Iron Age period, the site of al-Madam (Emirate of Sharja) also provided important information on the mud-brick architecture of the period and on the subsistence of its population.22 Unfortunately only partially published are the results of a comprehensive research program carried out along the Wadi al-Qawr, in the Emirate of Ras al-Khaima, which highlighted a complex pattern of sites, including Iron Age sites that cover the socalled Iron Age II and Iron Age III periods.23 Other sites were excavated to a lower extent or only identified in survey reports.24 However, none of these sites, despite an extension that is in cases remarkable (Muweila, Hili/al-ʿAin), possesses the prominence of Husn Salut’s architecture, not even other hill ‘forts’ located, for example, at Bithna (Bithna 24)25 and Husn Madhab,26 close to Fujaira. Moreover, during all these years there has been general acceptance of the three-fold chronological subdivision for the Iron Age period of the region proposed by P. Magee in 1996, mainly stemming from the results of Tell Abraq’s excavation,27 and none of the mentioned sites apparently provided evidence to revise it, which has probably largely to do with this misbalance between the data available from the UAE and those collected in central and eastern Oman, the two regions actually displaying much more different chronological trajectories than commonly thought. There-

Weisgerber 1981; 1982; Weisgerber and Yule 1999; Yule 2001; Yule and Weisgerber 1988; 2001. Elmahi and Ibrahim 2003. 14 Costa and Wilkinson 1987: 99-103. An extensive survey has also been conducted in more recent years, among which the work carried out in the frame of the ‘Transformation Processes in Oasis Settlements in Oman’ project deserves mention (Häser 2010, with bibliography). 15 Costa et al. 1996: 28-41, 49-71. 16 Yule and Weisgerber 2015. 17 E.g. Boucharlat and Lombard 1985; Ur-Rahman 1979; al-Tikriti and Haddou 2001. It must be noted that the different sites within the al-ʿAin oasis should conveniently be interpreted as parts of one and the same ‘polity’ (Magee 2014: 219), each distinguished by its own nature (productive, defensive/control, domestic). 18 Magee 1996a; 2001; 2007a; Magee et al. 2002. 19 Benoist et al. 2004; Benoist 2005; 2007. 20 Benoist 2013a. 21 Benoist 2010a; Benoist et al. 2012a; Benoit et al. 2012b; Benoist 2013b. 22 Cordoba 2003; 2010. 23 Phillips 1987; 1997; 1998a; 2001a. 24 E.g. the Awhala fort in Fujaira (Potts et al. 1996; Petrie 1998), and the enclosure at ed-Dur North in Umm al-Qaiwain (Phillips 1998b; 2001b). 25 Benoist et al. 2013. 26 Courboud et al. 1994; Corboud 2001. It is not the goal here to provide an exhaustive list of Iron Age sites in the Oman Peninsula, but only to sketch the background against which the IMTO’s initiative at Salut started. For a remarkably comprehensive list see Kroll 2013: 215-220. 27 Magee 1996b. 12

13

introduction

13

fore, the relevance of Husn Salut’s results lies not only in the outstanding architecture they revealed, but also in the data they provided in order to outline a different chronological development for central Oman throughout the second half of the second and the first millennium BC, as will be discussed in Chapter 10. As excavations at Husn Salut progressed, it became more and more evident that the site was far from a usual settlement, albeit strongly fortified by a massive enclosure.28 Only two actual buildings were outlined, the so-called Basement and the Burnt Building (see Chapters 2 and 3), appearing totally inconsistent with the necessity of housing at least as large a population as the one that would have necessarily been involved in the site’s construction. From a complementary perspective of the excavation, it thus became of great importance, after the first seasons of field work, to locate the settlement or settlements that likely gravitated around such a prominent site as Husn Salut. In 2009, a new program of surveys became included in the ever enlarging IMTO’s scope, finally attaining success as a series of smaller but anyhow not negligible areas of Iron Age settlement were located (fig. 3). While the results of these surveys were detailed elsewhere,29 worth underlining here is the fact that all these ‘satellite’ sites were placed along the lower slopes of other hills and rocky outcrops, thus suggesting that the more easily irrigable plain was saved for agricultural exploitation.

F igure 3 – Bronze Age and Iron Age occupation patterns in the Salut plain and adjacent wadi Bahla plain. Squares indicate Bronze Age remains, rounded shapes Iron Age sites (the larger, the more conspicuous; the line around Husn Salut shows possible extension of the associated settlement, i.e. Qaryat Salut). Site names follow IMTO survey coding (see Condoluci, Degli Esposti and Phillips 2014; also Phillips, Condoluci and Degli Esposti and 2011). SS11 comprise an area of field-levelling mounds (Arabic nadud), rich in Iron Age pottery.

See Chapter 9 for a discussion on the possible interpretation of the site’s function. Phillips, Condoluci and Degli Esposti 2010; 2012; Degli Esposti and Phillips 2012; Condoluci, Degli Esposti and Phillips 2014.

28 29

14

alessandra avanzini , michele degli esposti

However, the location of these sites, none of which was excavated, still left the query about the location of the larger Iron Age population’s houses largely unanswered. Indeed, the IMTO team discovered the solution was literally under their feet: as mentioned before, the recently started and ongoing excavation of the remaining part of the Husn Salut hill and adjacent plain is in fact revealing the large Iron Age settlement of Qaryat Salut, comprising hillside terraces and buildings (fig. 4).30 Its excavation, considered together with the layout and nature of the wider human-modified landscape, will hopefully provide for a more comprehensive picture of Iron Age society and contribute important information about how the settlements were sustained, their irrigation and husbandry practices, and their contacts with the wider world. The important place in South East Arabian ancient history that Salut is achieving through the works of the IMTO’s archaeologists is actually echoed in a number of legends and allegedly historiographic accounts, the most relevant of which were collected by al-ʿAwtabī in his Kitāb al-ansāb (the book of genealogies), of which translations and more detailed discussions have been published elsewhere.31 Strongly suggestive is the link established between the arrival of Sulaymān bin Dawūd in Oman, namely at Salut, and the excavation of 10,000 canals over 10 days by the demons that he controlled. This episode, said to have taken place at a time when in Oman “the people were nomads”,32 could be read as mirroring the fundamental importance of the introduction of complex irrigation systems in the region. Significantly, its association with King Sulaymān would somehow fit the Early Iron Age chronology for the adoption of the falaj technique, now widely accepted on the basis of a number of archaeological validations, including the indirect confirmation provided by the reconstructed demographic growth during the first half of the first millennium BC.33 In the Salut plain, a sophisticated system of water management surely existed as early as the second half of the third millennium BC, when huge ditches were excavated around the typical Early Bronze Age ‘towers’ that are located at a short distance from the Iron Age site, also associated with other substantial channels.34 So far, direct evidence for Iron Age hydraulic features has not been discovered at Salut, although there are clues pointing to their existence. In particular, recent measures of U/Th rates inside calcareous tufa in form of lateral flowstones and stalactites that were cut and removed during the Islamic phase of renewal of falaj Shaww, yielded a date which indicates it was originally built before the 4th century BC.35 If this legendary episode reflects a key innovation for settlement sustainability in the region, another event said to have taken place in the Salut plain laid the foundation for the development of South East Arabia culture in historical times: the battle of Salut, opening the way for the arabization of Oman. This semi-legendary clash saw Mālik bin Fahm, the first of the Azd to move from Yemen to Oman as a consequence of the Marib dam disaster, defeating the Persians that were occupying the coast and who were based at Sohar, under the command of “al-Marzubān, governor of the king of Persians”.36 It has to be emphasised that the Persian king that this governor was serving is identified as “Dārā b. Dārā b. Bahman b. Isfandiyār”,37 a name recognized as “a mythical construct which combines the names of Bahman, son of Isfandiar, one of the early Persian mythical figures described by Firdowsi, and Dara, as Darius III, Alexander the Great’s unsuccessful opponent”.38 This observation is of great importance in two instances, always bearing in mind that much caution has to be exercised in dealing with these sources.39 First, this would date the leg-

Avanzini and Tagliamonte in press. See most recently Amaldi 2017. 32 Amaldi 2017: 199. 33 E.g. Magee 2014: 214-222, with references; see also the critical review of available data in Charbonnier 2015. 34 E.g. Degli Esposti 2016. 35 Cremaschi et al. forthcoming. Falaj Shaww is an abandoned underground channel running close to the abandoned Islamic village of the same name, in the proximity of Husn Salut. See also Chapter 1. 36 Amaldi 2017: 177. 37 Amaldi 2017: 122. 38 Potts 1990a: 399. 39 The inconsistency of the link with the Marib dam for example was pointed out by von Wissman, recalled by Wilkinson (1977: 126). Wilkinson himself noted that in the narrative about Mālik bin Fahm “nearly a thousand years of history have been compressed” (Wilkinson 1977: 128). 30 31

introduction

15

F igure 4 – A general view of the new IMTO’s excavation at Qaryat Salut (April 2017).

endary Azd migration to Oman to the 4th century BC at the latest, much earlier than usually hypothesized.40 Second, such a date would find a rough correspondence in the chronological range of some ceramic affinities evident at Husn Salut, and in the presence of South Arabian items, thus contributing to the discussion on the final date of Iron Age Salut (see Chapter 10). Finally, it needs to be remembered that according to the Neo-Assyrian and Achaemenid written sources the Oman Peninsula was known as “the land of Qade (Kade)”,41 and that the king of this land, mentioned as a tribute-bearer to Assurbanipal around 640 BC, is said to be living in his capital, “the City of Iskie”.42 It is tempting to see in this toponym a reference to the town of Izki, not far from Salut in central Oman, an identity that has been proposed with some agreement43 and that finds a peculiar echo in the local tradition that considers Izki to be the oldest town in Oman.44 Archaeological surveys and limited test excavations have confirmed the presence of pre-Islamic remains, also including evidence for an Iron Age occupation that extends to the final part of the period.45 However, what is lacking is any substantial remains that could be, from an archaeological point of view, identified as the capital mentioned in Assyrian sources. The plausible chronological correspondence between this mention of an Omani ‘capital’ in a NeoAssyrian text, and the massive undertaking of Husn Salut’s re-arrangement and extension in the site’s phase III (see Chapters 2 and 10), gives way to some intriguing, although not verifiable, speculation, if one considers that toponyms may sometimes remain, but their geographic location may vary, or that similar polities as the one to be hypothetically

See the discussion in Potts 1990a: 398-400, specifically notes 234-235. Potts 1985a; see also Lombardi 2015: 29-30. 42 As reported in the so-called Ishtar slab, see translation in Lombardi 2015: 29. 43 Potts 1985b. 44 Wilkinson 1977: 130, 208. 45 Costa 1988; Schreiber 2004; Häser 2010; Yule 2015. 40 41

16

alessandra avanzini , michele degli esposti

identified with Izki could have existed throughout the region. What seems to be sure in fact, for the time being, is that no other Iron Age site in the Oman peninsula can become a stronger candidate than Salut for the role of Qade’s capital. The excavations at Husn Salut have already made a significant contribution to aspects of the Iron Age in the Sultanate of Oman, and in particular to a reassessment of the internal chronology of this long period, as discussed in Chapter 10. Further work at Qaryat Salut will no doubt expand upon this, and add remarkably more details also along different lines of research. Although the undertaking of the IMTO has been extremely demanding, in fact, it is clear from the initial results of its new project that much more work awaits, and this will make an important contribution to the regional, South East Arabian picture of the Iron Age period – a period that witnessed significant changes throughout the whole of Arabia. In this report the main results of Husn Salut’s excavation will be presented (Chapter 2), with a particular focus on the fundamental stratigraphic sequence revealed within the earliest edifices that were erected on the hill, namely the so-called Basement and the Burnt Building (Chapter 3). This sequence is of the uttermost importance in that it covers a remarkably long time-span, allowing the discussion of Early Iron Age chronology in central Oman and its correlation with the chronological model so far widely accepted for the whole Oman peninsula, but which is actually based on a bulk of data from the territory of modern day United Arab Emirates.46 A first outline of the critical implications of Husn Salut’s excavation for this issue was provided by C. Phillips,47 at the time when a general update on the site’s excavation was also published.48 However, the relevance of the matter requires a more exhaustive publication of the robust data set that served as a basis for the proposed chronology, including the comprehensive illustration of pottery and other finds, and additional radiocarbon dates that complement those already published. The latest field work in other parts of the site has yielded important results that enlighten its architectural phasing, partially modifying previous reconstructions and providing new 14C dates that can be considered together with those from the Basement and the Burnt Building. Therefore, in addition to the complete publication of the aforementioned contexts, a general outline will be presented of all the other areas of excavation, and the associated pottery and relevant objects will also be discussed at a more general level, while the most indicative finds and parallels will only be considered (Chapters 2 and 4). This outline will hopefully serve as a reference for a future, complete publication of the materials from these areas as well. For them in fact, data at hand are not as comprehensive as those for the Basement and the Burnt Building. This is due both to the nature of the contexts and deposits, often severely damaged by erosion and/or later activities, and to the extent to which the investigation was carried out in the different areas of the site. It is known that excavation necessarily entails a degree of destruction. As reconstruction and presentation to the wider public have always been among the main goals of the works at Husn Salut, in several cases it was not deemed worthwhile to proceed to further excavation, especially when the plausibly predictable results were likely going little way in clarifying key issues, and rather adding only minor details. This has been the case, for example, for the massive outer wall, specifically in its western part; of the main tower; or of the southern part of Area 4. A detailed discussion is dedicated to the nature and diachronic development of Husn Salut architecture, undoubtedly its most striking feature (Chapter 5). Apart from the technical description of the ways in which the ancient builders faced the constraints of working on such an inconvenient building site as the hill of Husn Salut originally was, this part of the report will deal with the implications of certain specific deposits discovered inside the supporting substructure of the main wall and terraces. Different approaches were followed depending on the different classes of materials recovered during the excavations. Stone vessels at Husn Salut were discovered from numerous contexts, and in a number that seems not to be so common at coeval settlement sites. Furthermore, their study reveals some interesting issues that bear contribution to

Magee 1996b; see Chapter 10. Phillips 2010. The suggestion for a different chronology for central Oman was also put forward independently by J. Schreiber (Schreiber 2010). 48 Avanzini and Phillips 2010. 46 47

introduction

17

the discussion on the site’s chronology, specifically for what concerns its later days. Therefore, a complete catalogue of stone vessels is presented (Chapter 6), that builds on the unpublished dissertation of E. Tagliamonte.49 Different is the case for the copper/bronze finds. On the one hand, they were not particularly numerous, with a large part consisting of arrowheads. On the other, the recent discovery or renewed excavation of large-scale metalworking sites like al-Safa near ʿIbri50 and Saruq al-Hadid51 in the Dubai desert, still largely unpublished especially for what concerns archaeological materials catalogues, are dramatically changing the knowledge related to this class of materials, with thousands of objects awaiting a proper study. These actually add to the numerous unpublished finds on display in the archaeological museums of the region. Besides, several metal items from Salut come from the uppermost, mixed contexts. As such, when their typology is not safely datable per se to the Iron Age on the basis of existing dated parallels, any attribution would be speculative at best. Thus, the catalogue presented here was kept as straightforward as possible, with very limited discussion. Basically, it is meant to provide an overview of the typology of finds recovered at Salut (Chapter 7). Animal bones were studied from selected stratigraphic units excavated in the Basement and other chosen contexts, in order to verify possible diachronic shifts in livestock and/or game consumption. Earlier studies of the animal bones collected from two safe contexts of Area 452 and from a peculiar, possibly ritual deposit in Area 153 have also been reviewed and their results considered in the light of new archaeozoological data presented (Chapter 8). Other artefact categories are not so widely attested at Husn Salut. This obviously does not diminish their importance but, as general references to them can be found in a recent publication that summarizes ten years of the IMTO’s work at Salut, not only concerning the Iron Age remains,54 it was chosen not to wait for the conclusion of their detailed study, in order not to postpone further the publication of this report.

A cknowledgements We would like to thank Carl Phillips. His passion and intuition, together with his great archaeological knowledge, have been essential help and inspiration over the fields seasons we spent together at the site. Unfortunately, we cannot accuse him with the possible mistaken made in this work ! Several people have passed by Salut during these years. The authors’ list of this book speaks for itself for the names of some of those who have played relevant parts in the Mission’s achievements. Hosting the works of Enrica Tagliamonte and Marzia Sasso, who both first joined the team as students, is particular pleasant, as it mirrors the efforts to build a working group that could also offer possibilities of intellectual growth to other younger students. Sandra Lombardi has been a wonderful (and patient) editor, and her countless advice and suggestions were, as usual, the quintessence of accuracy. All the work done by the IMTO at Salut would never be possible without the supervision and the constant collaboration of HE Abdulaziz Al-Rowas, Advisor for His Majesty the Sultan for Cultural Affairs: his contribution to the IMTO’s success is invaluable. The Office of the Advisor has been fundamental for the good outcome of the IMTO’s work at Salut. All its members, who, over the years, have worked at the site deserve mention and the warmest thanks. Dr. Said Al-Salmi, General Director, Mr. Hassan Al-Jabri, Director of Archaeological Sites Department, and Mr. Walid Al-Muzaini, representative of the Office at the site, coordinated every activity of the Office in our support.

Tagliamonte 2011. al-Bakri, Genchi and Tosi n.d. Despite the toponym al-Safa being initially used to indicate the site, and in the first communications of the results of its investigation, the correct toponym is ʿUqdat al-Bakra (Giardino 2017: 64). 51 Most recently Weeks et al. 2017; Contreras et al. 2017. 52 Rasile 2011. 53 Wilkens n.d. See also Chapter 5. 54 Salut 2015. 49 50

1. Geographical and environmental setting michele degli esposti , mauro cremaschi, alessandro perego, andrea zerboni

G eo - environmental

zones of the

O man

peninsula

Three prominent features characterise the geography of the Oman peninsula:1 the sea/ocean, the al-Hajjar mountain range, and the desert. On a physical map of the region (fig. 5), the first thing that catches the eye is the long crescent shape of the alHajjar mountain range, extending for almost 600 km from the Jaʿalān in the southeast to Ruʾus al-Jabal in the north, where the Musandam peninsula (Ras Musandam) closes the Persian Gulf and delimitates the Strait of Hormuz as the gateway to the Arabian Sea. Its average width oscillates between 30 and 50 kilometres, with points were it enlarges up to 90. Maximum heights are found on Jabal Shams and Jabal al-Akhdar, almost in the middle of the range, where the altitude is around 3000 metres. The barrier represented by these mountains, separating the eastern coast and adjacent plain from the piedmont plain on their west, is more apparent than real. In fact, all along the mountain ridge a series of wadis dissect the hills’ sides following fault lines, thus providing relatively easy transit routes. Studying the settlement networks that developed along these routes,2 both from a diachronic and a synchronic perspective, is of particular interest for understanding how fully-evolved an integration existed between the coast and inland Oman, a tight link that can be traced back at least to the V millennium BC.3 The main trans-mountain itineraries are, from north to south, the corridor connecting Fujaira with the western coast along Wadi Ham and then via al-Dhaid; Wadi Jizzi together with its bifurcation into Wadi Suq, connecting the oasis of al-ʿAin (nowadays split between the UAE and the Sultanate of Oman) with the port of Sohar; the Wadi Hawasina-Wadi Kabir system, connecting central-northern Batina with the area of Yanqul and Dank, and further southwest to ʿIbri; Wadi al-Fulaij, connecting the port of Sur with the northern end of Wadi al-Batha. Another, neatly wider

1 From a strictly geographical point of view, the use of the term ‘peninsula’ to indicate the territory of modern United Arab Emirates and northern Oman is rather inappropriate (cf. Yule 2016: 31-32): the only real peninsula is that of Musandam at the southern end of the Persian Gulf. However, ‘Oman peninsula’ has become a widely favoured expression in the literature, and can find some explanation. In fact, the desert belt formed by the eastern fringes of the Rubʿ al-Khālī and the Sands somehow separates this relatively densely populated area from inner Oman and Saudi Arabia, communication being allowed by the southern and northern coastal corridors and the central corridor that connects Nizwa to Salala. Hence, the UAE and northern Oman result in being more oriented toward the Persian Gulf and the ocean than toward the opposite side of the Arabian Peninsula, a situation that is partly mirrored in the archaeological records, at least from the Bronze Age to the Iron Age (although for the latter period some elements could indicate contacts were less ephemeral than usually believed – see Chapter 11; for a concise overview on the middle and long-range connection witnessed at prehistoric Salut, see Degli Esposti 2015a). 2 See for example Wadi al- Qawr in the Emirate of Ras al-Khaima (Phillips 1987; 1997). 3 Cleuziou and Tosi 2007: 7.

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

20

F igure 5 – Physical map of South East Arabia, with indication of the geo-environmental zones discussed in the text and of the main natural connection routes.

passage is the so-called ‘Samail Gap’, linking Nizwa and Izki to the Muscat area, most likely the favourite route for Salut’s dwellers heading to the coast already in ancient times (fig. 5).4 It is worth noting that during the so-called Iron Age II period (1100/1000-600 BC), when the region witnesses a substantial growth in settlement density and, apparently for the first time, all the landscape units outlined below are occupied,5 wadi valleys appear to be preferential places for settlement, most probably also in view of the possible control over, and exploitation of, communication routes. For those wishing to move from inner Oman to the coast, an alternative to crossing the mountains along wadi valleys is following the western piedmont either northward, reaching the Gulf coast at Ras al-Khaima, or southward, joining the mentioned Wadi al-Batha route in order to reach the Arabian Sea coast near the al-Suwayh and Ras alJibsh coastal lagoons. Not only are intra-regional connections influenced by the presence of the al-Hajjar range: the latter is also of some relevance in the definition of the different climatic zones in the region, reflected in different environmental areas. Moreover, such a remarkable elevation gradient, at such a short distance from the coast, could hardly not strongly condition air-mass circulation, as specified below. As outlined by M. Tosi,6 at least five main ecological niches can be distinguished in the region, on the basis of their geomorphological characteristics; significantly, their layout is mostly parallel to that of the al-Hajjar range (figs 5, 6).



Cf. Wilkinson 1977: 15-16. Cf. Magee 2000: 33. 6 Tosi 1975: 190-192. The same distinction was more recently adopted by P. Magee (Magee 2000: 29-30). See also Cleuziou 2007: 3-15. 4

5

and

Tosi

1 . geographical and environmental setting

21

F igure 6 – An altitude profile along an ideal section from the coast to the desert edges with indication of the ecological niches of the Oman peninsula (redrawn from Tosi 1975: fig. 4).

A first zone (Z1) corresponds to the al-Hajjar range itself. Here, mountain animals and vegetation occupy a habitat that has a tendency toward scanty vegetal cover. A great deal of human endeavor, aimed at the construction of terraces and irrigation systems, finally led to the establishment of numerous settlements; at the same time, high elevations favour a not negligible rainfall rate (see below). On both sides of the mountain chain stand two strips of land whose elevation is varied but always above 100 metres above sea level. To the east (Z2), the arrival on the less inclined slopes of the wadis coming down from the mountain valleys led to the deposition of gravelly fans. These make human settlement difficult, limiting it to the wadi valleys, to the smaller lateral ones and to the ancient terraces overlooking them. Conversely, west of the al-Hajjar range are the most favourable conditions for human settlement of the whole area found (Z3). Here, around the 500 metre a.s.l. contour line (thus including the territory around Nizwa and the Salut plain, whose average elevation is c. 440 metres a.s.l.), the most abundant precipitations of the whole eastern Arabia occur, feeding the catchment basins of the different wadis. This situation supports the existence of the numerous oases scattered from Buraimi/al-ʿAin to Nizwa, although other relevant ones stand also further south, as for example Adam. Moving farther from the mountains’ axis, to the east one encounters the coastal plain of the Batina, around 20 kilometres wide (Z4). Here, the settlement largely relies on marine resources, although the final parts of the different wadi courses, deeply incised, comprise wide terraces that allow intensive cultivation. It may seem a paradox that this area, left aside the coastal lagoons (or khors), is the archaeologically poorest of the region. Indeed, this is due to some extent to a more limited research7 and, above all, to the fact that abundant alluvial sediments and coast line variations most likely buried or eroded a large part of the ancient settlements. The last ecological zone barely fits the parallel distribution pattern of the others: it is the plain land of the Sharqiya (Z5), occupying the southeastern part of the Oman peninsula to the west and south of the southernmost reaches of the al-Hajjar range and adjoining the coastal region of the Jaʿalān. Though similar to western piedmont Z3, it differs from it in that it has direct access to the Indian Ocean and a more limited influence from desert areas, although some small ones (Wahiba Sands) are included within it. Even here, the settlement relies on agricultural exploitation of wadi basins, often within a frame of complementary exchange with marine resources from the coastal settlements. The differentiation between these zones results in “a wide variety of resources, which can be differentially exploited and provoke different responses and adaptive strategies”;8 of these responses, traces may be searched for in the archaeological records.

Recent years have seen an inversion in this tendency, see e.g. During and Olijdam 2015; al-Jahwari, Douglas and al-Belushi n.d 2012; 2016; Batinah 2016. 8 Magee 2007b: 85. 7

n.d.

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

22

R egional

climate and rainfall

Although the abovementioned geographical features play a relevant role in its definition, numerous other factors influence the climate of the Oman peninsula, their combination causing highly localized and erratic precipitations (fig. 7). Annual average rainfall in fact ranges between 36 mm at Sohar and 171 at Nizwa,9 but different figures are available for almost every locality;10 only occasionally does it exceeds 250 mm11 and thus the region classifies as arid.12 There is however one unique zone in the whole of South East Arabia where rainfall can reach 250 mm and even more: the drainage basin of Jabal al-Akhdar,13 one of the three massifs of the al-Hajjar range that reach above 2000 m a.s.l., together with the Musandam peninsula and the Sharqiyya massif.

F igure 7 – Average rainfall rates at three sites of the northern Oman peninsula, showing extreme irregularity, most exacerbated at Nizwa (redrawn from Tosi 1975: fig. 3).

Indeed, all these massifs are formed by calcareous rocks that, in response to precipitations, show the development of karst features and, subsequently, a very fast drainage of underground waters. Moreover, as the al-Hajjar range is characterized by scarcity of soil and vegetation, the run-off is rapid. The numerous wadis dissecting these reliefs transport alluvial deposits that in turn drain water towards the underground. Of these three massifs, Jabal al-Akhdar is the only one that is flanked by sediments allowing an accumulation of underground water at accessible depths.14 An overview of the available data on rainfall standards in the drainage basin of Wadi Bahla (including Wadi Saifam – Table 1), immediately brings to light some relevant aspects: the mountainside receives three times the amount of precipitations than the regional average, and this is affected by limited evaporation loss; the rainfall rate not only decreases as one gets farther from the mountains, but it also becomes less frequent and reliable; and high evaporation loss on lower lands must be also accounted for, since it is reflected in the high aridity index that characterises the region.15 In these areas, agriculture becomes viable only in the presence of superficial water flows or of accessible underground water.16 While the primary sources of rainfall for the region are Mediterranean frontal systems, most frequent between January and March,17 other interplaying factors are:18



Cleuziou and Tosi 2007: 8. E.g. Wilkinson 1977: table 4; Stanger 1994: tables 1, 2. 11 Cf. Fleitmann et al. 2007: 173. 12 Cf. Weyhenmeyer et al. 2002. 13 Stanger 1994: fig. 14; Weyhenmeyer et al. 2002. 14 Stanger 1994: 92-93. Both the Musandam peninsula and the hyper-arid Sharqiyya massif, in fact, have direct drainage to the sea, in such a way that all underground waters not directly conveyed into the sea are found at inaccessible depths, due to the fact that the carbonatic formations reach below sea-level. 15 Stanger 1994: 92. The aridity index expresses the evaporation/precipitation ratio. 16 Perennial superficial water flows are equally absent in the mountain areas, although some of the main wadis are almost permanent (Wadi Jizzi, Wadi Samail, Wadi Daiqa). Therefore, even there agriculture must rely mainly on underground water, recharged partly at the foot of the mountains and partly in the highest massifs of the al-Hajjar range. 17 Weyhenmeyer et al. 2002. 18 After Stanger 1994: 89. 9

10

1 . geographical and environmental setting

23

Locality

Mean annual rainfall (mm)

Elevation (m a.s.l.)

Distance downstream from the al-Akhdar watershed (km)

Sayq (J. al-Akhdar)

284

2000

8

al-Hamrā

105

625

12

Bahla

99

600

25

Bisya

69

400

52

Adam

72

227

111

T able 1 – Mean annual rainfall at a series of sites in the wider area surrounding Salut (after Stanger 1994: table 2).



- the region stands at a latitude, between 22° and 24° 30’ N, that is subjected to the influence of the descending part of a Hadley cell. Airflow from the Equator touches the area after discharging excess humidity above the tropical regions in the form of rains; moreover, it loses further relative humidity as it heats up. This process is the general cause for sub-tropical zones’ aridity; - seasonal shifts of the Intertropical Convergence Zone (ITCZ) can influence the peninsula, particularly between July and August, causing minor summer rains; - the Oman peninsula stands at the edges of two airflow systems: the Western Mediterranean flow on the one side, and the Indian Ocean Monsoon on the other. Both reach the area after losing most of their humidity, and as such precipitations linked to them can happen only when strong depressions occur; - the Persian Gulf, basically a large warm sea with reduced depth, creates a strong contrast with inner areas’ climate (for example, a remarkable gap in humidity at low distances); - the presence of a low airstream that crosses the Gulf during the summer, with a marked inversion layer, generates most of the precipitations at low elevations, so that the mountains of Oman block most of the air and moisture flow (and the Hadramawt and Zagros mountains as well); - although the Zagros mountains themselves receive the majority of the Western Mediterranean flow’s humidity, high pressure systems can occasionally form in the Caspian region, that deflect the eastern Mediterranean flow southward, bringing rare but abundant rains over the Omani mountains.



A recent isotopic and geochronological investigation on speleothems from several caves highlights the complexity of changes in precipitation occurring in Oman since the early Holocene,19 and the interplay between the northward (Mediterranean) and southward (Indian Ocean) climatic domains. According to d18O profiles, between c. 10.5 and 9.5 ka BP (the beginning of the Holocene), the mean latitudinal position of the summer Intertropical Convergence Zone (ITCZ) advanced northward, leading to an increased monsoonal contribution to rainfall. From c. 7.8 ka BP to present, the mean summer ITCZ continuously migrated southward; the intensity and duration of the summer monsoonal precipitation thus decreased gradually. Several short-term climatic events were identified in this phase, superimposed on a general, long-term gradual trend. During the Late Holocene the length of the monsoon season shortened in response to decreasing insolation and the summer ITCZ gradually retreated southward. As a consequence, the total amount of monsoon precipitation decreased in those areas located at the northern fringe of the Indian Ocean monsoon domain. For the first and second millennia BC few information are available on water availability in northern Oman, but a marked decrease in precipitation is registered at c. 2 ka BP.



Fleitman et al. 2007.

19

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

24

T he

plain of

S alut :

potentials for human occupation

Nowadays, the area of Bisya and Salut is just one in a series of oases aligned along the western piedmont of the alHajjar mountain range (Z3 in the scheme described above),20 with an extension that cannot compare with that of other near ones (most notably Nizwa and Bahla), likely as a consequence of a stronger impact of progressive aridification, starting in the final period of the Early Bronze Age.21 Indeed, over the last forty years a limited number of sites dating to the third millennium BC were excavated in South East Arabia, which provided robust evidence for the onset of the subsistence model based on oasis agriculture22 that constituted the base of local economy until recent times. The presence of abundant archaeological remains datable to that same period in the plain surrounding Husn Salut, and in particular the IMTO excavations at the Early Bronze Age site of Salut-ST1, only 300 metres northwest of it, bear witness to its inclusion in this picture of early land exploitation.23 Moreover, evidence for the same process of climate deterioration mentioned above was also provided by the study of the ancient sediments excavated inside the surrounding ditch of Salut-ST1, and of the geological stratigraphy cut by the ditch itself. 24 The choice of this area for ancient settlement appears to be indicative of an intention to place agriculture at the basis of subsistence strategies for the Iron Age population as well, as confirmed by the preferential location of minor sites along the hilly slopes around Salut. Archaeobotanical studies25 support the reconstruction of agricultural activities around the site, with date palms, cereals, sesame, and basil that would furthermore be consistent with a typical, fully developed oasis model.26 Indeed, a number of factors, connected more to its position in relation to the Jabal al-Akhdar watershed than to the wider scale conditions outlined above, makes of the Salut plain a favourable place for human occupation. The most important feature is its relatively good water accessibility, as it is ‘the distribution of fresh water resources’ that “above all has determined the settlement pattern in Oman”.27 Water availability is granted by the wadis that flow along the western side of the mountains, i.e. Wadi Saifam and Wadi Bahla. In fact, the low annual precipitation rate mentioned above would not be sufficient to sustain settled, agricultural communities. However, the majority of meteoric water falling on the Jabal al-Akhdar massif, apart from originating the numerous springs at its margins, infiltrates in piedmont sediments, recharging the shallow aquifers, and through them it supplies relevant modern settlements, such as Nizwa and Bahla. Thus, even if surface runoff in the wadis close to Salut is not constant, groundwater can be accessed in their fans. The case of Wadi Saifam is also fortunate, for its morphology laterally confines the aquifers,

This area can be also identified with the term bajada, that indicates a “zone of coalescing outwash fans at foot of the mountains” (Wilkin1977: 266). 21 That a major deterioration in local climate occurred around 4200 years BP is now largely accepted, based on relevant evidence coming from different contexts, ranging from lacustrine and marine deposits to speleothems (e.g. Mayewski et al. 2004; Parker et al. 2006; Fleitmann et al. 2007; Fleitmann and Matter 2009). 22 E.g. Cleuziou and Costantini 1980; Potts 1994a; 1994b; Cleuziou 2009; al-Jahwari 2009; Méry 2013; Desruelles et al. 2016. The beginning of oasis agriculture represented a real ‘economic revolution’, in the words of S. Cleuziou (2002: 200); however, an alternative, multistepped model has been recently proposed for the evolution of this agrosystem, which aims to re-evaluate the current interpretation of a rapid onset and subsequent unchanged perpetuation (Charbonnier 2017). 23 Early Bronze Age remains were first recognized as a relevant component of the archaeological richness of the Salut plain by the Harvard Archaeological Survey team (Hastings, Humphries and Meadows 1975). The IMTO team excavated one typical Early Bronze Age ‘tower’ at Salut-ST1 (Degli Esposti 2011; 2016) as well as a few tombs primarily dated to the third and second millennium BC but in many cases reused during the Iron Age period (Condoluci and Degli Esposti 2015). A concise overview of pre-Islamic human occupation at Salut is given in Degli Esposti 2015b. 24 Degli Esposti and Cremaschi forthcoming. 25 Bellini et al. 2011. 26 In the sense of a ‘multi-storied’ agrosystem as highlighted recently by J. Charbonnier (Charbonnier 2017: 54, 64), with crops vertically distributed under a palm trees cover that protects them from excess sun-rays and retains moisture. 27 Wilkinson 1977: 36. J.C. Wilkinson (ibid.) also pointed out that the location of hydrogeological resources has “more to do with the way run-off concentrates than with orographic influences on climate”.

20

son

1 . geographical and environmental setting

25

slowing down the loss of water sinking through wadi gravels and allowing access to it at a longer distance from the mountains.28 The location in a plain rather distant from the piedmont also ensures protection from the risk of flooding: as this is mainly connected with rush floods, it can have a much stronger impact on areas further upstream, and gets reduced as a consequence of infiltration through earlier wadi sediments and of superficial retention. At a less evident level, the site also stands at a distance from the mountains that allows water to retain an adequate level of salinity, before reaching an enrichment that would be noxious for cultivation, principally due to evaporation and superficial rocks’ alteration. Currently, waters draining from the Jabal al-Akhdar achieve a brackish salinity around 20-30 kilometres downstream from Bisya.29 In the frame of the generally semi-arid climate of the region, it is anyhow clear that these factors are not sufficient to sustain the subsistence of settled communities if adequate systems of water management are not developed.

T he

geoarchaeology of the

S alut

plain

The geoarchaeological stratigraphy of a certain area can be correlated with the phenomena that led to its formation, the latter often depending to a more or less large extent on climatic conditions. However, these are not the only variables to be taken into consideration, given the obvious influence, for example, of tectonics that can have localised effects. From this derives the uttermost importance of collecting data in the proximity of the investigated area, and specifically from open sections when available, rather than applying interpolations of data obtained in generically close-by areas and used for macro-regional scale reconstructions.30 With this in mind, a geoarchaeological study of the plain surrounding Salut has been carried out, in several phases, as a complementary achievement to the excavations and archaeological surveys. The main aims of this study were the investigation of the geomorphology of the area, and an evaluation of the irrigation systems existing there, their probable date included. Collected data show the potential of the area for a diachronic reconstruction of subsistence strategies in central Oman during the late Holocene, within a more general scenario of climate change and cultural reaction and adaptation to increasing aridity. The investigated area extends north-south from Bisya to Nattala, and is limited by the beds of Wadi Saifam to the west and Wadi Bahla to the East. From a geological point of view the area is included in the sedimentary apron at the southern margin of the Jabal al-Hajjar al-Gharbi, mainly consisting of gravelly alluvial fans and dated to the late Pleistocene.31 Most of the archaeological evidence and of the irrigation features identified so far – mainly represented by falaj (Arabic pl. aflaj) shafts and channels – are located in an area along the eastern bank of Wadi Sayfam occupied by fine, light coloured silty and fine sand deposits, which are indicated as Kabrha Formation on the geological map and are dated to the Holocene. By field walking over a more restricted area north of Husn Salut, roughly 12x8 kilometres in extension and delimited to the west by the fringes of the limestone hills that border the wadi plain and to the east by strongly dissected gravel terraces, five main geomorphological units have been distinguished (fig. 8): A. Piedmont alluvial fans along the western side of wadi Saifam. These are coalescent alluvial fans, gravelly in texture; darker patinas on the surfaces indicates older deposition dates. At the surface of the oldest fans several flint scatters have been observed, possibly Neolithic and even of an earlier date. They are associated with stone structures that indicate good preservation of archaeological features. B. Present Wadi Saifam bed, composed of gravel and sand. The green and olive grey colours indicate that its braided pattern is due to the ophiolite-dominated lithology.

Ibid.: 49. Stanger 1994: 95. 30 As underlined for example, for the region considered here, by G. Stanger (Stanger 1994: 97). 31 Indicated on the geological map as Qf (Geologic Map of Oman, 1. 250 000, sheet NF 4007 NAZWA).



28 29

26

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

1 - canal Jibrin 2 - falaj Salut 3 - falaj Shaww 4 - falaj Farud 5 - falaj Bisya 6a - falaj al-Hail 1 6b - falaj al-Hail 2 7 - falaj al-Dhabi

underground tunnel with wells (thuqaiba) open channel abandoned Islamic village Husn Salut and core of the archaeologiacal area

bedrock

alluvial fans of wadi Bahla of different ages

alluvial plain of wadi Saifam (Kabrha Formation)

recent fluvial deposits

F igure 8 – The geomorphological units distinguished in the surveyed area, and the surveyed aflaj system in the surroundings of Salut.

1 . geographical and environmental setting

27

C. Alluvial plain. It corresponds to the extension of the above-mentioned Kabhra Formation and it is easily identified on the satellite imagery due to its whitish colour, which contrasts with the dark colour of the surrounding alluvial fans. It is delimited to the west by the present course of Wadi Saifam, and to the East by the distal fringes of the alluvial fans descending from the north. D. Eastern alluvial fans. These are composed of fine-grained ophiolitic gravel, giving the dark green colour visible on the satellite image. The presence of some cairns, probably dating to Pre/Protohistoric periods, indicates that these landforms were already deposited in the middle Holocene and they may date back to its beginning. Some aflaj are cut in this unit, but no archaeological material has been observed associated with them. Their good state of preservation, however, suggests they are probably of recent date. E. Gravel terraces, made up of ophiolite and limestone large pebbles, coated by black desert varnish. Some possibly Palaeolithic and Neolithic flint artefacts have been observed on these terraces, as well as some cairns. It is worth mentioning that the limestone hills at the western edge of the survey area, upslope from the alluvial fans of unit A, are rich in flint outcrops, and several prehistoric quarrying sites and related workshops were also found, associated with them. Furthermore, a malachite outcrop has been identified in the same area, albeit in the absence of any other indication of possible ancient metallurgical exploitation. A rock shelter stands close to it but it is devoid of archaeological stratigraphy and/or material. However, a strip of travertine deposits has been found inside it, which is relevant for palaeoclimatic reconstruction since it may indicate a wetter climate existing at the beginning of the Holocene. The Kabrha Formation appears to be an essential component of the agricultural exploitation of the area, as mirrored in the concentration of irrigation features dug through it and in the dense network of now abandoned fields visible also on satellite imagery.32 This is due to its pedological characteristics: silty texture and light structure, easily ploughable, and good water retention. These conditions, provided that water availability exists, makes the Kabrha Formation the most suitable soil for agriculture in a region where soils are in general stony and shallow and consequently mostly unproductive. Its sedimentological characteristics indicate that the Kabrha formation consists of windblown particles; while the lithic industry, found mainly at its base, suggests an early Holocene age, consistently with the indication provided by the geological map. Its formation is to be related to a wetter climatic phase during which the area was covered by bush grassland, which trapped the dust brought by the northern winds. A similar phenomenon has been observed elsewhere in Oman and in particular along the southern margins of the Nejd desert.33 In several cases the Bronze Age monuments located on outcrops of bedrock and on small gravel bars inside the Salut plain appear to have been partly covered by alluvial sediments, indicating that the Kabrha deposits have undergone some colluvial redistribution after the middle Holocene. Additional information on the structure and thickness of the soils was obtained with the excavation of several test pits in the plain surrounding Husn Salut, both east and west of the site. At the base of each trench a hard and thick layer of caliche was reached, consisting of a silty-sandy to gravel deposit, strongly cemented by CaCO3; its origin is possibly pedogenetic, as it formed through evapotranspiration during a period with higher precipitations, interrupted by dry seasons. Even if no artifacts are included within this horizon, it could be dated to the early Holocene. A verification of this date is provided by the excavation of Salut ST1 Bronze Age contexts: the tower itself is in fact founded above, and the surrounding irrigation features cut through, this cali-

32 These are recorded on the topographic map as enclosures. Field walking highlighted a squared pattern of ‘relict’ agrarian ditches, aboveground masonry water channels, and stone walls meant to retain the fertile soil and prevent its erosion while possibly marking field boundaries. The cultivated fields appear to have been abandoned for years, and today the area is dry and almost deserted, with palm groves and cultivated fields being restricted to limited gardens where water is pumped from deep wells. 33 Cremaschi and Negrino 2002; 2005.

28

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

che layer, thus providing a terminus ante quem for its pedogenesis.34 Consistently, on top of the upper boundary of the caliche, which is generally erosive, a quite developed, clay-loamy and organic matter-rich soil is present that includes Bronze Age pottery. This, interpreted as the buried soil of the Bronze Age, was not continuously observed in the pits, as in many cases it is replaced by a coarse sandy to gravelly deposit, corresponding to a fluvial facies. The upper limit of the Bronze Age soil is in some cases represented by a small stones layer, eventually very rich in pottery fragments, that separates it from an upper, less developed soil; the latter is sandy-loamy, less rich in organic matter, and the cultural material collected inside it indicates an Iron Age date. Atop this soil stands a sandy-loamy loose sediment that is ubiquitous in the area and represents the outcome of recent wind erosion and re-sedimentation. Several residual, thin buried soil shreds were observed elsewhere in the Salut plain, covered by aflaj spoil heaps; they constitute the remnants of a poorly developed soil, possibly developed in the Iron Age, as would be suggested by the occasional archaeological material entombed within. The collected data provide a basis for some preliminary considerations on the evolutionary processes leading to soil formation; these will be addressed in more detail elsewhere, with the support of a robust analytical dataset.35 In the proximity of Husn Salut, the soils lying above the caliche layer display a thickness that is not consistent with that surveyed in other parts of the area: they are thicker close to the Iron Age settlement at the centre of the oasis and become thinner towards the open plain. As these soils are systematically connected to pottery and other evidence of human activity (i.e. charcoal, exotic stone fragments) they have to be regarded as artificial and related to long lasting agricultural activity (addition of fine textured sediment, manuring, irrigation, etc.). At least two distinct phases of pedogenesis and human activity can be envisaged: the older corresponds to the Bronze Age soil, locally eroded by later fluvial erosion, prior to the Iron Age soil formation. Worth pointing out is the lack of any soil that can therefore be dated to the Islamic exploitation of the plain, somehow contrasting with the abundance of potsherds dating to this period. A possible explanation can be found in the occurrence of a strong wind erosion during the last centuries: this caused the deflation of the soil cover dating to this phase, leaving a residual deposit of windblown sediments the surface of which is scattered by Islamic potsherds.

I rrigation

systems at

S alut

The different hydraulic features mentioned above form a substantial, multi-period irrigation system of which aflaj and canals appear to be two functionally closely connected components. The following aflaj were distinguished in the plain flanking Wadi Saifam to the east, north of Husn Salut: Falaj Salut, Falaj Farud, Falaj Shaww, and Falaj al-Hail 1 and 2. Two more aflaj were instead located in the alluvial gravelly fans of Wadi Bahla, near the junction with Wadi Saifam, and as such have to be considered part of the network serving the same larger area of Bisya and Salut: Falaj Bisya 1 and 2, and Falaj ad-Dhabi (fig. 8). Several of the surveyed aflaj (Falaj Farud, Falaj Shaww and Falaj Bisya) in fact end in open channels that serve the purpose of bringing water to the surface, where it would be distributed among the fields. Some more complex cases of this connection have been observed. The shafts of the middle part of Falaj Farud for example were excavated through the western levee of the adjoining open channel and therefore the falaj has to be considered more recent than the channel itself. In other cases the falaj was excavated inside the channel, again indicating its later date and showing that a substantial lowering of the water table occurred between the two phases. In one fortunate case the section of one of these open channels could be surveyed, thanks to a modern bulldozer cut. The section is located near the beginning of the open channel originating from Falaj Bisya: it is cut in the bedrock



The initial occupation of the Bronze Age structures at Salut ST1 can be placed around mid third millennium BC (Degli Esposti 2016; DeEsposti and Cremaschi forthcoming). 35 Cremaschi et al. forthcoming. 34

gli

1 . geographical and environmental setting

29

down to the depth of five metres and is filled by water laid sediments. Iron Age potsherds were exclusively observed in the lower fillings, while the upper deposits contained later, Islamic sherds. By way of a general point, the observed aflaj can be divided into two broad groups: very degraded ones, with backfilled shafts and deflated shaft spoil heaps (Falaj al Hayl 1 and 2, Falaj ad-Dhabi); and better preserved ones, with shafts still open down to two to ten metres and shaft spoil heaps up to two metres high. Potsherds of Islamic date are often found in association with the better preserved ones; for the others, archaeological material is completely absent. This suggests that the presence of Islamic material is more likely to be related to a later moment of re-use or restoration, rather than to the moment of falaj construction. Aflaj’ shafts along Falaj Shaww, Falaj Farud and Falaj Bisya also provided numerous potential stratigraphic sections that help to reconstruct the sedimentary history of the alluvial plain in which they are cut. It was thus possible to identify three main layers through which they had been excavated: - Topmost unit. About one metre thick, reworked Kabrha silts, mostly including Islamic pottery and occasional Iron Age sherds. - Cemented gravel unit. About 2-3 metres thick gravel deposit, strongly cemented by calcium carbonate (calcrete horizon), dominated by limestone and ophiolite lithotypes. - Gravel unit. About 2-3 meters thick weakly cemented gravel and sand deposit, dominated by limestone and ophiolite lithotypes. - Cemented marls (bedrock). Planar laminated marls or deeply cemented fine sediments, up to ten metres in thickness. Units 1 and 2 conveniently match the plain’s stratigraphy observed in the test pits mentioned above; most of the shafts cross these upper three units and their underground tunnel (in most cases not accessible) is excavated in the hard cemented limestone bedrock. It is worth noting that cemented conglomerates and caliches have been indicated as the principle aquitards throughout the region, affecting, in a determinant way, flow and infiltration rates, specifically for the drainage basins of central Oman.36

T he

ancient oasis of S alut : the age of the irrigation system and its relation to climatic change

The majority of the aflaj in the plain were in use during the Islamic era and were still active a few decades ago, when several of them were restored through the building of concrete structures.37 However, an older generation of aflaj can be recognized in the most deteriorated ones, as mentioned above. The pottery assemblage collected in association with the different aflaj also reflects two major periods, as it includes early Islamic glazed Abbasid pottery alongside later 14th – 15th century AD glazed sherds, but also a not negligible quantity of typical Iron Age pottery. Hence, it can be assumed with reasonable certainty that during the middle Islamic period a pre-existing, wellstructured irrigation system was re-used and in all likelihood extended or at least largely renewed. The large ‘fossil’ open channels in the central part of the Salut plain, and probably some of the most degraded aflaj alike, might be the remnants of the earlier hydraulic network. On the basis of occasional archaeological indications, namely the surveyed section of Falaj Bisya’s open channel and other scattered potsherds, this earlier network may be referred to the Iron Age.

36 Wilkinson 1977: 47. For a discussion of the caliche formations in the area of Bisya in relation to a tentative reconstruction of the sedimentary regime south of the al-Akhdar massif, see Stanger 1994: 96-97. 37 Such a practice was directly witnessed by the IMTO’s team in winter 2015, when the shafts of Falaj Salut were being lined with concrete. According to local informants, the last active aflaj definitively ceased to supply any water during the 1970s; wells were later excavated inside the abandoned shafts, with the aim of reaching the lowered water table, but they are also dry now.

30

michele degli esposti , mauro cremaschi , alessandro perego , andrea zerboni

More solid evidence is however provided by fragments of calcareous tufa in the form of lateral flowstones and stalactites,38 which were cut and removed during the Islamic phase of renewal of the falaj system and subsequently discarded on the shafts’ spoil rings. These fragments were particularly evident along Falaj Shaww, where a few were collected for U/Th dating: the results indicate a continuous deposition from the 6th century BC until the 12th century AD.39 The beginning of this chronological range represents a terminus ante quem for the construction of the tunnel, thus to be attributed at least to the late centuries of the Iron Age. Moreover, along Falaj Faruq an outcrop of a buried soil was spotted below the open channel levee, consisting of a reddish clay-rich B horizon, including rare potsherds, charcoal and land snails, which have been collected for radiocarbon determination. A TL age of 3900±200 years BP was so far obtained for one potsherd;40 this provides a terminus post quem for the open channel excavation, corresponding to the transition between the Early and the Middle Bronze Age in the region.41 On the basis of the collected data, the oasis of Salut in its heyday may have reached a size of about 75 square kilometres of cultivated land, sustained by water provided by irrigation facilities consisting of open channels and connected aflaj. Significantly, this extension corresponds to with that of the Kabhra Formation, the early Holocene deposit that, as seen above, constitutes the most profitable soil for agriculture in the region. Today, the surface occupied by palms and cultivated fields is reduced to about 6 square kilometres, reflecting a scenario of dramatic aridification and consequent collapse related to the inactivation of the irrigation system. Between these two ends, the reconstruction of diachronic land use in the plain of Salut is difficult, due to the scarcity of data. While an Iron Age date for an early exploitation of the area and the construction of the original irrigation network defined above seems probable, and must be correlated with soil exploitation, it is more difficult to indicate a date for the overall beginning of irrigated agriculture. However, data from the recent investigation of the Early Bronze Age site at Salut-ST1, and in particular from the study of the massive sediments accumulated inside the huge ditch that surrounds the central stone tower,42 harmonises with accepted palaeo-climatic reconstructions that places that period still close enough to the wet early Holocene where a larger water availability could be enjoyed, before the occurrence of a major climate deterioration that marked the end of the third millennium BC.43 In fact, this availability must be considered the necessary prerequisite for an effective and profitable onset of intensive agriculture that cannot be seen as a response to an already incipient aridification.44 The presence of a soil cover during the Bronze Age is indicated by the test pits described above and by the identification, inside the ditch’s filling at Salut-ST1, of micro-charcoals reflecting the existence of cultivated surfaces on the site’s surroundings, that were manured and upon which fires were lit.45 However, there is at present no undisputable evidence for the existence of irrigation facilities at that time. In fact, although the presence of water inside the ditch and other contemporary channels is shown by their inner sediments’ nature, there is no clear indication that may help in defining their precise function. What is so far ascertained is that the highly probable agricultural exploitation of the plain during the Early Bronze Age was not sustained by so well-developed an hydraulic system as the one established later in the area, i.e., there is no evidence for a possible falaj system prior to the Iron Age. Indeed, the development of such a system has to be recognized as the most probable basis for the dense occupation of the Salut plain during the Iron Age, and served as a response to a moment when aridity further increased, leading to climate conditions as arid as today.46 The reconstruction based on archaeobotanical data agrees in depicting such a dry palaeoenvironment, dominated by dry shrublands, with Acacia sp. (acacia), Tamarisk sp. (tamarisk) and Ziziphus sp.

Most significantly, they display a squared morphology that indicates deposition inside a man-cut tunnel. Cremaschi et al. forthcoming. 40 Ibid. 41 That is, between the so-called Umm an-Nar and Wadi Suq periods. 42 Degli Esposti and Cremaschi forthcoming. 43 See note 21. 44 Cf. Cleuziou 2009. 45 Degli Esposti and Cremaschi forthcoming. 46 Fleitman et al. 2007: 180. 38 39

1 . geographical and environmental setting

31

(jujube) probably diffused in the surroundings of the site.47 Significantly, these data coherently place the plain of Salut within the larger picture, now widely accepted, which places the introduction of an aflaj-based subsistence strategy in the local Iron Age II period.48 U/Th dating on calcareous tufa that encrusted their underground tunnels indicate that at least some of these aflaj were already in use from 540 BC and water availability inside them lasted until 1154 AD, when it was exhausted by the dropping of water tables.49 In the 14th century AD, water was derived through a canal from the area of Bahla, where water availability still subsisted and fed only a part of the oasis. However, most of the latter’s surface was dominated by wind erosion until the introduction of deep wells mechanically excavated, which currently sustain cultivated fields in the area.

47 Bellini et al. 2011. 48 For several of such structures, however, an Iron Age chronology is based more on their general context than on undisputable stratigraphic – or of other nature – evidence, as recently pointed out in detail by J. Charbonnier (2015: 50-68), who also provides an exhaustive overview of the different cases, with bibliography. 49 Cremaschi et al. forthcoming.

2. Husn Salut in the Iron Age michele degli esposti , chiara condoluci

A rchitectural

and stratigraphic phasing

The Iron Age architectural evolution of Husn Salut (pl. 1) can be better structured in three main phases, distinguished on the basis of a stratigraphic study, combined with the indications provided by radiocarbon dates and archaeological materials. At an earlier stage of the IMTO’s investigation, a preliminary two-fold division of this fundamental period of the site was proposed;1 it can now be amended and integrated in light of new data provided by further excavation. These phases will be referred to throughout this book as ‘Husn Salut I’, ‘Husn Salut II’ and ‘Husn Salut III’ phases, hereinafter abbreviated HSI, HSII and HSIII. HSI and HSII phases share an identical material culture and were distinguished on the basis of an evident change in the site’s layout, namely the obliteration of one of the earliest buildings (the so-called Basement) and the construction of the new Building 1 atop its buried remains (discussed below). HSIII also bears witness to a relevant architectural re-arrangement, comprising the modification of the outer walls structure in the eastern part of the site and a substantial reconfiguration of the Burnt Building’s plan. However, the definition of this third phase in the site’s evolution primarily depends on a change in the associated cultural material, mainly for what concerns the pottery assemblage and, to a minor extent, the stone vessels’ typology. Following the three-fold division of the Iron Age period in South East Arabia proposed more than twenty years ago by P. Magee,2 HSI and HSII phases would comprise the whole Iron Age I and Iron Age II periods (roughly 1300-600 BC in the commonly accepted chronology), while HSIII phase could be considered broadly to correspond to the Iron Age III period (c. 600-300 BC). Although apparently straightforward at first sight, this sequence, when considered together with the associated material culture and the now numerous 14C dates available, allows some relevant discussion of this commonly accepted chronological model, and in particular of its possible application to central Oman, as will be detailed in Chapter 10. Several sub-phases or re-arrangements are often evident only in particular places and are generally not so easily definable due to later structures’ impact as well as to an extremely complex stratigraphy. The first thing to be underlined is that the impressive stone wall that today characterises the site’s perimeter actually belongs to the HSIII phase,3 at least in its eastern part, east of the main mud-brick platform that lies atop the site (Terrace I); the same is true for the massive tower projecting onto the plain.

Avanzini and Phillips 2010: 97-99. Magee 1996b. 3 Indeed, a not negligible amount of Islamic pottery was discovered inside the trenches that were opened with the aim of investigating the chronology of the outer walls’ construction. However, these appear to be only scattered in the uppermost deposits and not close to the foundations, thus suggesting an Islamic re-arrangement of the ancient walls rather than a new construction. Moreover, the scale of these walls is totally 1 2

34

michele degli esposti , chiara condoluci

Husn Salut I Several consistent radiocarbon dates indicate that the site was established by the second half of the 13th century BC at the least (see Chapters 3 and 10). Its earliest layout (pl. 2) was characterised by a system of outer walls built in mud-brick over stone foundations, the plan of which appears to be much less linear than the later, outer stone wall. At least five different, connected walls were in fact distinguished just in the limited deep trench appositely excavated in the north-northeastern part of the site: M100, M232, M237, M239, M242 (see Chapter 4). Their maximum preserved height clearly shows that the original aspect of the site was by no means less impressive than the later one, further enhanced by the location on the upper slopes of the hill (fig. 9). This complex wall system embraced the eastern part of the site, while the western side (including Areas 1, 2 and 3, see fig. 10) was already delimited by a huge stone wall, M9, running all the way to the entrance located in the southeastern side of the site, and tied with M8 to close the northeastern corner of the Basement.4 The distinctive feature of the site, that is, a system of terraces built on compartmented foundations occupying the whole area encompassed by the perimeter walls,5 already existed during the HSI phase. Entrance to the site was through a passage in the southern central part of the outer wall, via two flights of stairs set at right angles to each other (fig. 11).6

F igure 9 – The original perimeter walls of the site, built in mud-brick over stone foundations, after restoration (cf. fig. 60).

inconsistent with the poor hut-like dwellings associated with the Islamic occupation of the site (see Chapter 11). Moreover, the construction of the outer walls, and of the main tower in particular, seems to be part of one and the same project as the larger terracing walls near the eastern foots of the hill, which in turn are undoubtedly linked to the extensive terrace system that defines Qaryat Salut, the settlement currently under excavation. 4 It is however unclear whether M8 already had the preserved layout during this phase, or if it the outcome of a later modification. 5 See also Chapter 4. 6 Due to the bad state of preservation in the area it cannot be excluded that a short ramp leant against the outer wall lead to a single flight of stairs placed at right angle with the latter, and laterally framed by walls M54 and M55. Immediately to the northeast of the entrance, beyond wall M55, the precise layout of the site during HSI phase remains elusive, as no specific investigation was carried out there.

2 . husn salut in the iron age

F igure 10 – Excavation Areas at Husn Salut.

F igure 11 – Husn Salut in April 2017, with the entrance to the upper part of the site in the foreground, after restoration.

35

36

michele degli esposti , chiara condoluci

These steps landed immediately east of wall M45, a long southeast-northwest wall that was connected with the eastern perimeter wall of Terrace I, M67. From this point, three possible paths diverged: to the right, Area 4 was reachable via another possible staircase, framed by walls M248 and M249 but very badly preserved; moving ahead, walls M213 and M212 formed a first large platform followed by a curving step between M212 and M211, in turn leading to another platform delimited by walls M45 and M210. From here, moving further ahead Terrace I was accessible, while turning left, over the step marked by M210, an area standing above the massive outer wall was probably reached, although its layout eludes any possibility of reconstruction due to substantial erosion and collapse. With the exception of Terrace I, the topmost and widest terrace, upon which the so-called Burnt Building was erected, no traces of other buildings were discovered over the lower, smaller terraces. From Terrace I, access was anyhow possible to another building located against the perimeter wall in the western corner of the site. This building, reachable walking down a flight of steps, stood at a notably lower elevation than Terrace I (and the Burnt Building), and was consequently labelled as the ‘Basement’ (fig. 12). The location of the Basement further underlines how the ancient builders partially adapted the plan of the site to the bedrock layout: the latter in fact is found just a few tens of centimetres below Terrace I’s surface, and sinks abruptly in correspondence with the Basement’s rooms. Both the Basement and the Burnt Building are characterized by a plan that has a corridor as its central node. Access to the Basement was from the eastern end of this corridor, here distinguished as room 1a,7 via the mentioned stairs. Two series of rooms opened on the opposite sides of room 1a, and their layouts were extremely different from one side to the other, pointing to different functional destinations, albeit not definable. To the north, three small premises were directly accessed from the corridor: rooms 2a, 3a and 4a.8 From room 4a, the small room 5a9 was reachable. All these rooms were delimited by the outer wall M9 in its original layout, later maintained and incorporated in the renewed perimeter wall. Beyond the northeastern wall of room 5a as said (M197), wall M9 was connected to the substantial stone wall M8, both enclosing a massive mud-brick and mud fill that almost gave to the area an aspect of a bastion. South of the corridor, only the large room 6a was identified, although its layout is rather uncommon and implicates a possible functional distinction of the room into two units. After entering from room 1a, in fact, a long northwest-southeast corridor was defined by walls M149 and M161, the latter separating it from a larger L-shaped space whose northern end forms a square room.10 The southeastern part of the long corridor stood at a higher elevation than the one close to room’s entrance, and a gently sloping ramp connected the two areas, as shown by the excavation of a trench F igure 12 – View of the Basement’s corridor (room 1a) during excavation (2007), with ruined stairs leading to Terrace I in the background.

1 m wide, 7.3 m in length (from the bottom of the stairs). 1.2x3 m (2a and 3a, almost identical); 1.9x2.4 (4a). 9 2.15x2.5 m. 10 Overall northwest-southwest length 12.85 m, width 6.3 m. Squared space 3.75x4.5 m. 7

8

2 . husn salut in the iron age

37

F igure 13 – SW-NE section through the western corridor of room 6a, showing the inclined ramp formed by US400, unfortunately largely dismantled by an Islamic cut.

between wall M149 and M161 (fig. 13). The portion of the room standing on the other side of wall M161 was not excavated, but one might assume that it stood at the same elevation as the higher part of the corridor.11 This situation might suggest that another entrance to room 6a existed, directly connecting it with the open area atop Terrace I, and likely located in the eastern corner of the room’s squared space, where there seems to be in fact a lacuna in the perimeter walls. While the central corridor of the Basement was oriented along a northeast-southwest direction, that of the Burnt Building (room 6b)12 stood almost at a right angle to it, the two elongated rooms thus mirroring the two main axes of Terrace I. Entrance to the Burnt Building was from the southeastern end of the corridor, leading to the opposite end to a small rectangular room standing on the right, perpendicular to the corridor itself, room 5b (fig. 14). The latter’s entrance was marked by a small pilaster strip at the northern end of wall M49, which might be indicative of the original presence of a door; this, together with the room’s reduced size,13 may indicate that it served some specific – although undeterminable – function. Four rooms opened on the western side of the corridor. From south to north, rooms 1b, 2b and 3b14 were three elongated rooms whose shape is suggestive of a storage function (fig. 15/a), although no solid proof comes from the collected materials. Room 4b,15 occupying the northwestern corner of the building (fig. 15/b), seemingly had a different, larger layout, apparently confirmed by the presence of mud-brick pillar M146 supporting the roof. 16

11 This reconstruction is clearly hypothetical, but seems more plausible than envisaging two opposite ramps on the opposite sides of wall M161. 12 Roughly 12.1x1.85 m of maximum width. 13 3.9x3.0 m. 14 All 8.6 m long, width ranging from 1.6 (2b) to 2 m (1b). 15 Roughly 4.6 x 7.15; northwestern side badly disrupted. 16 This mud-brick feature was in a bad state of preservation, badly damaged by a fire occurring in the whole building and by the construction of a later wall M138. It was only preserved for a height of 0.90 m for a length of 0.75 m.

38

michele degli esposti , chiara condoluci

F igure 14 – View of room 5b and of the northern end of the corridor (room 6b), looking north.

Most of the Burnt Building’s southwestern part, west of the corridor, stands above a massive mud-brick structure that has been referred to over the years as a large ‘buttress’. Indeed, it does not protrude isolated from the site’s perimeter walls, but rather occupies a wide indentation that resembles the opening of a snail shell, with the outer wall resembling its body whorl (pl. 2). A remarkable feature stands to the east of the corridor, framed to the southeast by wall M117 itself and by walls M143 and M4217 that also delimit room 5b: it is the massive, rectangular raised mud-brick platform SF49,18 composed of three adjacent compartments filled with mud-brick lumps and mud and capped by well-laid mud-bricks. These compartments, just like the similar features discovered in Area 4 (see Chapter 4) and elsewhere on the site (Chapter 5), were never used as real rooms, and the materials found inside their fills testify to a different function than the domestic one (discussed in Chapters 5 and 9). It must be noted that the central compartment, delimited to the northwest by wall M78, was not apparently flat levelled at the top, but showed a rectangular, hollow niche that was filled in at a later time (mentioned below). The building technique of wall M143 deserves mention, as it appears to be so far unique at Husn Salut. The upper part of this wall, in fact, was erected above a wooden beam that served as a leveller (fig. 16/a); above it, two different parts of the wall can be distinguished by the the use of bricks of different sizes, although built at the same time: rectangular in the south-central part (65x33x8 cm), square in the northern part (35x35x6/7 cm), the latter bond with a large amount of mud. That this layout was the outcome of partial rebuilding in one moment during HSI phase cannot be excluded. The dismantling of part of the wall allowed the observation of complete mud-bricks, and revealed the

17 Notably, deep trenches confirmed that walls M42 and M117 directly sit on the bedrock, further confirming their belonging to the earliest phase. 18 SF = Structural Feature. Dimensions in this phase 9.7x 3.6m.

2 . husn salut in the iron age

39

a

b

F igure 15 – a) Rooms 1b, 2b and 3b (left to right) in the Burnt Building, with later wall M138 in the background, which reduced their extension westward; b) Room 4b, looking at the corner between M138 and M127, during excavation. The thick, burnt layers that characterise the building are visible, as well as burning traces on the walls’ mud plaster.

40

a

michele degli esposti , chiara condoluci

b

F igure 16 – a) View of wall M143’ northern face, showing the wooden beam and the two differently built parts; b) detail of one of the mud-bricks removed from the wall, with a complex pattern of grooves realized to allow better bonding.

presence of an elaborated geometrical pattern of impressed straight and wavy lines, clearly a trick to assure a better bond between the bricks and the mud binder used to put them in place (fig. 16/b).19 Immediately east of room 1b, and presumably related to the Burnt Building, stands the trapezoidal room 7b, the largest among those dated to the HSI phase.20 The room is delimited by two substantial walls, M56 and M57, made of mud-bricks above stone foundation, both about 1 m thick. These walls actually articulate with the supporting structure of the outer wall, here comprising a succession of massive radial walls. The features that mainly distinguish room 7b are a well-laid, large flagstones floor21 and a central mud-brick pillar supporting the roof, M4822 (fig. 17). Against a small pilaster strip leaning on the northeastern end of M57, a small vertical, cylindrical furnace (SF25) was also unearthed, built with rough small stones and stone flakes embedded in a compact clay mud (fig. 18). Only 0.60 m high, SF25 has an hollow opening at its top, and a second semicircular one through its bottom, on the west side. Although this latter opening is remindful of an oven mouth, as well as recalling a possible furnace taphole, no material was found inside SF25 to shed light on its use.23 Admittedly, there is no way to safely date the flagstone floor to this first phase; besides, its nature is more coherent with that of HSII floors, namely those identified in the later Building 1 rooms (see below). It must also be underlined that both the central pillar of the room and the furnace/oven SF25 were erected above the stone slab floor: therefore, all these features should be better post-dated and attributed to the HSII phase. The flagstone floor, which also extends toward the northeast beyond the room’s limits, rests above a compact preparation made with mud lumps and mud, US95;24 on the edge of the not-eroded part of the room, an oval shaped pit was discovered (SF48), that had been outlined through US95 while this was being laid down. It has to be interpreted

19 Mud-bricks showing finger impressions are reported from the Iron Age I platform’s capping at Kalba (Magee and Carter 1999: 168), and a single example is also mentioned from Tell Abraq’s Locus 56 (Potts 1991: 108, fig. 187). Similar mud-bricks have been recently discovered by the IMTO at Qaryat Salut. 20 Assuming an extension until the outer wall’s external curtain, the room would have reached 11 m along the northeast-southwest axis, with a width of 4.2 m toward Terrace I and 6.1 m near the outer wall. 21 Due to considerable erosion and collapse of the outer wall’s structure, the southern limit of the room is missing, and the paved floor’s extension is also impossible to ascertain. 22 0.70x0.45 m in plan, it was preserved for a maximum height of 0.98 m. 23 See also Chapter 9. 24 US = Stratigraphic Unit (inverse spelling derives from original Italian labeling).

2 . husn salut in the iron age

41

F igure 17 – General view of room 7b, looking south.

as the housing for some now disappeared item, possibly a liquids container, if not as a basin itself. It may be tempting to connect this possible, indeed not large feature,25 to the fact that the room’s floor slopes southward (fig. 19). It is this compacted mud preparation US95 that has likely to be recognized as the actual floor of room 7b in its first layout. The impression of a woven basket discovered above the clay surface US391 between M140 and M14126 supports this interpretation (fig. 20), as it shows the existence of an earlier surface below the flagstone floor further confirmed by the discovery of a smashed jar above the same surface (fig. 21). Difficult is also the reconstruction of the L-shaped space between the northwestern end of the Burnt Building and the Basement. Here in fact, a passage could exist during this period, but extensive erosion of the mudbrick features hampers any reliable hypothesis.

F igure 18 – The small furnace (or tannur) SF25 in room 7b.

25

0.70x0.45 m and 0.40 m deep. In this area the stone floor had been dismantled in antiquity and its slabs likely used in later structures, see for an example fig. 111.

26

42

michele degli esposti , chiara condoluci

F igure 19 – Transversal profile through room 7b, showing its slight inclination southwestward.

F igure 20 – Woven basket impression on surface US391, northeast of room 7b. F igure 21 – Smashed storage jar in situ, immediately north of room 7b, resting above the mud fill contained by walls M140 and M141 (US391).

2 . husn salut in the iron age

43

Husn Salut II Over a period of time possibly spanning a few hundred years, the floor in the Basement was raised several times and the rooms became progressively in-filled, and this produced the sequence of superimposed archaeological layers excavated by the IMTO, together with the stratigraphy investigated within the Burnt Building perimeter.27 Based on the sequence of radiocarbon dates associated to the Basement’s stratigraphy and discussed in Chapters 3 and 10, at some point around the first half of the 11th century BC at the latest, large parts of the earliest buildings were affected by a large fire, that marked the end of the HSI phase. The Basement, probably no longer exploitable in a convenient way, was finally covered over, making space for new buildings to be erected (pl. 3). This is notably the case for what was labelled as Building 1, built above Terrace I and partially above the backfilled Basement. A radiocarbon date associated with US23, the original floor of Building 1’s room 2c, underlines how its establishment was immediately subsequent to the earlier buildings’ fire and partial obliteration (fig. 22).28 At the moment of discovery, Building 1 already lay in a bad state of preservation, largely destroyed by later occupation and natural erosion, which made the reconstruction of its plan rather unclear, occasionally having to infer it on the basis of small shreds of stone slab paving or isolated stone blocks. The same is true, to a larger extent, for the rest of the site, but it is anyway likely that no edifice actually existed, consistently with what seems to be the case during the HSI phase. Despite its bad state of preservation, a reconstruction of Building 1’s general layout can be put forward (fig. 23). The complex, with a roughly rectangular overall plan, was northwest-southeast oriented, coherently with Terrace I. The focus of the building was likely a rectangular pillared room located in its northwestern part (room 1c), connected with three ancillary rooms on its western (room 4c) and southern side (rooms 2c, 3c). Entrance to room 1c was from the large squared ‘vestibule’ room 5c,29 significantly aligned with the southwestern entrance to the upper part of the site, as is also the case for the short staircase that provided access to room 1c. While room 5c shared the same elevation of Terrace I’s floor in fact, rooms 1c to 4c stood at a lower elevation. The mentioned staircase comprised three low steps, almost 3 m large, built with large stone slabs, and landed in the southeastern corner of room 1c, where a sort of flat stone platform survives (fig. 24). Indeed, the presence of at least another small patch of a flat stones surface in the centre of room 1c suggests that originally it could have been entirely paved in the same way, the ‘platform’ in the southeastern corner rather being a surviving portion of the original pavement. Support for this interpretation is provided by the presence in the southeastern part of room 4c of another area similarly paved; a couple of large stone slab features were also found inside room 5c, that are difficult to interpret but likely represent the remnants of a flagstones floor as well. What distinguished room 1c was the presence of four pillar stone bases arranged on a two per two grid; however, an original three per two arrangement, with two other pillars standing northeast of the surviving ones would appear more consistent with the northeast-southwest extension of the room.30 One of these bases was found set in the platform-like structure located in the southeastern corner of the room. F igure 22 – Probability distribution for US23.

27 A similar raising of the occupational surface likely took place in the central part of the site as well (Area 4), until a substantial re-arrangement of the outer walls was carried out, which marks the onset of the HSIII phase. A correspondence between the different levels in the two areas is anyhow impossible to trace, given the absence of stratigraphic links. 28 All dates presented throughout this book have been re-calibrated using OxCal 4.3.2 software, on the basis of the most recent available calibration curve – IntCal 2013 (Reimer et al. 2013). 29 Roughly 5.4x4.4 m. 30 Roughly 7.2x9.3 m.

michele degli esposti , chiara condoluci

44

F igure 23 – Reconstruction of Building 1’s possible layout, including the hypothetical southeastern pillared room (drawing by P. Kruklidis).

From room 1c, a passage in its southern corner led to a irregularly trapezoidal structure,31 divided into two adjoining premises, room 2c and room 3c, by a central partition wall, M22. Room 3c was most probably completely separated from the area to the southeast of it by the mud-brick wall M24, although the latter was found severely damaged and did not physically reach M28, the western perimeter wall of the two small rooms. Another passage opened in the western wall of room 1c, M149, giving access to the elongated room 4c.32 On the opposite side of room 1c, parallel to its perimeter wall M1, the presence of wall M2 could suggest the existence of another elongated room (6c?), although this area could also be interpreted as a rectangular recess framed by walls M1, M4 and M250. Room 5c’s southeastern limit was likely roughly aligned with wall M24, the one closing room 3c from the southeast. Beyond this line, the plan of the building becomes more vaguely definable. Four large, isolated stone bases (p1, p3, p5) were discovered that, together with a few other supports, could suggest the presence of pillars in this space too. The presence of at least one of such pillars is witnessed by the circular impression of a column shaft clearly vis-



4.5x4.5 m in size overall. Roughly 7.6x2.7 m.

31 32

2 . husn salut in the iron age

45

F igure 24 – The short staircase linking rooms 1c and 5c, with the surviving part of the flagstone floor that likely covered the whole room 1c.

ible on the northernmost stone base (p1). A fourth pillar (p4) could indeed be sitting on top of a low buttress located at the corner between walls M24 and M23, outside room 3c, while a similar feature standing close to feature SF21 could be the support for another pillar, p6, aligned with p4 and p1. Another flat stone – p5 – was located east of this first possible alignment, standing on the same line as p2 and p3. Two parallel rows of pillars can thus be hypothesized in this area; if one considers that the standard suitable length for a palm trunk used as a horizontal roof beam is usually estimated around 2.5-3 m,33 the original presence of another three pillar bases, now lost, should be envisaged. The presence of yet another isolated large stone that could tentatively be recognized as an additional base, p7, could indicate that a more complex pillar pattern existed; however, in the extreme dearth of reliable evidence, any further reconstruction would only be speculative at best. This possible column arrangement, aligned with Terrace I’s orientation, is suggestive of the presence of a columned access to room 5c, possibly the main access to Building 1 as a whole. Almost against the corner between M24 and M28, leaning against the mud-brick platform SF49, stood a peculiar structure realized using stone slabs and mudbricks, SF21. Characterized by two adjacent circular pits, it comprised a possible small oven, inside which a complete bronze vessel was found – a cauldron with two vertical loop-handles –, and a storage jar-housing (fig. 25). Its connection with the preserved Building 1 layout is not certain, but the possible presence of the discussed pillared area would be consistent with its location within a well-ventilated space. The remaining part of Terrace I, although discovered lying in a bad state of preservation, was likely an F igure 25 – The copper-base cauldron discovered inside open space as it was during the HSI phase. structure SF21.

33 However, a recent visit by the authors to the restored Bahla fort, where traditional materials and techniques were used, showed that even longer trunks (up to roughly 4 m) can be used as floor/roofs support. Unfortunately, there was no opportunity to make an enquiry about their local or imported provenance.

46

michele degli esposti , chiara condoluci

One last feature that deserves mention in relation with Building 1 is a single flight of stairs encased between two lateral supporting stone walls, discovered to the west of room 4c, going from its perimeter wall M149 towards the edge of the main outer wall (fig. 26). As the steps seem to climb over M149, they could be recognised as later than, at least, Building 1’s original layout; however, they surely gave access to room 4c, and the resulting gap would actually be minimal, consistent with a single step’s depth. Moreover, at the opposite, lower end, these stairs land above the remains of another large flagstone floor. Although investigation of this area was made difficult by massive erosion and collapse, to the extent that no safe context was identified, this particular paving, characteristic of several Building 1 rooms, suggests that an additional one possibly existed, located at a lower elevation and remarkably placed on the outer edge of the outer stone wall (room 7c?).34 Although strongly impacted by the same fire that hit the Basement, the so-called Burnt Building remained in use (see also stratigraphic details in Chapter 3). The northern end of the corridor room 6b, where it was delimited by wall M136, and the squared room 5b, were backfilled with mud-brick lumps and loam deposits (US389, US392). Among these, one specific deposit laid down within the former perimeter of room 5b, US75, deserves specific mention as it can be interpreted as a ritual deposit, by virtue of the exceptional assemblage of cultural materials collected; as such, it is discussed in Chapter 5. A radiocarbon date coming from US75 (fig. 27), considered together with the associated materials that include possible Iron Age III items (see pl. 76/5-8), place its deposition between the HSII and HSIII phases.35 This structural re-arrangement is all the more significant as it originated one further compartment that went to form

F igure 26 – The late, single-flight stairs climbing over the outer wall and entering room 4c (possibly HSIII phase).

34 As HSIII phase appears to be mainly concerned with a re-arrangement of the peripheral parts of the site, along the outer wall, it cannot however be excluded that these stairs belong to that phase, although this must remain a hypothesis. 35 Although at first the possibility that US75 derived from the same large fire that marked the end of the HSI phase was considered, this hypothesis has clearly to be rejected as the date obtained for US75 is much later than those available from fire-related contexts (see Chapter 10).

2 . husn salut in the iron age

47

a northwestern addition to the large rectangular mud-brick platform SF49, already standing between the Burnt Building and the open area above Terrace I, now largely occupied by Building 1. The central compartment forming the original structure of SF49 was also completely levelled in this moment. Despite this, the platform, now L-shaped, was not provided a flat surface: in the space formerly occupied by room 5b itself, between the end of wall M49 and M42, a large and deep depression was left, as indicated by its laminated filling US72, clearly deriving from water decantation and thus mirroring exposure to weather agents.36 This space was later filled with two compact deposits (US61, US65) capped with mud-brick rows F igure 27 – Probability distribution for US75. – possibly during the HSIII phase. The central-pillared room 7b also continued to be in use and, as said, the laying down of its flagstone floor – and thus the erection of pillar M48 and the construction of the small oven/furnace SF25 – could actually date to the HSII phase, consistently with the discussed, very probable presence of flat stone floors in at least three rooms of Building 1 (1c, 4c, 5c). Moreover, this possibility would suggest that, whether the Burnt Building as a whole remained an independent building or not, this room should better be interpreted as connected with Building 1, rather than with the Burnt Building, at least for this later phase in the site’s evolution. This is all the more likely considering the presence, in the area in front of room 7b, of an extension of its flagstone pavement, part of which has been unearthed beyond the northern end of wall M56. In correspondence with the edge of the underlying substructure wall M140, this floor also shows the presence of a stone-clad step. Finally, the space between the Burnt Building and Building 1, namely the L-shaped space that during the HSI phase was possibly open, was in a later moment occupied by the reinforcement of wall M41b (M41), although this could have taken place during the HSIII phase as well.

Husn Salut III While the western and southwestern part of the main perimeter wall seems to have remained largely unaltered throughout the whole Iron Age life of the site, the HSIII phase is marked by the construction of the outer stone wall around the eastern and northern part of the site, which has also to be linked to the erection of the main tower that projects onto the plain, and thus to the construction of the stone staircase connecting it with the upper part of the site (pl. 4; see fig. 63). The construction of the main tower also provided the site with a new entrance at a lower elevation than the one already in use during the HSI phase. In fact, although the passage through wall M39 that is currently visible is likely to be linked to an Islamic re-arrangement (see Chapter 11 and fig. 116), the presence of a clear path leading to that same point, defined by a series of stone walls built on the southern slope of the hill of Husn Salut, strongly suggests that an entrance existed in the same spot at an earlier date. This relevant modification of the outer walls happened sometime around the 7th/6th century BC at the least; since no radiocarbon date is available for the associated contexts, such a date is indicated by the appearance in the related stratigraphy of the so-called Burnished Maroon Slipped Ware (BMSW), a new pottery type notably connected to the final part of the Iron Age,37 also discovered in late Burnt Building contexts, as well as in the disturbed layers that buried Building 1.

36 This layout and composition closely resembles those of US465, identified during the excavation of another ‘compartment’ in Area5 (see Chapter 5). The precise moment of this final intervention cannot be determined. 37 Magee 2005a. Fragments of such pottery were discovered also inside M138 itself.

48

michele degli esposti , chiara condoluci

The Burnt Building’s layout also underwent substantial changes during the HSIII phase. The earlier corridor room 6b still functioned as the central node of the building, but to its west the erection of wall M138 cut the northeastsouthwest extension of rooms 1b, 2b, 3b and 4b (fig. 15). During this phase, room 7b lost most of its functionality and became rather a peripheral, dumping zone. The flagstone floor was largely covered by a thick ash accumulation, as well as by additional compact loamy deposits (US99), that went to constitute a new surface, whose date actually cannot be safely ascertained. After the early oven/furnace SF25 had ceased to be used, a small tannur was built against the south side of pillar M48, cutting through US88 (SF22).38 A handled lid of the type usually connected with storage jars was found near the circular shaped mouth of the oven, probably used to close it while cooking (fig. 28). SF22 was probably in use for a short time, later buried by further dumping layers. Among these, the soft ashy loam deposit US86/87, mainly located against wall M56, was characterized by the abundant presence of burnt and disarticulated animal bones and scattered date stones, mixed with potsherds and other objects; this may suggest it represents the garbage coming from activities carried out nearby, possibly connected with cooking. At a later moment during the HSIII phase, one last significant change occurred in the layout of the Burnt Building, when the new mud-brick wall/platform M13039 was erected above the southern corner of wall M117. Indeed, there is no sufficient evidence on the ground to ascertain whether the Burnt Building was still in use at this moment. This addition, that at least implied a change in the circulation to and from the building, is distinguished by a different orientation from walls of the earlier phases (fig. 29). In the absence of other features that could be contemporaneous with M130 it seems incautious to envisage a real third architectural phase for the site,40 although it clearly cannot be definitely rejected.

F igure 28 – The small tannur SF22 with the probable closing lid laying beside.

38 Diameter of its mouth 38 cm. 39 Preserved dimensions 3.10x2.70 m. 40 Even hypothesizing a contemporaneity of M130 with the stairs leading to room 4c, this would not build the picture of a real, extended renovation of the site, of a kind that could be interpreted as a new architectural phase.

2 . husn salut in the iron age

49

F igure 29 – Wall M130, set with a different orientation from earlier walls.

T he I ron A ge

material culture : general remarks

Ten years of work at Husn Salut literally overwhelmed the mission’s store with archaeological materials, and pottery representing by far the most part of it, as is usually the case for archaeological excavations. Processing such a huge mass of materials has been an extremely demanding task, also considering the prolonged, yearly presence of the IMTO team on the field, which threatened to transform such work into a sort of tantalization. As far as a comprehensive study is concerned, it was thus opted to primarily tackle the materials coming from a selected sequence, the one that could provide the most interesting data from the point of view of clarifying the typological evolution of the site’s ceramics and stone vessels, as well as their correlation with the Iron Age chronology of South East Arabia. Once this choice was made, further excavation was preferentially targeted to complete the investigation of that sequence, although this did not prevent the investigation of several other relevant contexts. This main stratigraphic sequence is the one excavated within the so-called Basement, to which the one excavated within the Burnt Building has been later connected. These contexts, and the associated materials, are presented in full detail in Chapter 3. As mentioned above, the Basement came to a complete obliteration around the 11th century BC. However its use can be associated exclusively with the HSI phase. Although several stratigraphic phases were evident within HSI sequence (see pl. 18), no internal sub-phasing was possible and no differences were noted among the ceramics from each of the strata overlaying the different floors. Other significant contexts of this early phase were investigated during the excavations devoted to the study of the site’s architecture and building technique, and will thus be accounted for in Chapter 5, to avoid repetitions. Above the backfilled Basement, Building 1 was erected, most likely surviving until the abandonment of the site. Therefore, the materials discovered in the associated contexts can be added to the Basement’s sequence, of which they represent the latest part,41 dated to the HSII and HSIII phases. However, these materials have not been studied in the same detail as those from the Basement and the Burnt Building yet. Hence, at the end of this chapter, only a selection will be presented.



This relation was also physically existent in the northwestern part of the site.

41

50

michele degli esposti , chiara condoluci

Sealed contexts belonging to these later phases were extremely rare at best; however, pottery and stone vessels that, on typological bases, can be unmistakably dated to the Iron Age III period were collected from numerous spots, and support the proposed chronology for the main phasing of Husn Salut as described above. Among the rare unmixed contexts, some of the fills excavated between the outer stone wall belonging to the HSIII phase and the perimeter mud-brick walls built during the HSI phase (see Chapter 4) deserve particular mention, although the thorough study of the associated material has to be completed. In fact, the presence of diagnostic Iron Age III material inside them for the first time provided a safe means for dating the extensive re-arrangement of the site’s outer wall to its late period, strongly indicating that this has to be recognized as being much more considerable and longer lasting than previously envisaged. By way of a general point, the materials collected at Husn Salut perfectly fit the standard repertoire so far identified for the South East Arabian Iron Age. At the same time, it will be outlined in Chapter 10 how it is the materials’ absolute chronology that opens the way for a re-definition of the internal periodization of this Age, at least for central Oman. Pottery belonging to the HSI and HSII phases is generally handmade; occasional pieces occur that, by virtue of the presence of visible marks on their interior surfaces, suggest the use of some rotary device, possibly a slow wheel, for their finishing.42 For the sake of simplicity given the huge amount of potsherds collected, after a first, more detailed subdivision of the pottery fabrics attempted during the early years of excavation,43 it was chosen to simplify it further, to the extent that it finally comprises only three broad groups: coarse, medium and fine wares.44 As often noted on coeval sites in the region, the coarseness of the fabric is mostly linked with the pot’s dimensions and, specifically, with the walls’ thickness. Overall, coarse and medium fabric represent the majority of Husn Salut assemblage, mirroring the widespread presence of medium and large open bowls, large storage jars and smaller, high necked jars. Spouted jars represent a variant of the latter, but they generally come in a slightly finer fabric, although often not fine enough to be included in the ‘Fine’ group. The large majority of fine wares is represented by small cups, in most cases distinguished by the presence of a carination around mid-height of their body. One point that can be made is the paucity of multi-carinated (or wavy walls) cup/bowls, which somehow seem to be characteristic to the al-ʿAin area, attested as they are both at Rumeila45 and at Hili 2.46 This shape however appears to be of little chronological significance, as shown by its abundance in Iron Age III levels at AM-32, a funerary context in al-Madam,47 and by its occurrence, yet rarer, in period II contexts at Rumeila itself. 48 The presence of similar cups in two of the most noticeable ‘foundation deposits’ discovered at Husn Salut, US298 and US454, is reported in Chapters 4 and 5. Fabric colours range from dark brown to red-orange, with some light brown-yellowish and whitish specimens. Most of the pots display rather even exterior surfaces, and interior as well in the case of open shapes; distinction between an actual slip or the outcome of more or less accurate smoothing (self-slipped vessels) is often difficult. Surface colour varies consistently with that of the fabric, although a higher proportion of red and pale red-orangish surfaces is recorded, alongside a notable presence of brown and dark brown, almost blackened ones. The presence of rare sherds of incised grey pottery has to be pointed out. Three of them were found in US454 (fig. 86; pl. 74/1-3), thus providing them with an associated radiocarbon date that connects them with the early phase of the site (see fig. 79). This chronological indication is somehow in accordance with an association with Iron Age II contexts suggested on the basis of the admittedly limited available data from Rumeila, where only 7 samples were

42 A specific technical study of Husn Salut pottery has not been carried out so far. Related observations must therefore be considered as working hypotheses. 43 Cf. Iamoni 2009. 44 In the illustrated potsherds’ descriptions the first indication is the fabric coarseness group, abbreviated in C, M, F, and possible combinations. 45 Boucharlat and Lombard 1985: pls 45/2-5, 7, 10, 13, 15; 51/1-2. 46 al-Tikriti 1998: fig. 8/a-c. 47 Benoist and Del Cerro Linares 1998: fig. 9. 48 Boucharlat and Lombard 1985: pl. 51/1-2.

2 . husn salut in the iron age

51

discovered exclusively in early layers, while they are absent in Iron Age III contexts.49 Parallels for this peculiar type of pottery are not abundant, but are found both at settlement sites and in funerary contexts, and their distribution remarkably covers the whole Oman peninsula.50 A non-utilitarian destination for this pottery, and a possible preferential association with burials, could be reflected in the abundant documentation available from the long tomb at Sharm, where this pottery has been grouped under the label of ‘imitation soft stone’ ceramics in order to include the presence of red-slipped examples, a so far unique occurrence to the authors’ knowledge.51 Painted decoration is attested, mostly limited to small cups, medium size bowls and spouted jars, suggesting a particular significance of these shapes. Consistently, an uncommon concentration of painted vessels, comprising small cups, necked and spouted jars, tube-spouted globular jars, comes from US454, one of several deposits found buried inside the compartments that form the sustaining substructure of the massive perimeter wall of the site, and that is characterised by the presence of abundant dumped pottery (see Chapter 5). Most common painted patterns, again, mirrors the known repertoire for this region and period. Usually painted in black, with numerous occurrences of red and dark red paint, the decorations comprise geometric motifs (wavy or straight lines, hatched triangles, ‘running dog’ patterns, dotted circles, etc.) that are found both isolated and combined in different ways, with a certain recurrence of a star-like shape or asterisk on the bottom of small cups (on interior). Painted pottery represents less than 2.5% of the total drawn pieces from so far analysed contexts; since also nondiagnostic pieces were drawn if they bore such a decoration, this means that this small percentage is also overestimated. Nevertheless, its significance is enhanced by the presence of a painted sherd in one of the earliest foundation fills of the Basement’s first layout (US557, pl. 18/2). This picture is anyhow in accordance with what was reported for other sites, with the outstanding exception of Lizq with its 15% of painted sherds.52 If at al-Madam for example painted pottery is generally said to be present “in faible quantité”,53 at Rumeila it accounts for 0.8% of Period 1 assemblage only, to decrease “considerably” in period 2.54 Incised decoration is also witnessed, usually in the form of herringbone or ‘chevrons’ patterns, on long-handle bowls handles (see below) or on the raised cordons that often decorate large storage jars. For the latter, cross-hatched incised decoration is also widely attested. Incised, parallel wavy lines (‘combed’ decoration) are conversely quite rare. Incised decoration accounts for 7% c. of non-storage ware, i.e., not including the large jars with cross-hatched cordon.55 A different type of incised decoration is represented by a series of graffiti that were engraved on the pots after firing, in some cases spoiling the pre-existing painted decoration. These are usually found on simple and carinated cups’ bodies and bottoms or on larger cup rims, when they are flattened and have an almost horizontal lip. Several examples are known also from other sites.56 Overall, there appears not to be a standardised set of these graffiti, and this has clearly to be related to the fact that all of them were executed in an unskilled way after firing, thus excluding a relation with their production, and are rather to be explained as connected with the owner’s intention. A noteworthy exception is the image of a fish on the bottom of a small carinated cup from US269, which, though still quite basic in comparison, brings to mind the incised fish found on a jar wall from Building G at Rumeila, dated to the site’s period I (fig. 30).57 Another fish was found incised on the outer wall of a small carinated cup discovered inside the Burnt Building, part of a secondary decoration that went to replace the original, painted one (see fig. 47).

49 Benoist 2000: 147; Boucharlat and Lombard 1985: 55, 58, pl. 50/4-9. 50 For an overview see Benoist 2000: 315; Barker 2002: 49-50. 51 Barker 2002: 49, 51. 52 Kroll 1998. 53 Benoist and Del Cerro Linares 1998. 54 Boucharlat and Lombard 1985: 54, 58. 55 A remarkably abundant number of incised jar’s walls was discovered at the site. However, the reported figure is based on diagnostic, drawn pottery, among which these ubiquitous fragments are not included. 56 E.g. al-Tikriti 1998: fig. 2; Benoist 2000: figs 106-108. 57 Boucharlat and Lombard 1985: pl. 52/2.

52

michele degli esposti , chiara condoluci

Curious was the discovery of a shell fragment that also bore an incised representation resembling a fish (pl. 10/8); as it came from a late pit, it is unfortunately impossible to date it with certainty to the Iron Age occupation of the site. Based on their particular recurrence in pillared buildings discovered in the region (although this is not the case for Husn Salut), several tentative interpretations were proposed for these graffiti, including ex-voto and (elite) meeting commemoration.58 Be that as it may, that small, simple and carinated cups in themselves were endowed with some specific value is witnessed by their frequency in some of the most characteristic foundation deposits discovered at Husn Salut, as mentioned above. Apart from the mentioned applied ridges on storage jars and large basins, relief decoration is mainly represented by cordons in the 5 cm shape of crawling snakes’ bodies, rarely coiled, that are found on large jars walls and rims and on smaller vessels as well, with cases of relief snake representations on the interior of open shapes.59 F igure 30 – Incised fish decoration on the A distinct type of pottery vessel found at Husn Salut, usually bottom of a small carinated cup from US269. made in a coarse or coarse/medium fabric, is what has been defined ‘long-handle bowl’, that is, a small bowl protruding from which is a long straight handle (pls 5-6). The latter is most commonly decorated, as specified below. The same shape, although well-known, does not appear to have been found at the majority of Iron Age sites excavated in South East Arabia. When it was found, it was referred to as brazier, censer or incense-burner. At Husn Salut, where its frequency appears far greater than on other sites, it was opted for a more neutral label, although it is true that numerous examples bear evident traces of burning taking place inside them. Most notably, the handle and, less frequently, the side and rim of the reservoir of long-handle bowls are often decorated with an applied snake, or snake scales can be found represented on the handle in the form of incised dots; in some cases the body of the snake rises towards the rim of the bowl (pl. 6/1-6). Not rare is the indication of the snake mouth at the end of the handle, with cases in which the representation shows an undeniably anthropomorphic nature60 (fig. 31; pl. 6/9; see also fig. 56). The distal extremity of the handle can occasionally host plastic representations that may be remindful of different animals. Admittedly, the original interpretation of Cl86 as a ram head (pl. 6/7) could be somehow disputed despite the existence of possible, but later, parallels,61 as it could also be an unskilled portrait of a horned viper (Cerastes gasperettii gasperettii), a venomous subspecies still found in the Arabian Peninsula. Conversely, the horns shown on the handle’s end of Cl15, coming from a Basement context (pl. 33/3), are characterised by an oblique hatching that likely represents the corrugations of a ram’s horns. This particularly complex example is also decorated with two high relief fishes on the handle, located at both sides of a high relief axe whose head is unfortunately largely broken. More complicated is the interpretation of the configured pommel of lid Cl6 (fig. 32; pl. 7/1), where a horned animal’s head shows a decoration of incised small circles on the muzzle, while the horns are characterised by hatched decoration similar to that seen on Cl15. It is tempting, therefore, to suggest the representation of a fantastic animal with merging features of different real ones, something witnessed for example by a well preserved figurine from Bithnah that shows a combination of camel and, significantly, caprin and snake characteristics.62



Benoist 2010b: 129. Pottery decorated in this way has recently started to be referred to as ‘snake ware’ (i.e. Karacic et al. 2017). 60 Cf. Condoluci 2015a. 61 Condoluci 2015a: 201. 62 Benoist et al. 2012a: 158 and fig. 18. 58 59

2 . husn salut in the iron age

5 cm

53

5 cm

F igure 31 – A long handle bowl’s handle from US678 (Cl279) showing evident anthropomorphism. F igure 32 – Configurate lid Cl6.

Another unique example came to light during recent restoration works that also involved partial removal of collapsed debris. Amid the latter, a long-handle bowl handle was found, clearly portraying a bovine with large nostrils and evident brow ridge (pl. 6/8). Moreover, above the brow ridge, along the remarkably short handle, the horns were represented in high relief, decorated with a motif of incised crosses. When relief representations on pottery are found, different from the snake, they often are unique specimens, and as such their interpretation can be difficult. This is notably the case of a sherd coming from a massive accumulation connected with the outer wall collapse (US521). The sherd, Cl218, shows a sort of straight sided pi with an additional horizontal line at mid height and two oblique ones linking the lower ends of the pi to the centre of the middle horizontal line. Overall, the image could recall a stool, or a covered, light material structure (an altar?) or shelter (pl. 7/2).63 The high relief of a bull was found on another sherd coming from a substantial collapse deposit at the foot of the staircase connecting the main tower with the upper part of the site (fig. 33; pl. 7/3). The fabric and surface treatment leave little doubt about its Iron Age date,64 despite the style of the representation strikingly remindful of Bronze Age Harappan representations as widely attested, for example, on stamp seals. Animals are also the objects of a few examples of coroplastic discovered at the site. Leaving aside too fragmentary pieces from Husn Salut and from surveys in the surroundings, and a small bird (Cl22, pl. 7/4) unfortunately coming from a late, mixed context, the remaining examples can all be interpreted as camel representations. This is most clearly the case for T2, possibly the conformed handle of a vessel, where the head and neck are visible (pl. 7/5); it is also rather certain for T1, that shows the torso, part of the hind paws and part of the tail (pl. 7/6). Also T3, although initially interpreted as a female head,65 could more likely represent the central part of a camel’s body, including a saddle, above which a series of short incised traits can indicate either a decoration or thick stitching lines (pl. 7/7).66 A more fragmentary piece of camel figurine was found in US454, one of the ‘foundation deposits’ identified at the site (see pl. 77/9), and another one came instead from the surface collection (pl. 7/8).

This shape only vaguely recalls a series of signs that are found on potsherds usually dated to post 300 BC; however, these are incised postfiring and are usually seen in connection with pseudo-alphabetic signs or South Arabian letters (e.g. Schreiber 2005: fig. 10). 64 Possibly Iron Age III on stratigraphic ground. 65 Cf. Condoluci 2015a: 204. 66 However, this interpretation just remains a plausible hypothesis, since the part of the fragment where the hind paws would be attached appears to be remarkably thinner than in other known examples. 63

michele degli esposti , chiara condoluci

54

5 cm

F igure 33 – High relief decorated potsherd showing the profile of a bull. F igure 34 – A perforated clay object from Husn Salut, possibly a lid/lampshade or a device for water-flow measurements (tasa).

10 cm

This repeated occurrence is not surprising, as full relief, clay camel representations are well attested in the region during the Iron Age period,67 probably as a consequence of further acquaintance with this animal deriving from its domestication, for which a date around the beginning of the first millennium BC is now widely accepted.68 Unique is instead a painted human clay figurine collected during a survey in the surroundings of the site (pl. 7/10), although clay human representations are known from other sites, often with a similar painted decoration.69 Also from the survey came the figurine of a quadruped that seems more likely interpretable as a bull (?) rather than a camel (pl. 7/9). Another unique find at Husn Salut is represented by a fragmentary, perforated ceramic plate, most likely originally cylindrical in shape. Open at both ends, it clearly cannot be interpreted as a vessel fragment (fig. 34).70 A number of perforated ceramic objects have been discovered on Iron Age sites of the Oman Peninsula, which are interpreted as lids associated with large incense burners or braziers.71 This function was truly confirmed for a goblet-shaped example from Masafi 1 by the discovery of the associated brazier,72 and can be confidently assigned to a domed example from Muweila,73 to another more articulated specimen from the pillared hall at Bida Bint Saʿud,74 and to two flat ones from Bithna and, again, Muweila.75 At Rumeila, a more fragmentary object was found, that was finally interpreted in the same way.76 Overall, none of these published examples convincingly compares with the one from Husn Salut. Although the piece from Husn Salut could well have the function of a lampshade, an alternative, intriguing explanation

Among the best preserved examples in the Oman peninsula are the figurines from Muweila (Magee 2007a: figs 25, 30), Rumeila (BouLombard 1985: pl. 65/4-10) and Masafi 1 (e.g. Benoist 2010b: fig. 9/2; Benoist et al. 2012b: fig. 18), still bearing evidence of the original painted decoration. 68 Cf. Magee 2014: 204-213. 69 E.g. Benoist 2010b: fig. 9/2 (Masafi 1); Benoist et al. 2012b: fig. 5/9,15 (Bithna). The figurine from Salut finds also fitting parallels in one collection of the National and Military Museums in Sanʿāʾ (Yaseen and Shargabi 2007). 70 Perforated pots are known for example from Ras al-Hadd HD9 (Benoist 2000: fig. 259/4-7). 71 E.g. Benoist 2010b: 126. 72 Ibid., and fig. 9/1. 73 Magee 2001: fig. 11. 74 al-Tikriti 2011: fig. 87. 75 Benoist 2010b: fig. 5/1; Magee 1999: fig. 7. 76 Boucharlat and Lombard 1985: pl. 51/6; Benoist 2010b: 126. 67

charlat and

2 . husn salut in the iron age

55

was proposed by A. Benoist in her dissertation, which may be worth recalling here, although it must for the moment remain a working hypothesis until further research is possibly undertaken on this issue. A remarkable resemblance can in fact be noticed between the object from Husn Salut and the description of traditional objects used for measuring water flow from irrigation channels, which in the case of Oman are known as tasa.77 Considering other characteristic ceramic items, architectonic elements can likely be recognized in a few examples of right-angled, straight-sided, U-shaped sherds whose function is unknown but which can tentatively be interpreted as gutters (pl. 7/11-12). Published parallels are limited to one closely fitting example from Muweila,78 although an apparently similar one has been described from Hili 2.79 Different kinds of drainpipes, with a circular section, are however known from Bida Bint Saʿud, Hili 17 and Rumeila.80 Other specimens akin to those from Husn Salut have been discovered during the initial investigations at the connected settlement of Qaryat Salut. From the HSII phase and later contexts, namely US23 and the overlying US3, two possible nozzles or drains were collected (pl. 8/6-7). Together with other more or less fragmentary examples of the same object – unfortunately coming from mixed contexts, though all from Terrace’s I area – they display recurrent features (pl. 8). Of the original vessel only the drain itself is generally preserved, which was attached to the bottom of a large jar’s wall. The thick spout, almost circular in section and with a flat base, has a central hole more or less inclined outwards. Some of the sherds still comprised part of the jar’s wall where one or more small vertical loop handles are present. A peculiar specimen (pl. 8/4) is distinguished by a thick almond-shaped lug through which the possible nozzle hole runs. These drain/nozzles from Husn Salut are generally made in a coarse/medium fabric, occasionally red slipped, with abundant grits. A comparison is known from Mleiha,81 where hypothesis was advanced that the corresponding vessel could have served as a portable forge. Although suggestive, this interpretation is most likely to be discarded for one main reason that also applies to the Husn Salut’s examples, that is, they show no trace of burning, not to mention possible adhering slag. Likewise, a simple use of these containers as water tanks could be proposed, despite no waterproofing treatment was applied. Two other well preserved, comparable vessels (without the loop handles) have been found in two graves at Samad as-Shan,82 while one comes from Izki.83 With the transition to the HSIII phase, the pottery assemblage remains largely identical to that of the earlier phase. Among the previously predominant shapes, only spouted jars and carinated cups seem to be less frequently represented; painted pottery also becomes rarer. However, the main feature that distinguishes late contexts from the HSI and HSII phases, is the presence of a type of pottery previously absent (pl. 9). This pottery is characterized by a fine/medium fabric that is usually light brownbuff, sometimes reddish, and by its biunivocal association with a range of shapes that are not found in earlier layers. When non diagnostic fragments are collected, from which the shape cannot be inferred, the attribution to this pottery type is indicated by the presence of a distinctive burnishing of the slipped surfaces, that can vary in colour from dark brown to red, with darker burnishing strokes. Although different sub-groups in all likelihood exists among the collected assemblage,84 as a whole this pottery can be safely identified with the so-called Burnished Maroon Slip Ware (BMSW), first pointed out by P. Magee85 and of the utmost relevance from the perspective of dating the associated levels (further discussed in Chapter 10). Indeed, apart from the ‘drains’ just mentioned, other shapes were discovered

77 Detailed descritpion and a concise discussion are given in Benoist 2000: 351-353, and will not be replicated here. It may be of some significance that J.C. Wilkinson (1977: 107) reported the use of one of these devices from the oasis of Firq near Nizwa, thus not far from Salut. 78 Magee 2007a: fig. 39. 79 al-Tikriti 2001. 80 al-Tikriti, al-Haj and al-Niyadi 2001; al-Tikriti and Haddou 2001: fig. 9; Boucharlat and Lombard 2001a: fig. 74. 81 Ploquin et al. 1999: fig. 3 and pl. 3. 82 Yule 2001: pls 149/3; 423/11. 83 Schreiber 2007: pl. 38/4. 84 A more thorough study of the Husn Salut III phase pottery, starting from the isolation of all the contexts containing safely datable Iron Age III sherds, has just began within the frame of the renewed investigations at Qaryat Salut. 85 Magee 2005a.

56

michele degli esposti , chiara condoluci

5 cm

from late contexts that could hint to an even later date for the final, likely residual and limited occupation of Husn Salut; these are further discussed in Chapter 3. To the HSIII phase can also be dated two richly decorated fragments that are to be recognized as the legs of two cuboid incense burners (fig. 35; pl. 10/1-2). Both objects, when complete, likely measured around 10 centimetres in height, with sides ranging around 7 centimetres. Both were made in a light brown fabric with abundant small and medium dark brown-black grits, overall consistent with the common pottery found at the site; the decoration is made by deep incisions and impressions which together produce a sharp, excised pattern. While the stratigraphic provenance is only vaguely indicative of their late date, as they both come from late, reworked contexts, an Iron Age III chronology can be supported by the known parallels. In the Oman peninsula, the local ancient roots of aromatic burning are actually testified by a rectangular incense burner on four legs discovered at Ras al Junayz (Ras al-Jinz) RJ-2, at the easternmost end of the Arabian peninsula, in an Early Bronze Age context dated to around 2250 BC.86 More abundant and fitting are however Iron Age comparisons. One cuboid incense burner was discovered at Rumeila in contexts belonging to the site’s Period II,87 that is, dating to the Iron Age III period on the basis of associated pottery. Three other, unpublished examples have been excavated from Iron Age III contexts at Rafaq in the Wadi al-Qawr (southern Ras al-Khaima, UAE), while one was discovered in a multi-period long chambered tomb at Daba, at the border between Oman and the U.A.E, that also comprises Iron Age materials.88 Leaving aside the numerous specimens coming from the wider Middle-Eastern area and spanning a longer chronological range,89 a consistent chronology is shared by a cuboid incense burner found at Qalʿat al-Bahrain in a context dated between 600 and 400 BC.90 However, most examples from the Arabian Peninsula are later, including one found at Mleiha,91 one from Thaj,92 and several from South Arabia.93 It must be underlined that none of the specifically mentioned examples display a comparable excised decoration, it usually comprises a fine incised line and, in some cases, a painted surface.

F igure 35 – Fragments of two ceramic cuboid incense burners from Husn Salut.



Cleuziou and Tosi 2000: 53-54, pl. 33. Boucharlat and Lombard 1985: plate 65/1. 88 C. Phillips, pers. comm. (Rafaq); F. Genchi, pers. comm. (Daba). 89 E.g. O’Dwyer-Shea 1983; Millard 1984; Hassell 2005. Regrettably, only few come from excavated contexts. 90 Højlund 1997. 91 Taha 1974: pl. II/b. 92 Bibby 1973: 20, fig. 11. 93 E.g. O’Dwyer-Shea 1983; Hassell 2002. 86 87

2 . husn salut in the iron age

57

For what concerns non-ceramic materials, complete catalogues of stone vessels and metal objects are presented and discussed in Chapters 6 and 7 respectively. It is just worth underlining here that, while in the case of metallic items no real chronological distinction can be based on their typology, a set of stone vessels was discovered at Husn Salut that can be clearly connected with the late Iron Age phase. Moreover, they actually belong to a typology that is commonly held to date after the Iron Age III period, and as such they either indicate an earlier date for that typology, or a longer duration of the site (see Chapter 10). Several grinding stones and hammerstones were collected from different contexts. While these kinds of stone tool are informative about the conduction of domestic activities, they are of no help for chronological determinations, as they are so widely common at sites of different ages and location. Stone beads come in no large number from Husn Salut. Their study was began by J.M. Kenoyer and D. Frenez, and awaits conclusion. One particular bead deserves mention, as it was made in gold and comes from what has been recognized as a safe HSII phase context, US23, briefly discussed in Chapter 7 (no. 150). A few shell beads were also collected at the site. A number of them simply consists in specimens of cyprea shell to which the back has been cut off. While the majority come from late, mixed contexts, one example was found in US23, a good floor surface of the HSII phase, and thence they are mentioned here (pl. 10/4-7). Although shell objects still await a thorough study, it is also worth mentioning the presence of at least two large shell buttons or adornments of circular shape, S98 and B7 (pl. 10/8-9), the latter from US23 as well, together with a smaller example, Sh16 (pl. 10/10). Parallels to these objects are becoming more and more abundant, as a considerable number has been recently discovered at Saruq al-Hadid94 and in the collective tombs of Daba.95 However, the majority of the known examples is either incised or inlay decorated, and plain examples are scarce.96 Nevertheless, almost all share the same feature of the two Husn Salut objects, that is, a four hole configuration on the underside where three smaller holes are attached to a larger central one. These holes were most likely realized in order to allow sewing the ‘buttons’ to some other material. An identification of the shell used to produce them has not been accomplished so far at Salut; published parallels are usually reputed to be made of shell, and Fasciolaria trapezium has been indicated as the most likely species;97 however, a single sample analysed from the Sharm tomb was revealed to have been carved from a mammal tooth.98 A button of a different kind, Sh24, with a gently curved profile and only two holes in the middle (where it is broken) was found in US87, a context belonging to the late HSIII phase (pl. 10/12). Bone objects are extremely rare at Husn Salut, and only one example comes from a certainly dated Iron Age, sealed context. From US464 in fact, a context linked with the early construction of the outer wall in Area 5, comes B11, most likely to be recognized as half of the grip of some instrument, possibly a knife, as shown by the presence of two pairs of holes meant for fixing rivets (see pl. 65/4). An L-shaped plate, equally characterised by the presence of four rivet holes, was found in US3, and could represent the embellishment for some kind of wooden object (pl. 10/3). A limited number of small stone pendants were also discovered at Husn Salut (pl. 10/14-19). They are of irregular shape with flattened opposite faces on which a sketchy decoration is incised. Their interpretation is not univocal, but a link with an apotropaic function seems plausible, if not just a decorative one.99 Certainly decorative is a soft stone pendant that was obtained through reworking and piercing the fragment of a decorated stone vessel (pl. 10/14). More relevant to the general discussion on Husn Salut’s interpretation (see Chapter 9), is a pyramidal stamp seal unfortunately coming from a washed deposit of Area 4 (fig. 36; pl. 10/20). The seal is made of dark gray soft-stone. It is 1 cm tall, with an irregular squared face measuring 1.5 by 1.6 cm, and it has a slightly rounded apex, horizontally



E.g. Weeks et al. 2017: figs 3 (bottom left), 24. Caputo and Genchi 2016. 96 See also Hartnell and Barker 1999 for other examples. 97 See Weeks et al. 2017: 53. 98 Susino 1999. 99 Cf. Degli Esposti 2015c. 94 95

michele degli esposti , chiara condoluci

58

perforated (diameter 0.04 cm). The impression shows a simple design formed by eight incised and pierced triangles and an incised dot and circle, most likely representing a star/sun motif. Several seals of this pyramidal shape, which seems to be the actual local type, are known from other sites in South East Arabia and are usually considered as the indication of some sort of administrative/commercial system that might have existed, indeed the only indication that would be available so far. Published close comparisons to this seal are known; for example from the Iron Age sites of Rumeila100 and Qarn Bint Saʿud101 in the Abu Dhabi Emirate (UAE), from the Iron Age levels of Tell Abraq102 and from two Iron Age tombs on Jabal al-Buhais103 in the Sharja Emirate. In Oman, a pyramidal seal was discovered in a reused Hafit tomb at Tiwi.104 Still unpublished are the numerous seals recently found at Saruq al-Hadid, that significantly enlarge the corpus.105 Among small stone finds, two hemispherical, pierced items can be interpreted as spindle-whorl, thus possibly bearing witness to weaving at the site, although their use as simple pendants is also possible (pls 10/13; 34/12). Finally, at Husn Salut an exceptional find for ancient basketry has been discovered. The preserved, partially charred remains of at least one woven basket, and possibly a second one, deposited in US75 (discussed in Chapter 5) in fact underline once more how a possibly large quantity of perishable items should always be taken into account when reconstructing the material culture of ancient sites. Besides these rare remains, further evidence for basketry in the Iron Age is provided at Husn Salut by an impression on a storage jar base (fig. 37/a) and by two impressions on mud surfaces, all closely resembling the actual baskets from US75. One such impression was discovered in a filling layer of the western part of the external wall (fig. 37/b), while a second one, already mentioned, stood on the surface of the original floor immediately north of room 7b, between wall M140 and M141 (fig. 20). 1 cm

a

F igure 36 – The pyramidal stone seal found at Husn Salut (S305). b

F igure 37 – a) A woven mat impression found on a storage jar’s base from Husn Salut; b) impression of a woven basket discovered in one of the outer wall’s fills.



Lombard 1988, fig. 2/2-6. Stevens 1992. 102 Potts 1991: 95, fig. 135. 103 Jasim 2012: figs 100, 163. 104 Schreiber and Häser 2004: 323, fig. 6. 105 Boraik et al. 2017. 100

101

2 . husn salut in the iron age

P ottery

and other finds from

B uilding 1 (HSII

and

HSIII

59

phases )

By virtue of its location on the uppermost part of the site, the poorly preserved architectural remains of Building 1 were found below a moderately shallow soil cover, distinguished on the field as US1 and corresponding to the modern surface. Due to extensive erosion of the ancient structures; subsequent washing of the deposits; and above all human activities taking place over more than two thousand years after the site’s abandonment, most of the accumulated deposits also underwent extensive re-shuffling. Therefore, the identification of undisturbed, archaeological contexts associated with the building’s rooms resulted in being occasional at best, and limited to small areas. On the other hand, the elevated position assures that no material was washed in this area, coming from other parts of the site, and the collected artefacts can be reputed to mirror the building’s use, albeit without the possibility of specifying each room’s function. The whole area occupied by Building 1, and covering Terrace I as a whole, was buried beneath the thick, reworked deposits of US3. Moreover, as Islamic period occupation strongly impacted the Iron Age stratigraphy, due in particular to extensive excavation of pits and sunken hut-like structures (see Chapter 11), materials from the early phase of the site also ended up in most recent contexts. Given this stratigraphic situation, it was no surprise that from US3 archaeological material was recovered, whose variety covers almost the whole site’s assemblage. Nevertheless, as the quantity of Islamic sherds was not abundant, and mainly concentrated in the proximity of coeval structures, the association of ‘classic’ Iron Age materials with others that can be dated to the Iron Age III period appears to be sufficiently reliable, and underlines the fact that most of the artefacts classes diffused during HSI and HSII phases continued to be in use in the following HSIII phase, alongside newly introduced shapes and fabrics. Among the few safe contexts, the deposits backfilling rooms 2c and 3c deserve mention, as they appeared to be largely undisturbed. In particular, in room 2c it was possible to unearth the good surface of the US23 floor, belonging to the original layout of Building 1 and above which several potsherd, copper/bronze objects and fragments, were lying (fig. 38). The radiocarbon date obtained for this context (mentioned above, fig. 22) is actually the main evidence that allowed distinction of the HSII phase from HSIII, as its lower limit surely pre-dates the introduction of the diagnostic Iron Age III pottery types in the region.106 The materials from US23 thus represent a reliable, although unfortunately unique, reference collection for the HSII phase, and will be preliminarily illustrated here (pl. 11). They clearly show that there is no change in the material culture between the first two architectural phases of Husn Salut. Following them is a selection of materials from US3, mirroring a mixture of HSII and HSIII phase materials (pls 12-17). As stated above, not all the materials from Husn Salut have been studied in depth yet. Those related to Building 1 are therefore presented here for the sake of providing a preliminary, generic picture of the late phases’ assemblage. Individual description of the sherds are not given, but all conform to the general traits mentioned above for what concerns fabric colours, surface finishing, and fabric coarseness, the latter usually being directly related to the walls’ thickness. Pottery from US23 includes a limited number of shapes, including carinated cups, jars, bowls with inverted rims, one complete jar lid, a storage jar, and the nozzle mentioned above. Fabrics range from fine to medium. Slip is mainly reddish in colour although some sherds have smoothed, plain surfaces. Painted decoration is almost absent. Amid the carinated cups a variant with a more distinct carination and a thicker base differs from the more common types (pl. 11/6-7). A spouted bowl with a triangular, slightly grooved rim, on top of which a decoration consisting of impressed circles runs, seems to have been intentionally decorated with a ‘stained’ painted motif (almost a marbleized effect, pl. 11/10) which finds comparison in two dishes from US3 (pl.16/7, 10) but is unmatched at other Iron Age sites. Jars have triangular out-turned rims, in one case with up-turned lip (pl. 11/13). One sherd with a beaded rim (pl. 11/12) finds comparison from US337 (pl. 75/9). Stone tools such as hammerstones/crushers were also collected, suggesting that some domestic activity could be



106

Cf. Magee 2005a.

60

michele degli esposti , chiara condoluci

F igure 38 – Floor US23 in room 2c, with artefacts in situ.

carried out in the room, possibly grain processing or the like. Besides, the presence of copper-base slag of variable size, together with a large amount of tiny copper-base fragments from several objects and tools (plaques, wires, rings, arrowheads, needles, a shaft hoe, fragmented and also folded completed arrowheads) found scattered on US23 floor, strongly suggest that some activity related to the re-working (re-smelting?) was carried out on this spot. Similar scrap metal was also found in the uppermost layer US3 and in an area immediately to the east of Building 1 (US57). As discussed above, US3 represents a late, largely reworked context; however, a selection of the collected pottery will be presented here, as it further illustrates the overall assemblage of the site and, at the same time, includes shapes that are not known, or not so well attested, in other contexts. Although more study on this issue is undoubtedly needed, some of these shapes (and decorations) may also be an indication of a prolonged occupation at the site (see Chapter 10). Undoubtedly the most represented category in US3 are storage jars, coming in a wide range of shapes, mirrored in the different rims. Simple out-turned rims (pl. 12/1-2) can evolve with a more or less pronounced internal ledge (pl. 12/3-5) or with an outer bulge (pl. 12/6-7). Triangular in-turned or out-turned rims (pl. 13/1-3) can also sometimes be grooved (pl. 13/4-5). Decoration mainly comprises cross-hatched raised cordons, bands of impressed herringbone patterns below the rim or impressed circles on the rim. Ledge-rim jars (pl.13/6), globular jars (pls 13/7-8; 14/1), short necked globular jars (pl. 14/2), craters or deep basins (pl. 14/3-6) are also present together with jar lids with rounded or slightly flared rims (pl. 15/1), which sometimes bear an impressed dot decoration. Cups (pl. 15/2-5) and carinated cups (pl. 15/6-16) – one with an incised graffito possibly representing a boat (pl. 15/6) – are produced in fine or medium fine fabric, in some cases un-slipped. Bowls and basins, mainly with simple flat or nail-headed rims, are also largely present (pls 15/17; 16/1-9, 11-12) including some examples of carination (pl. 16/2), incised wavy line motifs and also painted decorations (pl. 16/5), although the presence of the latter is sporadic in the whole assemblage. A few scattered sherds possibly belonging to BMSW bowls were also found. Two dishes (pl. 16/7, 10) bear the same bizarre ‘stained’ decoration attested on a bowl from US23. The apparent use of a slow wheel for some bowls and jars needs mention, albeit in the absence of detailed technical studies. Jars are mainly represented by the variant with triangular in-turned rim with outer or upper bulge (pl. 17/1-6), although examples of simple out-turned (pl. 17/7) are also present. A single spouted sherd was collected (pl. 17/8) but few other badly preserved spout fragments were noticed during excavation. Some jars were also equipped with handles or lugs (pl. 17/9-10). Decoration on jars is scantily present, represented by incised wavy lines and by painted, nearly geometrical, motifs (pl. 17/12). Apart from a nozzle (pl. 8/6), a stand (pl. 17/11) represents a unicum at Husn Salut.

plates

62

P late 1 – General plan of Husn Salut.

michele degli esposti , chiara condoluci

2 . husn salut in the iron age

P late 2 – Plan of phase Husn Salut I structures.

-

plates

63

64

michele degli esposti , chiara condoluci

P late 3 – Plan of phase HSII structures, including Building 1.

2 . husn salut in the iron age

P late 4 – Plan of phase Husn Salut III structures.

-

plates

65

66

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

michele degli esposti , chiara condoluci

P late 5 – Long handle bowls. F orange. Abundant vegetal temper. Pale red slip. Incised chevrons decoration. US22,7 = Cl48. C orange with grey core. Abundant white and red grits. US1,130 = Cl51. C orange. Large white and red grits. Traces of red-brownish slip. Incised chevrons decoration. Surface,21 = Cl45. C red-orange. Large white and red grits. Traces of red slip. Incised chevrons decoration. Surface,26 = Cl78. M/C pinkish-orange. Medium white and red grits; vegetal temper. Red slip. Incised chevrons decoration. US3,59 = Cl39. C light orange. Abundant grey and red grits. Red slip. US111,1 = Cl62. M pale pinkish. Sparse white and grey grits. Traces of brown slip. US1,124 = Cl43. C pale orange with grey core. Abundant red grits; vegetal temper. Red slip. US1,125 = Cl49. C orange. Abundant red grits. Light-red slip. US128,1 = Cl69. C orange with grey core. Abundant white grits, some large. Traces of dark red slip. US85,1 = Cl50. M red. White and dark grits. Red slip. US29,1 = Cl9. M/C dark pinkish. Abundant white and dark red grits. Red slip. US36,2 = Cl11. C orange-red. Abundant white and red grits. Red slip. US237,1 = Cl84. M red. White and grey grits. Pale red slip. US36,5 = Cl14. M brown. Red and small white grits. US163,4 = Cl73.

2 . husn salut in the iron age

-

67

plates

3 1

2

4

5

6

9 8

7

10

13

11

14

P late 5 – Long handle bowls. Scale 1:3.

12

15

68

1. 2. 3. 4. 5. 6. 7. 8. 9.

michele degli esposti , chiara condoluci

P late 6 – Configured long handle bowls. M pinkish. Small grey grits; vegetal temper. Traces of dark red slip. US236,8 = Cl112. C orange. White and brownish grits. US1,147 = Cl79. C/M orange. White and red grits. Red slip. Surface,24 = Cl72. M red-brownish. Sparse grey grits. Dark brown slip. US 748,1 = Cl292. M orange. Sparse white and grey grits; vegetal temper. Brown slip US1,148 = Cl80. C orange-red. Abundant white and red grits. Brown slip. US256,1 = Cl85. C orange. Abundant white and red grits. Light brown slip US220,2 = Cl86. C/M pale brown-greyish. Abundant dark grey grits; vegetable temper. Dark slip. M54,1 = Cl299. M brown with grey core. Dark grey grits; vegetal temper. Red-brown slip. US698,1 = Cl279.

2 . husn salut in the iron age

-

69

plates

2

1 3

5

6 4

7

P late 6 – Configured long handle bowls. Scale 1:3.

8 9

70

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

michele degli esposti , chiara condoluci

P late 7 – Configured lid; high relief decorated sherds; terracotta figurines; gutters. M red. Small white grits, large grey grits. Red-brownish slip. US1,63 = Cl6. M pink. Black and white grits. Dark red slip exterior, red interior. US521,1 = Cl218. M brown. Red and grey grits. Red slip. US415,2. F light brown-pinkish. No visible grits. US36,10 = Cl22. M/C red with grey core. Red grits; vegetal temper. US3,12 = T2. F pink with grey core. Sparse black grits; vegetal temper. Smoothed. US5,2 = T1. F light buff-whitish. No visible grits. Pinkish-white slip. US23,24 = T3. M/F light red. Sparse small white grits. Traces of red slip. US0,27 = Cl88. M red. Sparse black grits; vegetal temper. Survey Site 7. M/F red. Sparse black grits. Traces of red painted decoration. Survey Site 1. C brown. White and grey grits; abundant vegetal temper. Traces of brown slip interior. US0, 25 = Cl77. C pink-orange with grey core. Red and white grits; vegetal temper. Dark brown slip. US1,123 = Cl38.

2 . husn salut in the iron age

-

71

plates

3 1

2

4 5

7 8 6

9

10 11

12

P late 7 – Configured lid (1); high relief decorated sherds (2-3); terracotta figurines (4-10); gutters (11-12). Scale 1:3, except nos 4-10: scale 1:2.

72

1. 2. 3. 4. 5. 6. 7.

michele degli esposti , chiara condoluci

P late 8 – Nozzles/drains. C orange with dark thick core. Red brown slip exterior and interior. US128,1, Area 3. C orange with dark core. Large amount of grit inclusions, red slip exterior and interior. US163,10, Area 3. C light orange with thin grey core. Abundant red and grey grits. Light red slip interior. US1,438, Area 2. C orange. Red and grey grits. Red slip exterior and interior. US1,435, Area 3. C dark orange. Grey grits; sparse vegetal temper. Surface,503. C red. Small white and grey grits. Red slip exterior and interior. US3,137, Building 1. C grey. Abundant red, grey and white inclusions. Flaked slip on ext. US23,4, Building 1.

2 . husn salut in the iron age

-

73

plates

1

2

3

5 4

6

P late 8 – Nozzles/drains. Scale 1:3.

7

74

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

michele degli esposti , chiara condoluci

P late 9 – Iron Age III pottery from un-stratified contexts. F brown-reddish. Rare small white grits. Red slip exterior and interior, burnished exterior. US44,1. F red. Rare small white grits. Red slip exterior and interior, burnished exterior. US134,1. F brown-reddish. Rare small white grits. Red slip exterior and interior, burnished exterior. US1,202. F brown. Tiny white grits. Red slip, vertical burnished exterior. US1,200. F light brown. Sparse tiny white grits. Red slip exterior and interior, burnished exterior. US163,11. F red. Rare small red grits. Red slip exterior and interior. US41,6. F brown-reddish. Rare small white grits. Red slip exterior and interior, burnished exterior. US44,2. F brown. Sparse tiny white grits. Red slip exterior and interior, burnished exterior. US1,201. F red-orange. Rare small white grits. Red slip exterior and interior, burnished exterior. US168,1. F orange. Rare tiny white grits. Red slip exterior and interior. Wheel made. US53,3. F red-orange. Occasional white grits. Red slip exterior and interior, burnished exterior. Wheel made. US85,2. F/M brown-reddish. Sparse white grits, rare grey grits. Red slip exterior and interior. US101,1. F red-brownish. Occasional tiny white grits. Red slip exterior and interior, burnished exterior. Wheel made. US36,1. F red. Occasional tiny white grits. Red slip exterior and interior. US36,3. F/M red-orange. Sparse white grits. Red slip exterior. US44,3.

2 . husn salut in the iron age

-

75

plates

1

2

3

4

5

6

7

8

9

10

11

12

13

14 15

P late 9 – Iron Age III pottery from un-stratified contexts. Scale 1:3.

76

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

michele degli esposti , chiara condoluci

P late 10 – Incense burners; bone and shell items; pendants and seal. M pale yellow. Abundant small and medium grey grits, rare small white grits. US1,176 = Cl135. M/C light buff. Abundant black grits, some large, rare pink grits. US559,1 = Cl220. Bone plaque. US3,8 = B1. Cypraea shell bead. US23,26 = Sh7. Cypraea shell bead. US1,30 = Sh5. Cypraea shell bead. US1,203 = Sh3. Cypraea shell bead. US3,300 = Sh2. Incised shell plaque. US48,1 = Sh13. Shell button. US1,61 = S98. Shell button. US23,25 = B7. Shell button. US56,1 = Sh16. Shell button. US87,1 = Sh24. Light grey soft stone. US345,1 = S199. Light grey-bluish soft stone. US1, 220 = S369. Dark grey soft stone. M57,1 = S304. Light grey soft stone. US0,37 = S300. Dark grey soft stone. US572,2 = S352. Dark grey soft stone. US383,1 = S314. Dark green-blackish stone. US24,5 = S172. Dark grey-blackish stone. US371,1 = S305.

2 . husn salut in the iron age

-

77

plates

1

2

3

4

5

7

6

10

8

11

9

15

17

12

19

13 18

14 16

P late 10 – Incense burners; bone and shell items; pendants and seal. Scale 1:2, except no. 3: scale 1:3.

20

78

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

michele degli esposti , chiara condoluci

P late 11 – Pottery from US23 (Building 1, HSII phase). F red. Small white and dark grits. Light red slip exterior and interior. US23,9. M orange. Small, medium and large white, red and grey grits. Red slip exterior and interior. US23,10. F brownish. Small grey grits, rare large white and grey grits. Pink-creamy, smoothed exterior, creamy smoothed interior. US23,11. F reddish. Small and medium white grits, rare mica. Pale brown smoothed exterior, reddish brown interior. US23,5. F with grey core. Small white grits, sparse mica. Creamy slip exterior, pale red interior. US23,7. F brownish. Rare medium red and small white grits. Pale red slip exterior, reddish-brown interior. US23,6. F brownish. Rare medium white and red grits. Reddish-brown slip exterior and interior. US23,14. M/F reddish. Small, medium and large red and white grits. Red slip exterior and interior. US23,15+16. M/F pale brown. Small white grits, medium and large grey grits. Pale yellow slip exterior and interior. US23,13. F brownish. Tiny and medium white grits, rare small brown grits. Pinkish, smoothed exterior and interior. Dark grey stained decoration exterior and interior. US23,22. F brownish. Small white grits. Red slip exterior and interior. US23,23. F reddish. Small and medium white and red grits. Red slip exterior and interior. US23,12. M/F reddish. Small, medium and occasional large big white grits. Light red slip exterior and interior. US23,18. F brownish. Small white and sparse small grey grits. Brownish smoothed exterior and interior. US23,24. C red with light grey core. Abundant white and red grits, some large; rare mica. Burnished exterior, light red slip interior. US23,2. C red. Abundant white grits, some large, abundant small and medium grey grits. Dark reddish-grey slip exterior, red interior. US23,1. C red. Abundant white grits, some large; abundant dark grits. Red slip exterior and interior. US23,50.

2 . husn salut in the iron age

-

79

plates

1 2

4

3

5

7

6

8

9 10

12

11

13

14

15

16

P late 11 – Pottery from US23 (Building 1, HSII phase). Scale 1:3, except nos 15-17: scale 1:4.

17

80

1. 2. 3. 4. 5. 6. 7.

michele degli esposti , chiara condoluci

P late 12 – US3: storage jars. M/C reddish. Abundant medium and large red grits, small white grits; sparse mica. Light brown slip exterior, light red interior. US3,2. F/M pale brown. Whitish glaze (?) exterior and interior. US3,14. M pale brown. Abundant small, medium and large grey grits. Grey slip exterior, reddish-grey interior. US3,48. M brownish. Red grits, small white grits. Pale brown slip exterior, pinkish, smoothed interior. US3,205. M pale brown. Medium and small white and grey grits; rare mica. Light red slip exterior and interior. US3,28. M/C reddish. Abundant dark grits, rare white grits. Pinkish slip exterior and interior. US3,78. M light red. Large dark grits, small white grits. Light brown slip exterior and interior. US3,64.

2 . husn salut in the iron age

-

81

plates

1

2

3

4

5

6

7

P late 12 – US3: storage jars. Scale 1:3.

82

michele degli esposti , chiara condoluci

1. 2. 3. 4. 5. 6. 7. 8.

P late 13 – US3: storage jars and globular jars. C pale brown. Abundant small, medium and large red grits, small white grits. Pink, smoothed exterior. US3,207. M brownish. Abundant medium and large red and grey grits, small and medium white grits. US3,89. M red with grey core. Small and medium white grits, medium red grits. Pale red slip exterior and interior. US3,203. M brownish. Abundant small grey grits. Pale red slip exterior and interior. US3,124. M greyish. Abundant small and medium red and grey grits. Pale red slip exterior and interior. US3,188. C red. Abundant grey and red grits, white grits. Pinkish slip exterior, red, smoothed interior. US3,184. M grey. Abundant grey white grits. Grey, smoothed exterior and interior. US3,136. M reddish. Small and medium white and red grits. Dark red slip exterior, red interior. US3,201.

2 . husn salut in the iron age

-

83

plates

1

2

3

4

5

6

7

8

P late 13 – US3: storage jars and globular jars. Scale 1:3.

84

1. 2. 3. 4. 5. 6.

michele degli esposti , chiara condoluci

Plate 14 – US3: globular jars; basins; crater. M light red. Small white grits, medium and small red grits. Red slip exterior and interior. US3,190. F greyish. Small white grits. Dark red slip exterior and interior. US3,57. C reddish. Abundant medium and large grey grits. Red slip exterior and interior. US3,144. M reddish-brown. Medium and large red grits, small and medium white grits. Red slip exterior and interior, burnished. US3,41. C reddish with dark grey core. Abundant small and medium white grits; vegetal temper. Red slip exterior and interior. US3,162. M greyish. Abundant medium and large red grits, sparse small white grits; rare mica. Light red slip exterior and interior. US3,208.

2 . husn salut in the iron age

-

85

plates

2 1

3

4

5

6

P late 14 – US3: globular jars; basins; crater. Scale 1:3.

86

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

michele degli esposti , chiara condoluci

P late 15 – US3: lid; small simple and carinated cups; bowls. M/C reddish. Abundant small and medium white grits, large dark grits. Light red, smoothed exterior. US3,55. F brownish. Small white and grey grits; rare mica. Pale brown slip exterior and interior. US3,13. F red. Small and medium white grits. Dark reddish-grey slip exterior, pale red interior. US3,167. F reddish. Sparse small and medium white grits. Light red slip exterior and interior, burnished exterior. US3,194. F red-brownish. Small white grits. Pale red slip exterior and interior. US3,243. F pale yellow. Small and medium white and grey grits; rare mica. Pale brown, smoothed exterior and interior. Incised graffito on exterior. US3,139. M/F red with grey core. Small and medium white grits, some large; grey grits; little mica. Light red slip exterior and interior. US3,138. F reddish. Small white grits, medium grey grits; little mica. Pink-whitish, smoothed exterior and interior. US3,153. M/F red. Medium white and grey grits. Red slip exterior, pale red interior. US3,244. F red. Small and medium white grits, small red and grey grits. Very pale brown slip exterior, red-brownish interior. US3,249. F pale red. Small white grits, medium red grits. Red slip exterior and interior. US3,248. F greyish. Small and medium white and grey grits. Red slip exterior, greyish interior. US3,154. M/F reddish. Small, medium and occasional large big white grits. Dark red slip exterior, pale red interior. US3,140. F greyish. Small and medium white grits. Red slip exterior, dark red-greyish interior. US3,156. F light brown. Small and medium white grits, small grey grits; rare mica. Pale brown, smoothed exterior, pinkish, smoothed interior. US3,141. F pale brownish. Small and medium white grits, small grey grits. Red-greyish slip exterior, red interior. US3,240. F reddish. Small and medium red grits, medium grey grits; rare mica. Grey, smoothed exterior, Pinkish, smoothed interior. US3,152.

2 . husn salut in the iron age

-

87

plates

1

4

5

3 2 6

7

8

12

13

11

10

9

14

16

15

17

P late 15 – US3: lid; small simple and carinated cups; bowls. Scale 1:3.

88

1. 2. 3. 4. 5. 6. 7.

michele degli esposti , chiara condoluci

P late 16 – US3: bowls and dishes. M/F yellowish. Medium grey grits, sparse medium white and red grits. Dark grey slip exterior and interior. US3,175. Very F light red. Light red-brownish, smoothed exterior and interior. Dark painted decoration exterior and interior. US3, 87. M red-brownish. Small and medium white and grey grits. Pale brown, smoothed exterior and interior US3,53. M buff. Dark grey and white grits; vegetal temper and mica. Pale brown, smoothed exterior and interior. US3,77. F pale red. Small white and grey grits. Pale brown slip exterior and interior. Red painted decoration on rim. US3,157. M/F light red. Dark grey and white grits. Red slip on exterior and interior, with traces of burnishing. Wheel made. US3,59. M brown. Small and medium white grits, vegetal temper. Reddish-brown, smoothed exterior and interior. Dark grey stained painted decoration exterior and interior. US3,51. 8. F/M red. Small white and red grits. Red slip exterior, pale re-brownish interior. US3,265. 9. F brownish. Rare mica. Traces of light red slip exterior and interior, burnished. Wheel made. US3,21. 10. M brownish. Small and medium white grits, grey grits; vegetal temper. Light red, smoothed exterior, pinkish, smoothed interior. Light red stained painted decoration exterior and interior. US3,189. 11. M red-brownish. Small and medium white grits, small red grits. Light red slip exterior and interior. US3,215. 12. M/C pinkish. Abundant large dark grey grits, sparse white grits; vegetal temper. Yellowish, smoothed exterior, light grey slip interior. Wheel made. US3,70.

2 . husn salut in the iron age

-

89

plates

1

2

3

5

4

6

7

8

9

10

11

12

P late 16 – US3: bowls and dishes. Scale 1:3.

90

1.

michele degli esposti , chiara condoluci

P late 17 – US3: necked jars; spouted jar; handle; lug; stand (?). F pale brown. Small dark grits, sparse small white grits, medium grey grits. Grey slip exterior and interior. Wheel made? US3,43. 2. F with grey core. Small and medium white grits, small dark grey grits. Light red slip exterior and rim. US3,33. 3. F pale yellow. Small and medium gray grits, rare medium white grits. Pale red slip exterior and interior. Wheel made? US3,44. 4. M red. Small red and dark grits; rare mica; vegetal temper. Light red slip exterior and interior. US3,183. 5. F light red. Small, medium and large white grits. Red slip exterior and interior. US3,137+158+159. 6. F light brown. Sparse small white grits, small and medium grey grits. Red-brownish slip exterior and interior, burnished. US3,181. 7. F pale brown. Small white grits. Pale red slip exterior and interior. US3,20. 8. F light brown. Sparse small white grits. Dark grey slip exterior, pale red interior. US3,177. 9. F reddish with grey core. Medium and small grey and red grits, small white grits; rare mica. Pale red, smoothed exterior, pinkish, smoothed interior. US3,172. 10. F light brown. Small white and grey grits; vegetal temper and mica. Light red slip exterior. Wheel made? US3,17. 11. C reddish with dark grey core. Abundant small white and large red grits. Light red, smoothed exterior, reddish-brown, smoothed interior. US3,135. 12. M red. Small and medium red grits; small white grits. Red slip exterior and interior. US3,36.

2 . husn salut in the iron age

-

91

plates

1

7 2

8

3

4

9

5

10

6

11

12

P late 17 – US3: necked jars; spouted jar; handle; lug; stand (?). Scale 1:3.

3. A complete sequence for the Early Iron Age of Central Oman. Excavations in the Basement and in the Burnt Building chiara condoluci , michele degli esposti , carl phillips

T he

excavation

Excavations at Husn Salut have revealed an archaeological stratigraphy the depth and complexity of which is all but common for coeval archaeological sites in South East Arabia, these being often discovered under few tens of centimetres of aeolian accumulation and collapsed material. Although in the central part of the site (Area 4, see Chapter 4) the bulk of the deposits derives from ancient structures collapse, secondary reworking of debris associated with later phases of occupation, and substantial erosive and colluvial events, the remains of the earlier structures built at the site were unearthed in the northwestern corner of the fortress, that were associated with mostly undisturbed archeological layers.1 In this area, one first building was identified and became the focus of early field seasons: given the fact that it stood at a lower elevation than the mud-brick terrace that occupied the top of the site (Terrace I), it was distinguished as a ‘Basement’. Only later did it became clear that this Basement actually coexisted with Terrace I in its original layout, i.e., before the erection of so-called ‘Building 1’ (see Chapter 2). Moreover, in the western part of the site, the presence of the ‘Burnt Building’ was revealed, which, instead dates back to the early layout of the site and can, therefore, be considered alongside the Basement with the intent of building a stratigraphic sequence that, as the radiocarbon dates have shown, covers the whole Iron Age I and Iron age II periods as usually intended, and further extends into the Iron Age III period (see Chapter 10).2 The structures that throughout some of the mission’s preliminary reports3 were referred to as belonging to a ‘buttress’ located on the western edge of the site, actually belong to the original extension of the Burnt Building’s rooms. The stratigraphic sequence outlined inside the Basement and the Burnt Building will be described here, and all the associated materials illustrated,4 in order to provide an exhaustive documentation supporting the discussion of the chrono-cultural issues prompted by the association of these materials with the radiocarbon dates obtained for the corresponding layers.

1 This is not to say that pristine, early phase contexts were completely absent in Area 4, but it was not possible there to identify an undisturbed stratigraphic sequence of such relevant comprehensiveness as those excavated inside the Basement and the Burnt Building. 2 Sub-phases are indicated in the Harris Matrix (fig. 40) referring to the associated main events, instead of introducing HSIa, HSIb, etc. labels, as these would find no correspondence in other areas of the site. 3 Preliminary Reports on the IMTO’s works at Salut are available on the Arabia Antica portal at http://arabiantica.humnet.unipi.it/index. php?id=imto-salut-campaigns. 4 As stone vessels and metal items are indeed of little chronological help, it was chosen to only include their drawings in the respective catalogues.

94

chiara condoluci , michele degli esposti , carl phillips

Foundation substructures Although not everywhere was it possible to reach the virgin bedrock below archaeological deposits, due to space constraints and in order to preserve significant contexts in place,5 in several points it was nonetheless exposed (fig. 39). Deep excavations thus allowed the identification of the extensive preparatory work for the actual site’s erection. It comprises a series of stone or mud-brick walls that were built in order to form closed compartments that were subsequently deliberately in-filled in order to level the bedrock’s unevenness and obtain flat surfaces, above which further construction could take place. Depending on the point of the site, different levels of such substructures were needed, and containment walls were discovered, built on top of one another: it is clear that in those cases they are not indicative of different building phases but have to be recognized as part of one single project. For the sake of clarity, only a few of these walls have been shown in the Harris matrix of the upper part of the site (fig. 40). Some of these layers, intentionally laid down inside the site’s substructures, show a series of characteristics that lead to their identification as foundation ‘deposits’ endowed with some sort of ceremonial value, as opposed to normal foundation fills, as discussed in detail in Chapter 5.6 Even when their fill is not so characteristic, from several of these layers potsherds were recovered that are of the uttermost importance for the discussion on the site’s chronology and pottery assemblage (i.e., US20, US274, US275, US279, US557). It is worth pointing out the presence of six contexts belonging to the very earliest construction phase for which a radiocarbon date was obtained: US6, US20, US274, US275, US337, US468 (fig. 41). With the erection of these substructures, two areas that stood at two considerably different elevations, in all likelihood as a consequence of a natural difference in altitude of the bedrock, were prepared for further construction. In the lower area the Basement was built, while the Burnt Building occupied the upper one. The space between the two was in all likelihood occupied already in this phase by the mud-brick platform of Terrace I, which covered the dismantled remains of two ruined Early Bronze Age tombs.7

F igure 39 – The bedrock exposed in room 2a, looking at the corner between wall M38 (front) and M87 (right).

5 Subsequent reconstruction and presentation to the public have always been among the primary aims of the IMTO, in a strong synergy with the Office of His Excellency the Adviser to His Majesty the Sultan for Cultural Affairs in Muscat (see Bizzarri 2015). 6 All these foundation fills and deposits could, therefore, be plotted inside a general site’s matrix at the same level as the walls and fills reached during the Basement and Burnt Building’s excavation. As an exemplification, the layers deposited below the compartment forming the northern part of platform SF49 (see Chapter 5) are shown in the matrix presented here. 7 Degli Esposti and Phillips 2012; Condoluci and Degli Esposti 2015: 6-12.

3 . a complete sequence for the early iron age of central oman

F igure 40 – Harris Matrix for the Basement, Burnt Building and Building 1 sequences.

95

96

chiara condoluci , michele degli esposti , carl phillips

F igure 41 – Calibrated dates for early foundation substructures.

Original layout The overall plan of the two buildings remained largely unchanged throughout their existence, with only a not-negligible raising of the successive floors inside the Basement. The Burnt Building plan only underwent changes during the HSII and HSIII phases (see also Chapter 2). In the Basement a central corridor, room 1a, was connected to the upper platform by means of a single flight of stairs at its northeastern end (fig. 42), and gave access to five rooms located at the lower level, four on its northern side (rooms 2a-5a) and a larger one on its southern side (room 6a). In the northern corner, wall M197 closed room 5a and delimited a sort of buttress whose original layout cannot be precisely identified due to later modifications. The original floors were named with several numbers as a consequence of interruption in the excavations from one season to the other, although they were often physically contiguous and therefore equivalent. In the Basement, US18fl, US19fl, US487 and US551 represent the earliest floors and all stand at almost the same elevation; in room 6a instead, a sort of stepped ramp was realized to span the elevation gap with the corridor, as the bedrock stood at a higher altitude (see fig. 13). Here, floors exposed in a small, appositely excavated trench were named as US319fl, US334, US409 and US458, and were often realized with a thin mud surface laid down on top of substantial mudbrick fills. This higher elevation of the southeastern part of room 6a, together with an apparent gap in its perimeter walls between wall M260 and M1, also suggests the possible presence of a second access to the Basement, directly in connection with Terrace I. Indeed, in rooms 2a and 3a, and in the western part of room 1a, below floor US18fl two previous floors were distinguished (US18Afl and US51fl), that indicate the existence of an even earlier surface. Also, a stone threshold was evident at this level between wall M36 and wall M19, oriented as the former and buried below floor US18 (fig. 42). A corresponding, lower floor was not exposed in the remaining rooms, either because this part of the Basement stood at a lower elevation (roughly 35 cm lower) than the eastern part or, more probably, because it disappeared completely due to later works. Be that as it may, the partition walls of the Basement, at least those for the northern rooms, seems to have been founded above the corresponding early floors.8 A series of layers did, however, accumulate above these floors

3 . a complete sequence for the early iron age of central oman

97

during the Basement’s initial occupation (see fig. 40); among the most significant in terms of collected pottery are US18, US240, US269. In room 5a, US473 has been interpreted as the evidence for a small fire that only hit this room and partially room 4a as well, where the corresponding US277 actually only shows scattered ash lenses. US473 has also been radiocarbon dated (fig. 43).9 In the Burnt Building, the central corridor (room 6b) opened onto three elongated and one larger room to the west (rooms 2b-4b), and on a small squared premise on the opposite side (see figs 14, 15). Adjacent to the building and in all likelihood connected with it was the large trapezoidal room 7b. The corridor was separated from the open part of Terrace I by a raised, rectangular mud-brick platform, SF49. The corresponding early floors were US100,10 US368, US 437 and their equivalents. A few deposits associated with their life of use were distinguished, unfortunately very poor in pottery or other materials. Among these, US561, the filling of a pit certainly prior to the large fire, is of relevance as it was radiocarbon dated (fig. 43). The remains of a charred wooden post were discovered against the eastern end of wall M129 (fig. 44), and may indicate the presence of some kind of wooden F igure 42 – The corridor inside the Basement (room 1a), partition. In room 7b, the original floor seems to be with stairs leading to Terrace I in the background. A stone represented by a compact mud surface, US95, in phase threshold aligned with wall M36 is visible in the foreground, with US391, a surface exposed immediately north of buried below floor US18. the room. Above US391, the imprint of a woven basket was discovered (see fig. 20), and below it the crests of walls M140 and M141 appeared, that can be interpreted as walls delimitating one of the foundation compartments which distinguish the construction technique of the site discussed in Chapter 5. At some point during this phase, probably as a consequence of the small fire in room 5a mentioned above, a modification in the layout of the northern corner of the Basement occurred, where wall M8 was connected to wall M9.11

8 The walls of room 6a, specifically M161, were instead founded deeper and directly reached the bedrock. 9 Since in these two rooms a floor level perfectly corresponding with that on which the large fire took place in the remaining rooms (i.e., US15fl, US16fl, etc.) was not identified, the possibility was considered that US277=US473 could actually be an extension of the same ‘destruction layers’ distinguished in the adjacent rooms (mentioned below). A test was made modeling US473 together with the other fire-related radiocarbon dates (cf. Chapter 10) but low agreement confirmed that its date appears too early when compared to the others, without a plausible explanation. The resulting lower limit for the fire event would be shifted around 1120 BC. Given these results, the interpretation of US473 as the outcome of a different, earlier and less extended fire has been preferred, although it is clear that a different one cannot be absolutely excluded. 10 It must be noted that US100 showed no evidence of having being hit by fire. As it actually overlaps with the footings of M143, whose face is conversely burnt, this may indicate that the floor was re-plastered after the large fire, and a previous floor, now lost, actually represented the first living surface in room 5b. 11 Uncertainties remain about the original layout of wall M9. This has originally been considered as one wall running all the way around the site, but recent excavations showed that at least in its north-eastern part the outer stone face of the site’s perimeter wall belongs to a late construction phase (Chapter 2 and Chapter 4). Therefore, a different label for the two parts of the walls should more correctly be assigned (in the Harris matrix this is indicated with ‘M reb.’). Also, the original layout of M8 could have been different, but it is impossible to ascertain.

98

chiara condoluci , michele degli esposti , carl phillips

F igure 43 – Calibrated dates for original layout’s contexts (a, b), prefire floors (c), destruction layers (d, e, f), and deposits linked to the Basement’s obliteration (g, h).

F igure 44 – Wooden post remains discovered against the northern end of wall M129 in the Burnt Building.

3 . a complete sequence for the early iron age of central oman

99

Floor raising and a large fire A new series of floors was established above the earliest ones inside the Basement, significantly only discovered in the corridor and in the northern rooms, i.e., those standing at the lower elevation. This also caused the burying of the lower steps of the stairs linking the basement with the upper platform. These floors were named as US15fl, US16fl, US261, US265fl and US289. Remains of a door socket set into US16 were found in the entrance to room 4a (fig. 45/a); in a later moment, the socket was replaced by a horizontal wooden beam, the impression of which was also exposed, probably to be interpreted in connection with a hole on the eastern face of wall M36 (fig. 45/b,c); this most likely indicates the presence of a wooden door or partition, and the existence of at least two sub-phases in the room’s use. In the Burnt Building the earlier floors were apparently still in use. A most significant event impacted the whole Basement and the Burnt Building around the mid 11th century BC (considering the latest possible date). Evident traces of a large fire are in fact visible on almost all the surviving walls, and rather thick layers of burnt matter (‘destruction layers’) were discovered in most of the rooms, lying above the mentioned floors and further buried beneath slightly baked clay accumulations that can be interpreted as further collapse of the damaged walls. A peculiar situation was that of a mud-brick stack found collapsed on top of US16fl and covered by the accumulation of US16 (fig. 46). Proposing an interpretation for these mud-bricks is difficult, but they could have been bricks prepared for further construction/renovation work inside the Basement that were then aborted after the large fire. Interestingly, all these mud-bricks showed the presence of deep finger impressions on one side, meant to provide better bonding between the different courses.12 In some areas, the evidence for these burnt layers was weaker, possibly mirroring the random diffusion of fire. This is the case for example of the southeastern end of room 6a, where a substantial burnt layer, US318+US319 was discovered north-east of wall M161, while on its opposite side the corresponding US331 only showed little traces

F igure 45 – Remains of a door socket (a) and of a later wooden beam (b) were discovered at the entrance of room 4a in the Basement, the latter probably connected with the presence of a hole in wall M36’s western face, likely the housing for a horizontal beam (c).



See also what was reported for M143 in Chapter 2, and footnote 122.

12

100

a

chiara condoluci , michele degli esposti , carl phillips

b

c

F igure 46 – Collapsed mud-bricks stack on top of US16fl (a), and other examples of finger-impressed mud-bricks from the Basement’s early contexts (b: US477; c: US551).

of burning. Different is the situation for room 7b, where an ashy, dark grey soft accumulation (US108) was distinguished, confined to the interconnection between the stone slabs of the whole floor. Interestingly, mixed with the loam, different materials have been found, such as Iron Age potsherds, small fragments of animal bones, a copperbase arrowhead fragment and a few iron fragments. This layer can either represent the remains of the fire-related deposit, subsequently covered by the newly laid flagstone floor, or it could derive from later activities inside the room.13 Radiocarbon dates were obtained from five contexts that can be related to this phase. Among these are US16, US318 and US367, three of the so called ‘destruction layers’; floor US289; and a late fill in room 4a, US276/242,14 indeed difficult to place in a specific phase but whose date perfectly fits with the other four layers mentioned here (fig. 43). Pottery from many of these fire-related contexts was abundant, as is notably the case for US15, US16, US400. Of great interest is also the pottery assemblage recovered from pit US35, cut through US16. In fact, almost the entirety of it is represented by carinated cups and spouted jars, two shapes for which a special significance seems strongly suggested when found in specific contexts and, moreover, mutually associated. The presence of long-handle bowls is also of relevance in this instance. These elements, further discussed elsewhere in this volume (Chapter 9), suggest that this pit could be interpreted as the evidence for some augural ceremony carried out prior to the final burying of the fire-destroyed Basement, although the archaeozoological remains recorded inside do not specifically differ from those from other Basement’s contexts (see Chapter 8).

13 In the Harris matrix the choice has been to place it in connection with the ‘destruction layers’ discovered in other rooms, but it could well be placed above the slab floor, in phase with US107. The pottery collected inside it is only of typically Iron Age II types. 14 Erroneous double naming attributed in different campaigns.

3 . a complete sequence for the early iron age of central oman

101

The Basement’s obliteration and late modifications to the Burnt Building’s layout Subsequent to the large fire, the Basement fell into disuse and was completely obliterated and leveled with the upper terrace’s surface by laying down a series of massive deposits (US7, US9, US11, US12, US13),15 for one of which a radiocarbon date is available that indicates this operation was concluded at least before the mid 10th century BC16 (US12, fig. 43). Accessory to this substantial in-filling operation was the erection of a series of walls that were to act as containments for the newly deposited material, providing it all with more stability (e.g. M173, M174, M6). Pottery from these layers further confirms the great uniformity in the assemblage associated with the HSI phase. Immediately after this operation, Building 1 was erected on top of the widened and levelled platform of Terrace I (see Chapter 2), marking the beginning of the HSII phase. Judging by the available date for the only safe floor of this building, US23, its construction had been completed by the mid 10th century BC (see fig. 22). A first re-arrangement concerned the Burnt Building, namely room 5b. This in fact was filled and transformed in a northern extension of platform SF49 by the deposition of a series of layers that also went to bury the north-western end of the corridor, contained by wall M136. Among these layers, that include US119, US389, US392 and US407, the most significant is undoubtedly US75+US74, interpreted as evidence of another ceremony connected with building activities (see Chapter 5). For this, a radiocarbon date is also available (see fig. 27) that was extremely useful for the interpretation of the sequence of activities in this small portion of the building. Pottery was not so abundant from US75, although other characteristic materials were discovered, all illustrated in Chapter 5 (pls 76-77); the other associated contexts were likewise very poor in pottery, of which a few sherds have been collected in US392 and in US407. The flagstone floor in room 7b was most likely laid down during this phase, and the earlier accumulation above it, US107, can be tentatively attributed to this phase as well, given the absence of typical Iron Age III sherds that would otherwise place it into the HSIII phase (see Chapter 2). Pillar M48 and furnace/oven were also realized above the flagstone floor, the pillar giving to the room a roughly similar layout to room 4b, at the other end of the Burnt Building. The re-structuring of platform SF49 also involved levelling the original niche/tank that had been realized in its centre; this was obtained by laying down layers US567, US568 and US569, from where some potsherds was discovered, not illustrated here. The place of this original feature was possibly taken by a new depression obtained in the northwestern corner of the newly enlarged platform, later filled by the laminated deposit US72. The final part of the Burnt Building’s sequence is of relevance because it offers an insight into the last phase of the site and the associated introduction of shapes and fabrics that are typical of the Iron Age III period. During this late phase, a series of pits was excavated in different points of the Burnt Building, most of which is characterised by the mentioned Iron Age III pottery (US385, US390, US393). Two of these pits, US385 and US393, have been tentatively interpreted as the foundation trench for the erection of wall M138, an HSIII phase wall (Iron Age III pottery was found inside the wall itself) that cut the walls of rooms 1b, 2b and 3b causing their considerable reduction. Iron Age III material was further discovered in US370, a floor located southwest of wall M138 in what had become an outer extension of room 2b.17 In room 7b, the ashy accumulation US99 and later dumping layers US86+US87, US88 and US89 also gave back occasional Iron Age III sherds. The same is true for US387 and US388 in the area in front of room 7B, where the late wall M130 was erected above US387 surface.

15 Floors US11fl, US12fl and US13fl, also representing one same level conventionally divided during excavation, are to be recognized as the surface above which levelling operations started, rather than actual living surfaces related to one additional sub-phase in the Basement’s life. 16 Modeling together the available dates, this limit gets shifted to the end of the 11th century BC; the date for US13 has actually been excluded from the model due to poor agreement (see Chapter 10). 17 Due to substantial erosion, it is not clear whether this space actually became a new small room or if it was just backfilled.

102

chiara condoluci , michele degli esposti , carl phillips

After the site’s abandonment, collapse and erosion strongly impacted the archaeological stratigraphy, finally resulting in a few massive layers that covered the ancient structures. A large part of Terrace I’ s extension was exposed below US3, while in the area of the Burnt Building a very recent small building was erected, whose foundations were only removed in 2009. In this area, US348 and US356 were distinguished below topsoil US1, and they stratigraphically correspond to US3.18

P ottery

and other objects from the

B asement

and the

B urnt B uilding

Overall, comparisons for the pottery assemblage collected during the excavation of these two buildings can be found at a number of Iron Age sites spread all over the Oman peninsula, with an abundance that makes precise references rather pointless.19 However, this ordinary nature is worth underlining in itself, as it distinguishes the material culture of Husn Salut’s early layers (i.e., those belonging to the HSI and HSII phase, prior to those connected with the late Burnt Building’s re-arrangement) as perfectly corresponding to what can be defined the ‘classic’ Iron Age (i.e., Iron Age II) repertoire for South East Arabia, first outlined with some consistency thanks to the French excavations at Rumeila, and specifically mirrored in that site’s Period I assemblage.20 These data are of the uttermost importance, as their combination with the available radiocarbon dates allows a re-assessment of the Iron Age chronology for the area of central Oman, as discussed in Chapter 10. Likewise, the assemblage associated with the later layers of the Burnt Building is perfectly coherent with the known repertoire distinguished for the later phase of the Iron Age, the so-called Iron Age III period, for which the materials from Rumeila’s Period II contexts also provided a reference during the early years of archaeological research in the region.21 However, from some of the uppermost deposits, some shapes are witnessed that could hint at a longer duration of Husn Salut occupation, as they recall examples that have been dated to the so-called Samad Late Iron Age.22

Foundation substructures Pottery from the foundation substructures and fills of the Basement and the Burnt Building is definitely scarce, yet of great relevance. In fact, typical Iron Age II shapes are witnessed in these very early contexts, notably including spouted vessels and carinated cups, together with different types of bowls. Two of the small cups also show a painted decoration that in one case comprises a bichrome net pattern on the exterior of the vessels (black and dark red – pl. 36/9), and in another a complex motif painted on the interior and developing with a radial layout (pl. 36/12), both known from Iron Age II sites throughout the region. Worth pointing out is also the attestation of two sherds showing traces of burnishing on the exterior surfaces (pl. 18/1-2). Among the substructural contexts of the Burnt Building, the sequence excavated between walls M143 and M201 was recognized to be connected to one of the so-called foundation deposits of the site, and is discussed in Chapter 5. It must be remembered that all the foundation compartments and associated materials actually belong to the same subphase of the contexts discussed here.

18 See Chapter 2 for a discussion of the materials from this last US. 19 The most comprehensive publications about Iron Age pottery in South East Arabia still remain the proceedings of the conference “La céramique de l’Age du Fer dans la péninsule d’Oman” held in Lyon in 1997 (Céramiques 1988) and the PhD dissertation of A. Benoist (Benoist 2000). Several single site’s reports and full publication appeared in more recent years corroborate the wealth of available data. 20 Boucharlat and Lombard 1985: 53-57, pls 45-52. 21 Ibid.: 58-59, pls 53-59. 22 ‘Samad LIA’, sensu Yule 2016: 32. These potsherds are preliminarily presented here in order to provide an exhaustive overview of Husn Salut’s material culture (see also Chapter 2), but their proper study has only just began and will be carried out together with that of possibly late materials that are emerging from renewed excavations of the IMTO at Qaryat Salut.

3 . a complete sequence for the early iron age of central oman

103

Original layout Materials connected to this phase of occupation, prior to floor raising, are definitely more abundant inside the Burnt Building than the Basement. They include a considerable portion of the repertoire of the main shapes witnessed throughout the two buildings’ sequence (and at the site in general): bowls with out-turned or flat rim, large basins, jars, hole-mouth and globular jars, storage jars and lids, bridged spouted jars, carinated cups, long handle bowls, and also baking pans. A medium fabric with variously sized grits and a widespread use of vegetal temper is generally employed, with the rather obvious exception of small carinated cups, always made in a fine fabric. Slip, when present instead of simple smoothing, is most commonly reddish-brown. Decoration – painted and incised – is scarcely attested (e.g. pls 36/5; 37/8; 38/1), and rare graffiti are found on carinated cups (pl. 36/8). In one particular case, secondary graffiti, including a fish, were incised above the original painted decoration (fig. 47).

5 cm

F igure 47 – A carinated cup from US 367 with a secondary graffito over the original painted decoration.

Floor raising and a large fire The largest part of the pottery assemblage collected in these two buildings comes from fire-related layers23 (see fig. 40), and in particular from US16 in the Basement and US394 in the Burnt Building. In room 3b no pottery was found associated with contexts of this phase. Carinated cups are well represented in all the burnt layers of the Basement – again, the largest amount coming from US16 – as well as spouted jars. Different is the case for the Burnt Building, where these two shapes are less commonly attested and are notably absent from the assemblage collected from US394, in room 1b, as said above, the context that gave back the largest amount of potsherds. Conversely, US377 in the corridor (room 6b) is the only one among the burnt layers of the Burnt Building where these two shapes are attested together (pl. 41/1-4, 8). Even long-handle bowls follow a similar pattern. In fact, although in a small number, they are represented in almost all the layers of the basement, while only one single fragment was discovered in room 2b of the Burnt Building, in US445 (not illustrated). Given the general association of this peculiar shape with snake representations, especially if one wishes to interpret the simple chevron incised examples as elementary stylized snake bodies, this long-handle bowl distribution may be seen in correlation with the almost total absence of snake-decorated pottery from the Burnt Building, with the exception of one sherd from a late context of room 7b (pl. 45/13).

23 This could be considered of some relevance, given also the presence of pottery types that can in some situations be associated with ceremonial/rites, although hypothesising a ceremonial burial of the whole two buildings seems a bit too hazardous (see Chapter 9).

104

chiara condoluci , michele degli esposti , carl phillips

Worth pointing out is also the presence of baking pans, almost exclusively discovered in US400.24 Carinated cups span a wide range of shape variants, deriving from a more or less accentuated, handmade carination; they are generally made in a brown or pinkish fabric with sparse white grits. Slip is very rare, smoothing being more common, with the occurrence of burnishing. Decoration is restricted to a few graffiti, while a unique wheelmade sherd shows a well done ‘running dog’ black and red painted decoration (pl. 21/30). Bridge-spouted jars, usually made in a medium fabric sometimes showing a grey core, can be red slipped or plain (smoothed). Painted decoration is sporadic and feeble. The few out-turned rim necked jars share the same medium, reddish fabric and are usually red slipped. Worth pointing out is a necked jar from US394, which bears a painted decoration comprising a row of dots that remains peculiar within Husn Salut documentation (pl. 39/8). Bowls and deep bowls with in-turned, simple flat or nail headed rim, together with large bowls/basins are widely attested in the assemblage. The fabric varies significantly from medium to coarse, with the occasional use of sand or vegetal temper. Noteworthy is the rather unusual employment of a medium/coarse fabric for a few bowls with incised decoration (pl. 25/13-14). Small globular jars, storage jars with incised criss-cross decorated cordon and associated lids are also present, the latter made in coarse, heavily gritted, usually red fabric as it is largely attested for the local Iron Age. As mentioned before, dishes are peculiarly numerous in US400, and represent one of the shapes that finds less frequent parallels in the Iron Age of the region.25 They are made in a distinct crude fabric with a large amount of large grits. One of the sherds includes an almost complete shell that is recognizable to the naked eye. The dishes diameter ranges from 20 to 26 cm. They all bear a pale brown slip and show evident traces of burning on their exterior, likely related to their use. Possibly connected to functional reasons is also the groove visible on the rim of one of the sherds (pl. 28/7). From the same layer comes the fragment of a jar in a fine fabric showing a red painted decoration comprising pending triangles (pl. 28/3), which finds good parallels for example in some painted jars from Nizwa,26 and a wall fragment decorated with painted vertical wavy lines (pl. 28/2). Worth mentioning is the presence, specifically inside the assemblage from US394, of several sherds for which an Iron Age III date is reasonably ascertained (pl. 39/2,4,10-11,13). They can in all likelihood be reckoned as intrusive, deriving from the excavation of a large pit with pit-fill US393. As mentioned, pits appear to be the main features related to the HSIII phase, identified in rooms 1b, 2b and 6b.27 Also related to the excavation of such pits are a few sherds from US386 (pl. 36/4-5), US378 (pl. 38/3), US377 (pl. 41/5) and US389 (pl. 43/5). Above, it was outlined how the assemblage from room 1b is slightly different from those from other rooms, and from those witnessed in the Basement’s rooms. In particular, possible ceremony-related objects and vessels come from rooms 1b and 6b, although with different distributions; no pottery comes from room 3b; while a particular concentration of animal bones comes from room 4b. Whether this might mirrors specific functions of the different rooms, possibly within the ‘protocol’ of ceremonies of some kind, has clearly to remain a subject for speculation, although it appears feasible.

The Basement’s obliteration The stratigraphically earliest context that can be related to the obliteration of the burnt Basement, meant to be a preliminary operation for Building 1’s construction, is the pit characterized by pit-fill US35. From this layer materials of a certain relevance and abundance were collected, which indicate it was not a simple pit but can rather be listed among the numerous examples of ‘foundation deposits’ identified at the site (see Chapter 5).

24 Only one example comes from the pre-fire occupation context US473, belonging to the previous construction phase (pl. 20/1). For US400, only diagnostic sherds belonging to this type of vessel, and very few painted and incised other shapes have been illustrated, since the remaining ones only replicates what is widely attested in contemporary layers. 25 See however the specimens from Shimal SX (Velde 1998: fig. 4/Epl 01.1-7) and al-Madam (Benoist 2000: fig. 134/11). 26 Schreiber 2007: 611, pl. 61/9-10. 27 Their possible connection with some wall re-foundations remains only hypothetical.

3 . a complete sequence for the early iron age of central oman

105

Numerous carinated cups, spouted jars, and long-handle bowls have been found together with bowls and large basins. Carinated cups from US35 are made in a fine light brown to pinkish fabric, with small white grits, and, occasionally, differently coloured grits; vegetal temper, and rare mica. Just like the carinated cups from US16, slip is uncommon, and they are usually smoothed. Several examples show a shaved bottom or string cut base. Graffiti are also testified both on the interior and exterior (pl. 30/1, 19, 23). Spouted jars are generally made in a medium brownish to reddish fabric, with small, medium and large grits; red slip usually still covers portions of the vessels. A cream slipped jar with a slightly different spout is distinct from the other not only shape-wise, but also for its fabric, which is light brown with abundant grits and vegetal temper (pl. 32/5).28 Bowls with flat or in-turned rim are all made in a medium brown fabric with abundant grits, cream or red slipped. Noteworthy is the earliest evidence for more carefully portrayed snake figures on pottery, namely on jars and long-handle bowls. Among these, a red slipped wall sherd made in a medium, reddish fabric with chamotte inclusions deserves mention (fig. 48). Its applied decoration comprises two snakes with triangular heads, elongated tongues, and bodies decorated with incised dots. Although snake representations on pottery are numerous in South East Arabia during the Iron Age II period, and the most various,29 the scene portrayed on this sherd shows a small snake approaching to a larger one, apparently simulating a typically human attitude of intimacy, usually referred to mother-child relationships. A similar kind of ‘humanisation’ of the snake occurs on two long-handle bowls also from US35 (pl. 33/5-6). In these examples, the whole flat handle symbolizes the snake body, as shown by the scattered circular impressions and by the distal part of the handle, which is shaped in the form of a snake head, with accentuated eyes, nose, nostrils and mouth. The latter is represented by a long longitudinal groove, which claims to be the closest thing to a smile.30 The uniqueness of these representations has to be highlighted, as they are unmatched not only at Husn Salut31 but also in South East Arabia as a whole, longhandle bowls representing in all likelihood a strictly local vessel typology, unparalleled outside the Oman peninsula. One further point worth noting may be the fact that, throughout the Basement’s sequence, longhandle bowl handles are always decorated with simple chevron impressions, with the exception of US35, and only in later contexts at Husn Salut (HSII phase) do they bear applied snake decorations – generally along the handle or climbing on the reservoir rim. As this is the case for the simply chevron-decorated ones, also this latter typology of long-handle bowls find large parallels both at worship sites such as Bithna, Masafi, alQusais,32 but also from surveys around Nizwa and on F igure 48 – A small wall sherd from US35 with applied the Jabal al-Akhdar.33 snake decoration (see pl. 34/8).

28 From the same layer three similar, fragmentary spouts are known ( not illustrated). 29 Suffice it to mention the possible scene of snake hunting with nets visible on two large jars from Saruq al Hadid (Nashef 2010: 220, figs 7, 8; see also Weeks et al. 2017: fig. 16). 30 Another handle (Cl15) has the distal end with a ram’s head, while two fishes, parted by an axe, are applied on the handle length (pl. 33/3). 31 Here, other animals represented on long-handle bowls were discovered, but none show similar traits (cf. Chapter 2). 32 E.g. Benoist 2013a: fig. 112/7-8; Benoist et al. 2012a: fig.14; Taha 1982-83: fig. 15/A. 33 Schreiber 2007: 523, pl. 67; 547, pl.79/9-10.

106

chiara condoluci , michele degli esposti , carl phillips

The assemblage collected from the other deposits laid down to obliterate the Basement, and thus sealing US35, comprises a wide range of shapes, although slightly more restricted than the underlying fire-related layers: carinated cups and cups, bowls, jars, globular jars, storage jars and lids, and censer-lumps. Of relevance is above all the occasional evidence for the use of the slow wheel for jars and bowls (e.g. pl. 33/9). Carinated cups are generally made in fine pinkish or orange-brown fabrics, with little grits and sand or limestone temper (little mica is also testified), and are always smoothed without traces of slip. The few examples of cup and bowls have fabrics ranging from fine to medium, with the inclusion of vegetal temper, mica and grits. Slip and painted decoration are exceptional. Large bowls, mostly with flat or rounded rim, are made in fabrics ranging from coarse to medium/coarse, with the presence of vegetal temper, mica and variously sized grits. Slips range from grey to pale yellow. Incised criss-cross decoration is witnessed only on one rim. Jars usually have out-turned triangular or simple rounded rims. Fabric ranges from medium/coarse to coarse; slip is uncommon. The presence of a single spouted jar sherd, made in a medium fine pale yellow fabric, from one of the latest accumulations (pl. 33/7), is possibly the most clear evidence on the uniformity of the pottery assemblage throughout the Basement’s stratigraphic sequence. To this phase has possibly to be attributed US329,34 from which an interesting sherd was collected, Cl298 (fig. 49). This is an incised sherd belonging to a closed vessel, likely a small jar or jug, hard fired and made in a medium/fine fabric with a gritty touch, including small and medium white grits and sparse vegetable temper. A red slip (turned brown in firing) was possibly applied on the exterior. Its main feature is, however, the presence of post-fired incised signs. They form a symmetrical composition, with two external signs 5 cm bracketing a more complex one; the graffito seems to be something more than a simple drawing, although its F igure 49 – Incised sherd Cl298, coming from one of the latest meaning remains elusive.35 layers preceding the construction of Building 1 (US329).

Burnt Building’s re-arrangement Indeed, placing these contexts in a fixed place inside the general Harris matrix of the excavation is rather difficult; the choice of indicating them as part of the HSII phase instead of HSIII is mainly based on the absence of unmistakably Iron Age III sherds. Therefore, it is not surprising that also the overall assemblage from these layers replicates that of the earlier phases, including carinated cups, spouted jars (pl. 43/16), and painted sherds (pls 43/1; 44/1-2; 46/8), with frequent attestation of necked jars. To this phase also belong the deposits laid down inside former room 5b, that have been interpreted as the outcome of a possible re-foundation ceremony (see Chapter 5). Consistently with the stratigraphic position of these layers, among the material from the main deposit, US75, a few sherds display an ‘Iron Age III’ look, although not belonging to the typical repertoire for that period.

34 In the small area of room 6a investigated by means of small trenches, it is impossible to distinguish the layers that are connected to postfire collapse, then pertaining properly to the HSI phase, from those belonging to the series of obliteration deposits marking the beginning of the HSII phase. This also applies to US399, a context from which numerous potsherds were collected that are anyhow not illustrated here, as they would just replicate the assemblage witnessed in corresponding layers. 35 Further study of the sherd and of the incised motif is on-going.

3 . a complete sequence for the early iron age of central oman

107

Later modifications to the Burnt Building’s layout It has been already underlined (see Chapter 2) that the materials collected from the very latest contexts inside the Burnt Building are the principal means of datation for the last phase of the site, HSIII, by virtue of them comprising typical Iron Age III shapes and/or showing a specific surface treatment. These are associated with a dense, fine fabric that is consistent with their inclusion in the ‘Burnished Maroon Slipped Ware’ class, as defined by P. Magee,36 which in turn corresponds to the ‘burnished slipped-ware’ that was recognized as a newly-introduced category in Rumeila’s Period II contexts.37 Probably the most distinguished Iron Age III shapes are large, shallow dishes (pl. 47/10, but see also pl. 39/4, intrusive), together with bowls with flaring walls and slightly incurving, pointed rim (pls 47/3; 48/24-26; 45/15, the latter significantly joining with one intrusive fragment recovered from US394),38 further characterized by surface burnishing. In a similar ware is made a so-called tulip-bowl from US88, also displaying an analogous surface treatment (pl. 52/3); other, rather occasional specimens were discovered at the site (see pl. 9/9-14). Again, Rafaq 2 provides abundant comparisons.39 From US86 a globular jug was found (pl. 51/10) that is unmatched elsewhere at the site except for one sherd coming from US44, a very late pit in the upper part of the site (see pl. 9/15). They broadly recall one jug from the singleperiod, Iron Age III settlement of Rafaq 2, although those examples had no handle,40 while a second vaguely similar one also had a spout on the opposite side of the handle and finds a comparison from Rumeila Period II.41 Another shape scarcely attested at Husn Salut are large basins with out-turned rim, distinguished by the presence of horizontal lugs in the upper part of the walls, one specimen of which comes from US99 (pl. 50/1). The fact that other parallels from the site only come from the late, reshuffled context of US3 (see pl. 14/3, 5-6) may indicate a chronological relevance for this shape. However, although a comparable vessel is consistently reported from Rafaq 2,42 at Rumeila this shape appears both in Period I and Period II contexts.43 Finally, from US377 two vessels have been discovered which may be indicative of a very late occupation at the site, post 300 BC, and therefore have to be reckoned as intrusive just like one possibly Iron Age III bowl (pl. 41/5). One is a particularly oblong vessel (pl. 41/16) recalling the shape of some Samad LIA ‘balsamaria’, although the latter are usually made in a completely different, notably untempered ware.44 The other is a lentoid flask (pl. 41/17), that finds parallel in post 300 BC contexts from both central Oman and the Emirates,45 at the same time recalling, in terms of fabric and surface treatment, typical Iron Age III wares. The presence of these vessels, although needing further study, is anyway intriguing, also in light of the reported attestation in a few late contexts of Husn Salut of peculiar jars provided with drains at their base, that only find parallels in Samad period sites (see Chapter 2).

Other finds Finds from the Basement and the Burnt Building mainly comprise a few copper-base fragments of difficult interpretation; soft-stone vessels; hammer-stones; clay objects; and shells, plus some rare copper-base weapons and tools. Among the objects connected to daily life activities it is interesting to note the presence of some loom weights, that



Magee 2005a. Boucharlat and Lombard 1985: 58. 38 For both shapes, see for example Phillips 1998a: fig. 8. 39 Phillips 1998a: fig. 9, more properly no. 7. 40 Phillips 1998a: fig. 19/5. 41 Phillips 1998a: fig. 19/4; Boucharlat and Lombard 1985: pl. 53/12. 42 Phillips 1998a: fig. 15/8. 43 Benoist 2000: fig. 53/3, 10 from Period I; figs 62/3-4; 69/1 from Period II. 44 See for example Yule 2005: fig. 4/2, 5 and p. 306. 45 E.g. Yule 2016: figs 32/20; 41. 36 37

108

chiara condoluci , michele degli esposti , carl phillips

were obtained re-shaping and piercing stray potsherds (e.g. pl. 18/8). Their shape ranges from quadrangular to rhomboidal or almost circular; a stone specimen was also collected from US400 (pl. 28/12). Analogous clay loom weights were found in several other contexts at Husn Salut, and a fair number also come from Iron Age levels excavated at the Bronze Age tower site of Salut ST1,46 possibly testifying to a weaving activity on a domestic scale. The presence of two spindle whorls would be consistent with this hypothesis. Both are made in a dark grey soft stone; one (pl. 34/12) has the shape of a truncated cone, with incised vertical lines, the other (pl. 35/14) is biconical and is also decorated with incised radial lines. However, the small dimensions of the latter would also be consistent with a use as a bead. Simple stone tools, such as hammerstones, handstones and pestles (not illustrated) have been found in several layers of the Basement and Burnt Building sequences, and further bear witness to daily life activities at the site. Two large shells, one belonging to the Caridoidea family (pl. 28/13, Plagiocardium pseudolima?), the other to the Pectinidae family (pl. 28/14, Chlamys?), have been found in the burnt layer US400. Although their function is uncertain, their use as a liquids container can be supposed for the deepest one. Pectinidae shells are known to have been used as oil-lamps in the past, but no burning traces were found inside Sh43. A few fragments of soft-stone vessels were discovered inside the Basement, testifying the presence of conical vessels, bowls and lids (see Chapter 6; nos 21, 29, 38, 45). Two heavily corroded, flat copper-base bands (not illustrated)47 recall a number of similar items found throughout the site and specifically in the Building 1 area, together with other copper-base scraps, possibly stored to be re-melted. These items have been interpreted as clamps, in some cases possibly used for broken vessels’ repair.48 A possible ingot fragment, showing one flat and one slightly convex surface (see Chapter 7, no. 145), a conicalshaped undefined object, and one fragment of a probable earring were also found (both not illustrated). More significant may be the presence of a few weapons. Apart from a single arrowhead form US386 (Chapter 7, no. 64), a bronze spearhead was found in deposit US391, outside room 7b (Chapter 7, no. 75). A shaft-hole axe (Chapter 7, no. 83) comes from the burnt layer distinguished inside room 6b, the Burnt Building’s corridor. As carinated cups and one spouted jar were collected from this same layer, this association is suggestively remindful of the one that characterises several ceremonial deposits identified at the site (see Chapter 5). Interestingly, together with a more ‘secular’ loom weight (Cl95, not illustrated) and the fragment of a copper-base pin (Chapter 7, no. 112), a collection of objects that may have been endowed with some specific, ritual meaning, was discovered inside US394. These comprise a copper-base snake (Chapter 7, no. 9), a copper-base ladle/spoon? (Chapter 7, no. 14), an iron blade (Chapter 7, no. 149); however, the associated pottery lacks the two shapes that appear to be more likely connected with libations – whatever their nature – i.e., spouted jars and carinated cups. Finally, animal bones are widespread throughout the stratigraphic sequence and are common in both buildings, with the notable exception of room 3b that, as was shown above, strongly differs from the others rooms also for the absence of potsherds.49

46 Degli Esposti 2011: fig. 13. 47 MB4 comprises a horizontal small rod with two broken vertical elements on its ends (h. 1.9, l. 4.3, w. 1.1 cm); MB159 comprises two rectangular bands, flat in section, joined by a rivet. Other small circular holes are visible on both bands. The edges are slightly rounded, one is folded. Bands measures: l 2.8 and 3.6 cm.; w. 1.4 and 1.2 cm. 48 Similar plates with rivets found at Jabal al-Emala were interpreted as fragments of a metal gauntlet (Potts 1998: 201), but the presence of more completely preserved specimens at Husn Salut strongly contrasts such a reconstruction (see Chapter 7, no. 108). 49 The particalur exclusive presence of sheep/goats bones in pit-fill US35 is discussed in Chapter 3. See Chapter 8 for the archaeozoological study of the animal remains from the discussed buildings.

plates

110

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

chiara condoluci , michele degli esposti , carl phillips

P late 18 – Basement: US20, 557, 279, 275, 274, 457, 51. M orange with grey core. Medium and small grey grits, rare red large grits; rare vegetal temper. Rotation marks on interior (slow wheel?). Burnished, reddish exterior and interior. Burning traces on exterior. US20,1. M pale brown-whitish. Red and white grits; sparse vegetal temper. Pale brown-greenish slip exterior, whitish interior. Burnishing traces exterior. Dark red painted decoration exterior. US557,1. M pinkish. Red grit, occasional black grits. Incised decoration on rim. US279,1. M dark grey. Dark grits, sandy texture. US275,1. M grey. Dark grits; vegetal temper. US275,2. M greyish-brown. White and dark grits. US274,1. M yellowish-white. Dark and red grits. Dark brown slip exterior, reddish-brown interior. US274,2. Loom weight made from coarse ware sherd. US457,1 = Cl169. M brownish, dense. Small, medium and large red grits, small and medium grey grits. US51,3. F reddish. Small white and grey grits. Smoothed exterior and interior. Painted dark reddish and incised decoration on exterior. US51,2. M/F brownish. Small, medium and large grey and red grits. Washed surfaces smoothed on inside. US51,1. M reddish, dense. Medium and large red grits. Light red slip exterior and interior. US51,4. M reddish, dense. Small and medium white and red grits, sparse large red grits. Pinkish-white slip exterior, reddish-yellow interior. US51,5.

3 . a complete sequence for the early iron age of central oman

-

111

plates

1

2

3

4

5 6

7

8

10 9

11

12

13

P late 18 – Basement: US20, 557, 279, 275, 274, 457, 51. Scale 1:3, except no. 13: scale 1:2.

112

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

chiara condoluci , michele degli esposti , carl phillips

P late 19 – Basement: US18, 287, 477, 473. M brown. Black grits. Orange-brown, smoothed exterior and interior. US18,3. M/C light brown. Brownish red and grey grits. Light brown-buff slip exterior and interior. Incised decoration exterior. US18,2. C brown. Abundant red and black grits; vegetal temper. Smoothed exterior and interior. Light brown slip traces on upper and lower surfaces. US18,5. C pale-brown. Abundant red grits, some large. Incised herringbone decoration. Burnt traces on base and walls. Long handle bowl. US18,1 = Cl46. C red. Abundant large white and dark grey grits; vegetal temper. Dark red-brownish slip exterior and interior. US477,1. F brown. Abundant, small brown and black grit temper. Brown slip exterior and interior. US473,4. F orange-brown. Sparse black grits. Light brown slip exterior and interior. US473,2. F brown. Brown grits. Light brown slip exterior and interior. US473,3. F/M brown. Abundant brown and white grits, some large; vegetal temper. Light brown slip exterior and interior. String-cut base with ‘shaved’ sides. US473,1. M brown. Abundant black grits; vegetal temper. Black-brownish slip exterior and interior. US473,8. F/M brown. Brown and white grits; sparse vegetal temper. Light brown slip exterior and interior. US473,9. C brown. Abundant black, brown and white grits; vegetal temper. Blackened exterior and interior. US473,16. M/C brown. Abundant angular black grits; sparse vegetal temper. Dark brown slip exterior and interior. US473,10. M/F orange-brown. Abundant brown grits; sparse vegetal temper. Lightly burnished, orange-brown slip exterior and interior. US473,11. M brown. Vegetal temper with abundant brown, black and white grits. Brown slip exterior and interior. US473,12. C orange-brown with grey core. Abundant black and white grits; vegetal temper. Burnished light brown slip exterior and interior. US473,13.

3 . a complete sequence for the early iron age of central oman

-

plates

113

2 1

3

4

5

6

8

7

10 9

11

12

13

14

15

16

P late 19 – Basement: US18, 287, 477, 473. Scale 1:3, except no. 5: scale 1:4.

114

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

chiara condoluci , michele degli esposti , carl phillips

P late 20 – Basement: US473, 513, 472, 269, 240, 478. C brown. Abundant black, brown and white grits; vegetal temper. Blackened exterior and base. US473,15. C brown. Abundant brown and white grits; vegetal temper. Light brown slip on upper and undersides. US473,17. F light red-brownish. Small white and grey grits; vegetal temper. Red slip exterior, red-brown interior. US513,1. F black. Sparse white grits. Black slip exterior and interior. US472,2. F brown. Sparse white grits; vegetal temper. Black-grey slip exterior and interior. US472,3. M orange. Brown and white grits; sparse vegetal temper. Light brown slip exterior and interior. US472,1. F pink with inner greyish-brown section. Sand tempered with fine limestone. Red slip exterior and interior. Incised decorations on the base (exterior and interior). US269,2. M orange-brown. Rare white grits. Red slip exterior and interior. Incised decoration on base interior. US269,3. M light brown. Red grits; occasional vegetal temper. Light red slip exterior and interior. US269,1. F light brown. Send tempered. Red slip exterior and interior. US240,1. M/F yellow. Dark grits, occasional white grits. Red slip exterior and interior. US240,4. M/F light brown with pink inner section. Small dark grits. Red slip exterior and interior. US240,2. M light brown with pink inner section. Red slip exterior and interior. US240,3. C reddish-brown. Abundant red and grey grits Incised herringbone decoration. Consistent burnt traces inside reservoir. Long handle bowl. US240 = Cl82.

3 . a complete sequence for the early iron age of central oman

-

115

plates

1

2

5

4

3

6

9 7

8

10

11

12

13 14

P late 20 – Basement: US473, 513, 472, 269, 240, 478. Scale 1:3.

116

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

chiara condoluci , michele degli esposti , carl phillips

P late 21 – Basement: US16. M light brown. Small red and grey grits. Buff-brown, self-slipped exterior and interior. US16,63. M light brown. Sparse red and white grits, occasional large red grits. Buff-light brown smoothed exterior and interior. US16,178. M light brown. Sparse red and white grits, occasional large white grits. Buff, smoothed exterior and interior. Incised cross exterior. US16,114. M light brown. Occasional red and white grits; vegetal temper. Whitish slip exterior and interior. US16,185. M orange-brown. Red and dark grey grits. Mottled brown-grey exterior. Blackened-grey interior. US16,175. F light brown. Occasional white grits. Light brown-buff, smoothed exterior and interior. Faint traces of red paint exterior. US16,66. M orange-brown. Abundant white grits. Brown, smoothed exterior and interior. US16,120. F brown. Sparse small white grits. Red self-slipped exterior and interior. US16,76. M pinkish-brown. Red and white grits; vegetal temper. Whitish slip exterior and interior. US16,194. F brown. Light brown smoothed exterior and interior. US16,65. M brown. Sparse red and white grits. Whitish, green-pinkish, smoothed exterior and interior. Incised decoration on exterior and interior. US16,116. M brown. Red grits, occasional white grits. Orange-brown slip exterior and interior. US16,84. M orange-brown. White grits. Brown slip exterior and interior. US16,115. M brown. Some grit plus white grits, occasional red grits. Red-brown slip exterior and interior. US16,63. M brown. Sparse red and white grits. Brown, smoothed exterior and interior. US16,176. M red-brown. Red and white grits. Red, smoothed exterior and interior. US16,112. F light brown. Occasional white grits. Brown, smoothed exterior and interior. US16,79. F light brown. Red and white grits. Brown-whitish, smoothed exterior. Pinkish-brown interior. Incised decoration interior. US16,122. F brown. Rare white grits. Brown, smoothed exterior and interior. US16,177. F light brown. Rare white grits. Light brown, smoothed exterior and interior. Traces of incised decoration interior. US16,179. M orange-brown. Sand temper, occasional white grits. Red, smoothed red interior, light brown exterior, smoothed above rim. US16,128. F brown. Rare white grits. Brown, smoothed exterior and interior. Traces of incised decoration interior. US16,180. M brown. Sparse red and white grits, some larger white grits. Orange-brown, smoothed exterior and interior. Half ‘shaved’ bottom with slightly burnished appearance. US16,67. M brown. Sparse red and white grits. Orange-brown, smoothed exterior and interior. US16,136. M brown. Sparse red and white grits, some larger white grits. Orange-brown, smoothed interior and upper rim. ‘Shaved’ base. US16,62. F reddish, dense. Rare white grits; rare vegetal temper. Red, smoothed exterior and interior. String-cut base. Rotation marks on interior (slow wheel?). US16,1. M orange-brown. Sparse red grits. Orange-brown, smoothed exterior and interior. US16,138. M grey. Small red and white grits, sand temper. Red, smoothed exterior and interior. US16,148. M light brown. Sparse red and white grits, occasional large white grits (3mm). Brown, smoothed exterior and interior. US16,125. F pinkish, very dense. Small black, rare small white grits. Red slip exterior and interior. Back painted decoration exterior. Rotation marks on interior (slow wheel?). US16,23. F pinkish, dense. Small, medium and large red grits, small white and grey grits. Red slip exterior and interior, smoothed exterior. ‘Shaved’ bottom. US16,40. M orange-brown. Abundant red grits. Orange-brown, smoothed exterior and interior. US16,169. M/C brown. Abundant brownish-red grits. Brown, smoothed exterior and interior. US16,58. M red-brown. Occasional white grits, sand temper. Whitish slip exterior and interior. Purplish-brown painted decoration on rim. US16,200. M orange. Abundant small and medium white grits, rare small black and red grits; vegetal temper. Pale yellow slip exterior, light red interior. Dark red painted decoration interior and rim. US16,32. M brown. Abundant brownish-red grits. Light brown slip exterior and interior. Burnt exterior. US16,57.

3 . a complete sequence for the early iron age of central oman

1

2

6

11

12

17

21

22

27

117

plates

4

8

7

16

3

-

5

9

13

10

15

14

18

19

23

24

20

25

28

30

29

26

33 31

32

35 34 36

P late 21 – Basement: US16. Scale 1:3.

118

1. 2. 3. 4. 5. 6. 7.

chiara condoluci , michele degli esposti , carl phillips

P late 22 – Basement: US16. M brownish, dense. Small and medium white and red grits, small grey grits. Pinkish, smoothed exterior and interior. Soot on exterior. US16,47. M brown. Abundant red and black grits. Light brown slip exterior and interior. US16,110. M reddish, dense. Small and medium grey grits, small white grits. Light reddish-brown slip exterior, red slip interior. US16,53. M pinkish, dense. Small and medium grey and red grit. Pink, smoothed exterior, pinkish-white, smoothed interior. US16,52. M/C pinkish. Red large and medium grits, rare small white grits; vegetal temper. Pale brown slip exterior, smoothed interior. US16,30. M/F pinkish, dense. Small and medium red and grey grits. Pink slip exterior, white interior. US16,111+170. M brown. Abundant small red and black grits. Light brown slip exterior and interior. US16,45.

3 . a complete sequence for the early iron age of central oman

-

119

plates

1

2

3

4

5

6

7

P late 22 – Basement: US16. Scale 1:3.

120

chiara condoluci , michele degli esposti , carl phillips

P late 23 – Basement: US16. M red with grey core. Small, medium and large red grits, medium grey and white grits; occasional mica. Light red slip exterior and interior. US16,4.1. 2. M pale brown. Small and medium red grits, small white grits. Red, smoothed exterior and interior. US16,107. 3. M greyish. Small, medium and large red grits, small white grits. Brown, smoothed exterior, grey, smoothed interior. Traces of red painted decoration exterior. US16,105. 4. M reddish. Small white and medium red grits. Light red slip on exterior and interior. US16,92. 5. M reddish. Small white and grey grits. Red slip exterior and interior. US16,103. 6. M reddish. Small and medium white grits, rare medium red grits. Red slip exterior and interior. US16,171. 7. M reddish. Small white grits, rare mica. Light red slip exterior and interior. US16,108. 8. M reddish. Small and medium red grits, small grey grits. Red, smoothed exterior and interior. US16,100. 9. F light brown. Small grey grits. Red slip exterior and interior. US16,109. 10. M brownish. Small white grits, small and medium red and grey grits. Red slip exterior and interior. US16,49. 11. M brownish. Abundant small and medium red grits, small white grits. Red slip exterior and interior. US16,42. 12. M/C orange-brown. Red and grey grits, occasional larger white grits. Red slip exterior and interior rim, smoothed exterior. US16,165. 1.

3 . a complete sequence for the early iron age of central oman

-

121

plates

1

3

2

5

4

7

6

9 10

8

12 11

P late 23 – Basement: US16. Scale 1:3.

122

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci , michele degli esposti , carl phillips

P late 24 – Basement: US16. C red-brownish. Abundant red and grey grits. Incised herringbone decoration. Long handled bowl. US16, 4 = Cl67. C brown. Abundant red and grey grits. Burnt traces all over. Long handled bowl. US16,94. C pale red. Red and white grits; vegetal temper. Light red slip on upper surface of handle. Incised herringbone decoration. Long handled bowl. US16,3 = Cl54. M brownish. Abundant small, medium and large red and grey grits, medium and small white grits. Reddish, smoothed exterior and interior. Incised decoration exterior. US16,46. M brownish. Small, medium and large white and grey grits, abundant medium and large red grits, rare mica. Red slip exterior and interior. US16,95. C brown. Abundant red, grey and white grits. Light brown slip exterior and interior. US16,61. C dark grey. Small white grits, medium and large red and grey grits. Dark red slip exterior, light red slip interior. Incised decoration exterior. US16,97. M brownish. Abundant small, medium and large red and grey grits, sparse small and medium white grits. Reddish-yellow slip exteriors, light brown interior. US16,101. M reddish. Small, medium and large white and red grits, rare mica. Pinkish slip exterior, pink interior. US16,102.

3 . a complete sequence for the early iron age of central oman

-

plates

2

1

123

3

4

6

5

7

8 9

P late 24 – Basement: US16. Scale 1:4, except nos 1-3: scale 1:3.

124

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

chiara condoluci , michele degli esposti , carl phillips

P late 25 – Basement: US15. F pinkish-orange. Small white and red grits. Light reddish-brown, smoothed exterior and interior. US15,69. F pink, dense. Small white grits. Pinkish, smoothed exterior and interior. US15,6. F pinkish. Dark and white grits. Pale brown, smoothed exterior, pale brown interior. US15,5+70. F greyish. Abundant white grits, occasional dark grits. Pale brown slip exterior and interior. US15,12. M/F pinkish-orange. Rare white, red and dark grey grits, rare mica; occasional vegetal temper. Buff slip exterior, pinkishorange interior, smoothed near rim. US15,1. F pink, dense. Rare mica. Pale brown, smoothed exterior and interior. US15,8. F pink, dense. Occasional dark and white grits. Pinkish, smoothed exterior and interior. US15,9. F light red, dense. Occasional small white grits. Light reddish-brown, smoothed exterior and interior. US15,16. M/F pinkish-red. Rare white and dark grey grits; occasional vegetal temper. Pale red slip exterior, pinkish-orange interior. US15,72. F light red, dense. Occasional small white and red grits. Light reddish-brown slip exterior and interior. US15,79. F light brown. Abundant small white grits, rare small red grits. Pale yellow, smoothed exterior and interior. US15,77. F/M pale yellow. Small white grits, rare dark and red grey grits; vegetal temper. Pale yellow, smoothed exterior and interior. US15,80. C pale brown. Large red grits, large and medium grey grits; vegetal temper. Pale brown, smoothed exterior and interior. Incised decoration on rim. US15,53. F/M orange-brown. Small white grits, rare dark and red grey grits; vegetal temper. Brown-reddish, smoothed exterior and interior. US15,54. M reddish-orange. Abundant small red and grey grits. Res slip exterior. US15,59. M light brown. Small white grits, large and medium red grits, occasional brown grits; occasional vegetal temper. US15,71. M/C brown. Red and white grits. Soot on the interior. Long handled bowl. US15,42. C/M pinkish. Sparse white grits, abundant dark grits. Pinkish exterior and interior. Incised herringbone decoration. Deep groove in section at the end of the handle. Long handled bowl. US15,41. C pinkish. Red and grey grits. Incised herringbone decoration. Long handled bowl. US15,89. C greyish. Abundant dark and white grits. Very pale brown exterior and interior. Incised herringbone decoration. Soot on the interior. Long handled bowl. US15,40.

3 . a complete sequence for the early iron age of central oman

1

3

2

plates

125

4

5

10

9

8 6

-

7

13

11

12

14

15

16

17

19 18

P late 25 – Basement: US15. Scale 1:3.

20

126

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci , michele degli esposti , carl phillips

P late 26 – Basement: US15. M/F reddish-brown. Small white and red grits. Light brown slip exterior, brown, smoothed interior. US15,56. M/C light brown. Abundant small white grits, rare red, rare red grits, large and small grey grits, rare mica; rare vegetal temper. Light brown, smoothed exterior and interior. US15,78. M orange-reddish. Sparse white and red grits. Reddish, smoothed exterior and interior. US15,55. M brown-reddish. Red and white grits. Brown, smoothed exterior, reddish, smoothed interior. US15,60. M brown. Red and grey grits, rare vegetal temper. Red slip exterior, brown, smoothed interior. US15,65. M light brown. Red and black grits; sparse vegetal temper. Light brown slip exterior and interior. US15,66. M/F orange. Sparse red grits, rare white grits. Red slip exterior and interior. US15,50. M orange. White and red grits, some large, occasional vegetal temper. Orange-reddish, smoothed exterior, orange-reddish interior. US15,88. M light brown. Sparse white grits, occasional grey grits. Buff-brown slip exterior, light brown interior. US15,86.

3 . a complete sequence for the early iron age of central oman

-

127

plates

1

2

4

3

5

6

7

8

9

P late 26 – Basement: US15. Scale 1:3.

128

1. 2. 3. 4. 5. 6. 7.

chiara condoluci , michele degli esposti , carl phillips

P late 27 – Basement: US15. M light brown. White and red grits. Red slip exterior and interior. US15,87. M/C brown-reddish. Red and white grits, occasional grey grits; rare vegetal temper. Light brown, smoothed exterior, light brown slip interior. US15,85. M/C brown. White and red grits, some large, sparse vegetal temper. Light brownish slip exterior and interior. US15,83. M red-brownish. Small red and grey grits, sparse vegetal temper. Red slip exterior. US15,74+75+76. M red-orange. White grits, small dark grey grits. Red slip exterior. US15,48. M/C reddish with grey core. Abundant dark and white grits; vegetal temper. Pale brown, smoothed exterior and interior. US15,34. M reddish with grey core. Small red and grey grits, sparse small white grits. Pale red slip exterior. US15,90.

3 . a complete sequence for the early iron age of central oman

-

129

plates

1

2

3

5

4

6

7

P late 27 – Basement: US15. Scale 1:3.

130

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

chiara condoluci , michele degli esposti , carl phillips

P late 28 – Basement: US400, 318, 331, 258. F orange with grey core. Small and medium white grits. US400,19. F pale buff with some small grey inclusions. Light orange slip exterior and interior. Red painted decoration exterior. US400,8. M pinkish-orange medium. Abundant large and small grey grits. Creamy slip exterior and interior. Dark red painted decoration exterior and rim. US400,7. M orange. Grey grits, some large, sparse small white grits. Light brown slip interior. Burnt on the outside. US400,31. C light orange. Abundant white and grey grits, some large; vegetal temper and rare large shell inclusions. Light brown slip interior. Burnt on the outside. US400,28. C light orange. Large and medium grey grits, sparse white grits. Light brown slip interior. Slightly burnt on the outside. US400,32. C light orange. Abundant white and grey grits, some large. Light brown slip interior. Burnt on the outside. US400,33. C light orange. Large grey grits, small white grits. Light brown slip interior. Burnt on the outside. US400,29. C orange. Abundant large and medium grey grits. Light brown slip exterior and interior. US400,30. C orange. Large and medium grey grits, sparse white grits; rare vegetal temper. Light brown slip interior. Burnt on the outside. US400,27. M reddish. Small white and grey grits. Traces of dark brown slip on lateral surfaces. Long handled bowl. US400,36 = Cl57. Loom weight obtained from a small wadi pebble. US400,37 = S316. Complete shell belonging to the Caridoidea family. US400,38 = Sh44. Almost complete shell belonging to the Pectinidae family. Pierced in antiquity. US400,39 = Sh43. M light grey with grey core near the base. Small white grits; sparse vegetal temper. Light beige slip exterior. US318,1. M brown, badly fired. White grits, some large; vegetal temper. Pale red slip exterior, slightly burnished, dark grey slip interior. US331,1. M reddish with brown inner section. Red and white grits. US258,3. M pinkish-brown. Red grits, occasional white grits. Smoothed exterior. US258,1. M/C orange. Red and dark grits. US258,4.

3 . a complete sequence for the early iron age of central oman

-

131

plates

2

1

3

4

5

6 7

8 9 10

12 11

15

13

16

14

17

18

19

P late 28 – Basement: US400, 318, 331, 258. Scale 1:3.

132

chiara condoluci , michele degli esposti , carl phillips

P late 29 – Basement: US276, 242, 313, 330. C orange. Dark grits and red grits. US276,6. C orange. Dark grits. Smoothed on upper side. US276,4. C light brown. Red and grey grits; occasional vegetal temper. US276,1. C orange. Dark grey grits. Smoothed on upper side. US276,3. C red-brownish. Abundant red and grey grits, some large. Incised herringbone decoration. Thick burnt sediment inside reservoir. Long handled bowl. US242,1 = Cl83. 6. M brown. Abundant black and white grits, sparse red grits; vegetal temper. Light brown-beige slip exterior and rim, transition to red on interior. US313,1. 7. M/F brown with grey core. Occasional white, red and grey grits; vegetal temper. Dark red-brownish slip interior, red exterior. US313,2. 8. F red. Small black and red grits, sparse large white calcareous grits; vegetal temper. Beige slip exterior, gradually turning to light brown-pinkish on interior, moving downward. US330,2. 9. M/F red-brownish. Tiny white grits, red grits; occasional vegetal temper. Red slip interior and exterior, dark red and black painted decoration exterior. US330,3. 10. M orange red. Abundant white, red and black grits; vegetal temper. Light brown slip exterior and interior. US330,1. 1. 2. 3. 4. 5.

3 . a complete sequence for the early iron age of central oman

-

133

plates

1

2

3

4

5

7 6

8

9

10

P late 29 – Basement: US276, 242, 313, 330. Scale 1:3, except no. 5: scale 1:5.

134

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

chiara condoluci , michele degli esposti , carl phillips

P late 30 – Basement: US35. F/M light brown. Sparse white and red grits. Light buff-brown, smoothed exterior and interior. Incised decoration exterior. US35,21. F/M brown. Occasional white and red grits. Light orange-brown, smoothed exterior and interior. ‘Shaved bottom’. US35,90. F reddish. Small white grits. Pinkish, smoothed exterior and interior. US35,41. F brown. Occasional white grits. Brown, smoothed exterior and interior. US35,28. F brown. Occasional white grits. Light brown, smoothed exterior and interior. US35,91. F brown. Occasional white grits. Orange-brown, smoothed exterior and interior. ‘Shaved bottom’. US35,62. F brown. Rare white grits, some large. Orange-brown, smoothed exterior and interior. ‘Shaved base’. US35,92. F/M orange-brown. Red and white grits. Red-orange, smoothed upper exterior and interior. US35,63. F/M brown. Occasional red and grey grits. Light brown, smoothed exterior, orange, smoothed interior. ‘Shaved’ bottom. US35,67. F/M brown. Abundant red and grey grits; vegetal temper. Light brown-buff, smoothed exterior. String cut base. US35,61. F reddish. Small and medium white and red grits. Pinkish slip exterior and interior. US35,22. F pinkish. Small white grits, rare mica. Pinkish, smoothed exterior and interior. US35,25. F brown. Small white grits. Pinkish, smoothed exterior and interior. US35,27. F reddish. Small and medium white grits. Light red, smoothed exterior and interior. US35,30. F reddish. Small and medium white grits. Light brown slip exterior, light red, smoothed interior. US35,31. F reddish. Small and medium white grits, rare large white grits, medium red grits. Light red, smoothed exterior and interior. US35,56. F reddish. Small and medium white grits, small grey grits. Pink, smoothed exterior and interior. US35,71. F reddish. Small white and grey grits. Pale red, smoothed exterior and interior. US35,26. F pinkish. Small white grits, rare red grits. Pinkish-white, smoothed exterior and interior. US35,32+193. F reddish. Small and medium white and red grits. Pink slip exterior and interior. US35,23. F pale brown. Small white and red grits. Pale brown exterior, pinkish-white, smoothed interior. US35,36. F reddish. Small gray grits. Light red, smoothed exterior and interior. US35,54. F pinkish. Small white grits, rare mica. Pink, smoothed exterior and interior. US35,24. F reddish. Small and medium white grits. Pink, smoothed exterior and interior. US35,55. F brown. Small white grits, rare mica. Pink, smoothed exterior and interior. ‘Shaved’ bottom. US35,53. M brown. Abundant small grits. Buff slip exterior and interior. US35,87. M orange-brown. White and red grits. Red orange slip exterior and interior. US35,82. M brown. Abundant black grits. Buff-light brown slip exterior and interior. US35,78. M/F brown. Brownish red and white grits. Red slip exterior and interior. Light burnishing exterior. US35,68.

3 . a complete sequence for the early iron age of central oman

-

plates

135

3 2

1

6

11

4

7

13

17

16

9

8

12

22

10

15

14

19

18

21

5

20

24

23

26

27

28

29

P late 30 – Basement: US35. Scale 1:3.

25

136

1. 2. 3. 4. 5. 6. 7. 8.

chiara condoluci , michele degli esposti , carl phillips

P late 31 – Basement: US35. M light brown. Red and black grits; sparse vegetal temper. Light brown slip exterior and interior. US35,83. M/C brown. Abundant red and black grits. Buff slip exterior and interior. US35,79. C dark brown with black core. Abundant white grits; vegetal temper. Light pink-brownish slip exterior and interior. US35,85. C brown-orange. Abundant red and white grits. Red slip exterior and interior. US35,58. M light brown. Abundant small red and grey grits; vegetal temper. Light buff-brownish, self-slipped exterior and interior. US35,19. M orange-brown. Abundant small red and grey grits. Buffed slip exterior. Incised decoration exterior. US35,81. M brown. Occasional brownish-red and grey grit. Orange-brown slip exterior, light brown interior. US35,76. M Reddish. Dark and white grits, large red grits. Light red slip exterior, Incised and relief decoration. US35,93.

3 . a complete sequence for the early iron age of central oman

-

137

plates

1

2

3

4

5

6 8

7

P late 31 – Basement: US35. Scale 1:3.

138

1. 2. 3. 4. 5.

chiara condoluci , michele degli esposti , carl phillips

P late 32 – Basement: US35. M reddish. Abundant white grits; sparse medium red and grey grit. Pinkish slip exterior, light red, smoothed interior. US35,2. M reddish. Small white grits, medium red grits. Red slip rim interior and base, red slip exterior; dark reddish-grey rim exterior and spout. US35,1. M brownish. Abundant small, medium and large red grits, small white grits. Light red slip exterior and interior. US35,65. M light brown. Small white, medium red and grey grits. Red slip slip exterior and interior. US35,3+8. M/C light brown. Abundant red and grey grits; vegetal temper. Buff-light brown slip exterior and interior. US35,4.

3 . a complete sequence for the early iron age of central oman

-

139

plates

3

1

2

4

P late 32 – Basement: US35. Scale 1:4.

5

140

chiara condoluci , michele degli esposti , carl phillips

P late 33 – Basement: US35, 11, 12. C pinkish. Occasional large red and grey grits; vegetal temper. Red slip exterior surfaces. Incised chevron decoration. Burning traces on reservoir. Long handled bowl. US35,7 = Cl42. 2. M/C pale pink with grey core. Red and white grits. Burning traces on reservoir. Long handled bowl. US35,94 = Cl52. 3. M/C reddish. White and dark grey grits. Light red slip exterior surfaces. High relief decoration, with ram protome at the handle’s end. Long handled bowl. US35,95 = Cl15. 4. M/C pale pink. Dark and red grits. Pale red slip external surfaces. Incised herringbone decoration. Deep groove in section at the end of the handle. Traces of burning all over. Long handled bowl. US35,97 = Cl41. 5. M/C dark red. Dark and white grits. Reddish brown slip upper surface. Handle’s end shaped as a snake head, with eyes portrayed as oval depressions with two pointed circles in the centre; nose with two nostrils is slightly projecting; deep groove in section at the end of the handle. Long handled bowl. US35,98 = Cl16. 6. M light brown. Dark grits. Impressed circles decoration on handle. Deep groove in section at the end of the handle. Long handled bowl. US35,96 = Cl12. 7. M/F pale yellow. Dark grey grits, some large. Smoothed exterior and interior. US11,20. 8. F dark orange. Rare white grits, some large, rare small red grits. Reddish slip exterior and interior, smoothed exterior. US12,19. 9. M light brown. Small and medium white grits, rare red grits, occasional mica; abundant vegetal temper. Beige-orange, smoothed exterior, orange, smoothed interior and rim. Rotation marks on interior (slow wheel?). US12,18. 10. C grey. Abundant grey and white grits, occasional red grits, mica; vegetal temper. Pink, smoothed exterior, grey slip on rim interior. US12,17. 1.

3 . a complete sequence for the early iron age of central oman

-

141

plates

1

2

3

5

6

4

8

9

7

10

P late 33 – Basement: US35, 11, 12. Scale 1:4, except nos 7-10: scale 1:3.

142

chiara condoluci , michele degli esposti , carl phillips

P late 34 – Basement: US13, 14, 251, 272, 241. M orange-brown. Red and grey grits. Light brown, smoothed exterior and interior. String cut base. US13,8. M light brown. Sparse red and grey grits, some large. Red brown slip exterior and interior. US13,4. C reddish-brown. Large, medium and small black and white grits, rare red and brown grits; rare vegetal temper. Buff- pinkish exterior surfaces, light brown interior. Incised herringbone decoration. Burnt traces all over. Long handled bowl. US13,3. 4. C dark orange. Small white grits, rare dark grey grits; vegetal temper. Light brown, smoothed exterior, dark red slip interior. Rotation marks on interior (slow wheel?). US14,2. 5. C light orange. Occasional large and small white grits, small and medium red grits, rare small grey grits. Pale brownpinkish slip exterior and interior. Incised decoration on rim. US14,4+US14,5. 6. M greenish-yellow medium. Grey grits. US251,1. 7. C brown. Dark grits. Brownish, smoothed upper surface. US251,2. 8. C grey with orange inner section. Dark and red grit grits. US272,2. 9. C pinkish-brown. Dark grits, occasional vegetal temper. Traces of burnt exterior and interior. US272,1. 10. C orange. Red grits. US241,1. 11. C/M brown-reddish. Dark and white grits; vegetal temper. Traces of light brown slip exterior and interior. Incised herringbone decoration. Long handled bowl. US272,3 = Cl90. 12. Conical shaped spindle-whorl. US241,2 = S178. 1. 2. 3.

3 . a complete sequence for the early iron age of central oman

-

143

plates

2

1

3

5

4

6 7 8

9

10

11

P late 34 – Basement: US13, 14, 251, 272, 241. Scale 1:3, except no. 12: scale 1:2.

12

144

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

chiara condoluci , michele degli esposti , carl phillips

P late 35 – Basement: US9, 7, M174. F pink. White grits, sand temper. US9,67. F brown. Rare small white and dark grey grits, rare mica. Pinkish, smoothed exterior and interior. US9,55. F reddish-pink. Small and medium white grits, rare mica. Pinkish, smoothed exterior and interior. US9,60. M/F orange. Red and white grits. US9,69. M greenish-yellow. Dark and red grits. Traces of brown slip on interior. US9,66. C orange. Abundant red grits, rare grey and small white grits; vegetable temper. Pale brown slip exterior, light red slip interior. US9,1. M orange. White and red grits, rare mica. Pale red slip exterior and interior. US9,3. M/C orange. White and red grits, some large, dark grey grits, rare mica. Pale yellow, smoothed exterior, pink, smoothed interior. US9,11. C orange. Red, dark grey, and white grits, sand temper. Pale yellow slip exterior and interior. US9,7. M brown. Rare red and grey grits. US9,68. C dark red. Red and white grits, rare large red grits, sand temper. Light grey slip exterior, red-brownish interior. US9,9. M orange. Large and medium red grits, sand temper. Light red slip exterior and interior. US9,34. M pinkish-beige. Rare small white and red grits, dark grey grits, sand temper. Pale yellow, smoothed exterior and interior. US9,23. Soft stone biconical spindle-whorl. US9,81 = S176. F pink. Rare small white and black grits. Pinkish slip exterior, reddish-brown interior. US7,22. M orange. Rare small white and red grits, sparse mica. Reddish brown slip exterior, pale red interior. US7,12. C pinkish with grey core. Rare small white grits, large red grits. Whitish exterior and interior. Incised herringbone/chevrons e decoration. Burning traces on reservoir. Long handle bowl. US7,19.

3 . a complete sequence for the early iron age of central oman

1

-

145

plates

3

2

4 5 6

7

8

9

13

10 12

11

16

15 14

17

P late 35 – Basement: US9, 7, M174. Scale 1:3, except no. 14: scale 1:2.

146

chiara condoluci , michele degli esposti , carl phillips

P late 36 – Burnt Building: US475, 386, 428, 427, 474, 437, 368, 95. F brown-reddish with red core. Small white and red grits; vegetal temper. Brown slip exterior, red-brownish interior. Red painted decoration interior. US475,1. 2. F red. Small white grits; vegetal temper. Light red slip interior and exterior (altered to white). US475,2. 3. M brown. Small dark grey grits, sparse white grits, some larger; vegetal temper. Dark red-brownish slip interior and exterior. US475,3. 4. F brown. Some vegetal temper. Light orange-brown slip exterior and interior. Burnished exterior with horizontal brown streaks. Iron III. Intrusive. US386,3. 5. F brown. Sparse vegetal temper. Light orange-brown slip exterior and interior. Burnished exterior horizontal brown streaks. Iron III. Intrusive. US386,2. 6. F/M brown. Tiny brown grits; sparse vegetal temper. Orange-brown slip exterior and interior. String-cut base. US386,8. 7. F/M brown. Occasional tiny brown grits; vegetal temper. Brown slip exterior and interior. US386,7. 8. M orange-brown. Brown and white grits; vegetal temper. Orange-brown slip exterior and interior. US386,1. 9. F brown. Occasional large brown grits. Orange-brown slip exterior and interior. Red-brown painted decoration exterior and rim. US428,2. 10. F light brown. Occasional brown and white grits. Light brown-whitish slip exterior and interior. US428,1. 11. F light brown. Occasional small white grits; vegetal temper. Light brown slip on exterior and interior. US428,3. 12. F brown. Occasional tiny brown grits. Orange-brown slip exterior and interior. Brown-black painted decoration interior. String-cut base also shaved. US427,4. 13. F brown. Occasional brown grits. Black slip exterior and interior. US427,2. 14. F orange-brown. Occasional brown grits. Brown slip exterior and interior. US427,5. 15. M orange. Brown grits; vegetal temper. Light brown-black slip exterior. US427,1. 16. F brown. Occasional brown and white grits. Light orange-brown slip exterior and interior. US427,3. 17. F red. Tiny white and red grits, sparse grey grits. Red slip interior and exterior. US474,2. 18. F brown. Tiny white grits, red grits; vegetal temper. Thick red slip interior, red exterior. US474,1. 19. C orange-brown. Abundant red-brown grits; vegetal temper. Brown slip exterior and interior. US437,3. 20. F brown. Occasional large grits; sparse vegetal temper. Black slip exterior. US437,2. 21. M orange with grey core. Occasional brown grit; vegetal temper. Red-orange slip exterior. orange interior. US437,1. 22. F orange brown. Occasional grits and tiny white inclusions. Orange brown slip exterior and interior. Burnished. Incised signs decoration (graffito). US368,1. 23. F light brown. Occasional small white and white grits; sparse vegetal temper. Light brown slip exterior and interior. Dark brown painted decoration exterior. US368,2. 24. M orange-brown. Sand temper. Orange-brown slip exterior and interior. US95,1. 1.

3 . a complete sequence for the early iron age of central oman

1

-

147

plates

3

2

4

5

7

6

8

10

9

12

11

14

13

15

17

16

18

19

21

23

20 22

24

P late 36 – Burnt Building: US475, 386, 428, 427, 474, 437, 368, 95. Scale 1:3, except nos 20-21: scale 1:4.

148

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

chiara condoluci , michele degli esposti , carl phillips

P late 37 – Burnt Building: US381, 378. F orange-brown. Small brown and white grits. Brown slip exterior. US381,6. M brown. Abundant small brown grits; vegetal temper. Brown slip exterior and interior. String-cut base. US381,5B. F orange. Small white grits; sparse vegetal temper. Dark brown-black slip exterior and interior. US381,8. F brown. Small brown and white grits; sparse vegetal temper. Brown slip exterior and interior. String-cut base. US381,7. M brown. Tiny brown grits; sparse vegetal temper. Dark brown slip exterior and interior. US381,3. M brown. Small brown and white grits; vegetal temper. Light brown slip exterior and interior. Slightly burnished exterior. US381,1. M orange. Brown grits, smaller white grits; vegetal temper. Light brown slip exterior and interior. Incised decoration exterior. US381,2. M orange. Small brown grits; vegetal temper. Dark brown slip exterior, lighter brown interior. Incised decoration exterior. US381,4. C brown. Abundant black and brown grits; vegetal temper. Orange-brown slip exterior and interior. US381,10. M light brown. Small black and brown grits. Light brown slip exterior and interior. US381,9. C brown. Abundant black and brown grits; abundant vegetal temper. Orange-brown slip exterior and interior. US381,11. F/M brown. Small brown grits and occasional large grits; sparse vegetal temper. Light brown slip exterior and interior. String-cut base also shaved. US378,6. F grey-black (due to firing?). Occasional small grits; sparse vegetal temper. All surfaces blackened. US378,7. M/F brown. Small white grits; vegetal temper. Orange-brown slip exterior and interior. US378,8.

3 . a complete sequence for the early iron age of central oman

1

2

-

149

plates

3

4

5

7 6

9 8

10

12

13

11 14

P late 37 – Burnt Building: US381, 378. Scale 1:3, except no. 11: scale 1:4.

150

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

chiara condoluci , michele degli esposti , carl phillips

P late 38 – Burnt Building: US378. M light brown. Occasional small brown grits; vegetal temper. Light brown slip exterior and interior. Slightly darker brown painted decoration exterior. US378,13. M orange. Abundant small brown grits; vegetal temper. Orange slip exterior and interior. US378,4. F brown. Sparse tiny white grits; vegetal temper. Brown slip exterior and interior. Iron III. Intrusive. US378,1. F brown. Sparse white grits; vegetal temper. Red-brown slip exterior and interior. US378,5. M brown. Abundant small black and brown grits; vegetal temper. Dark brown-black slip exterior and interior. US378,2. F light brown. Occasional small brown grits. Black-brown slip exterior and interior. US378,3. M orange-brown. White grits; vegetal temper. Dark brown slip exterior. US378,17. M brown-green. Rare brown grits; abundant vegetal temper. Partially vitrified. May have had a light brown slip. US378,14. F light brown-whitish. Occasional brown grits; vegetal temper. Brown slip exterior and interior. US378,12. M orange. White grits; vegetal temper. Light brown slip exterior. US378,15. M light buff-green. Grey-black grits. Smoothed exterior. US378,16. C brown. Large white grits; vegetal temper. Brown slip exterior. US378,10. Very C orange-brown with grey core. Abundant large brown and black grits; vegetal temper. Orange-brown slip exterior and interior. Incised decoration on raised cordon. US378,11. C brown with thick grey core. Abundant brown grits; vegetal temper. Orange-brown slip exterior and interior. US378,9. C red-brownish with grey core. Large amount of red, grey grits, sparse vegetal temper. Orange brown slip exterior and interior. US378,18.

3 . a complete sequence for the early iron age of central oman

-

151

plates

2

4 3

1

6 7

5

9

10

8 11

12

13

14

15

P late 38 – Burnt Building: US378. Scale 1:3.

152

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

chiara condoluci , michele degli esposti , carl phillips

P late 39 – Burnt Building: US394. F orange-brown. Occasional tiny white grits; vegetal temper. Orange-brown slip exterior and interior. US394,21. F orange-brown. Sparse tiny black grits, occasional large white grits; vegetal temper. Orange-brown slip exterior and interior. Iron III. US394,1. M brown with black core. Abundant small grits. Orange-brown slip exterior and interior with abrasive texture due to grits. US394,19. M/F orange-brown. Occasional small white grits; sparse vegetal temper. Light orange slip exterior. Burnished, with horizontal burnished streaks. Iron III. US394,2. M light brown. Small brown grit; vegetal temper. Dark brown slip exterior and interior. Dark brown painted decoration interior. US394,11. M brown. Brown grits, tiny white grits; sparse vegetal temper. Orange-brown slip exterior and interior. US394,20. F brown. Vegetal temper. Light brown slip exterior and interior. Slightly burnished. US394,12. F light pink-brown. Abundant exploded yellow grits; abundant vegetal temper. Light brown slip exterior, faint black painted decoration exterior. US394,6. F brown. White and brown grits; vegetal temper. Black-brown slip exterior and interior. US394,7. F light brown. Occasional small black grits; vegetal temper. Orange-brown slip exterior, black-brown interior. Iron III. US394,3. F brown. Occasional white grits; vegetal temper. Dark brown slip exterior and interior. Iron III. US394,4. F orange. Abundant small white grits; vegetal temper. Light brown slip interior (exterior eroded). US394,9. F orange-brown. Occasional small brown grits; vegetal temper. Orange-brown slip exterior and interior. Iron III. US394,5. F light brown. Occasional tiny brown and white grits; sparse vegetal temper. Light brown slip exterior and interior. US394,8. F grey-black (due to firing?). Small white grits; vegetal temper. Dark-grey exterior and interior. Exterior possibly burnished. Incised decoration. Wheel-made? US394,22. M orange-brown. Abundant vegetal temper. Brown slip exterior. Light brown painted decoration interior. US394,24. M/F orange. Occasional brown grits; vegetal temper. Orange-brown slip exterior. Black-brown painted decoration exterior. US394,14. M dark brown with black core. Brown grits; vegetal temper. Light brown slip exterior and interior. US394,10. F orange-brown. Occasional small white grits; vegetal temper. Dark brown slip exterior and interior. String-cut base, also shaved on exterior. US394,13. M light brown. Small white grits; sparse vegetal temper. Light yellow-brown slip exterior. US394,18. C orange-brown. Abundant large brown grits. Red-brown slip exterior and interior. Incised decoration on raised cordon. US394,15. Very C brown with thick black core. Abundant large brown grits; vegetal temper. US394,23.

3 . a complete sequence for the early iron age of central oman

-

153

plates

1 2

3

4 5

6 7

10 9

12

11

8

13

14

18 15

16

17

19

20

21

22

P late 39 – Burnt Building: US394. Scale 1:3.

154

chiara condoluci , michele degli esposti , carl phillips

P late 40 – Burnt Building: US394, 384, 367. C orange-brown. Abundant medium brown grits; vegetal temper. Orange slip exterior, top of rim and upper part of rim interior. US394,16. 2. C orange-brown. Abundant small brown, black and white grits; vegetal temper. Brown slip exterior. Incised decoration on slightly raised band around collar. US394,17. 3. M/F brown with black core. Small brown and white grits; abundant vegetal temper. Light brown slip exterior and interior. Exterior possibly burnished. US384,1. 4. C orange-brown. Abundant large brown grits. Black-brown slip exterior and interior. US384,2. 5. F brown. Brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. US367,2. 6. F brown. Occasional tiny white grits; vegetal temper. Orange-brown slip exterior. US367,4. 7. F orange. Small white grits. Orange-brown slip exterior and interior. US367,3. 8. M brown. Occasional brown grits and sparse vegetal temper. Brown slip exterior. US367,9. 9. M orange. Occasional large brown grits; vegetal temper. Brown slip exterior. Black painted decoration exterior. US367,5. 10. F/M orange with thin grey core. Abundant white grits; vegetal temper. Red-brown slip exterior. String-cut base. US367,8. 11. M/C brown. Abundant brown and large white grits; vegetal temper. Dark brown slip exterior. US367,6. 12. M brown. Brown grits; abundant vegetal temper. Brown slip exterior. String-cut base. US367,7. 1.

3 . a complete sequence for the early iron age of central oman

-

155

plates

2

1

4 3

5

6

7

9

8

11

10

P late 40 – Burnt Building: US394, 384, 367. Scale 1:3, except nos 1-2: scale 1:4.

12

156

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

chiara condoluci , michele degli esposti , carl phillips

P late 41 – Burnt Building: US377. F brown. Sparse vegetal temper. Brown slip exterior and interior. Red-brown painted decoration exterior and interior. US377,22. F light brown. Abundant small brown grits; sparse vegetal temper. Dark brown slip exterior and interior. US377,16. F orange-brown. Occasional tiny brown and white grits; sparse vegetal temper. Light brown slip exterior and interior. String-cut base, also shaved on exterior. US377,15. F grey-brown (due to firing?). Tiny brown and white grits; sparse vegetal temper. Burnt interior. US377,17. F orange-brown. Tiny black grits, occasional larger white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Iron III. US377,14. M light brown. Abundant small brown and white grits; vegetal temper. Orange-brown slip exterior and top of rim. US377,18. M/C light brown. Abundant black, brown and white grits. Brown slip exterior and interior. US377,19. M brown. Small brown grits, sparse white grits; vegetal temper. Red-brown slip exterior. Black painted decoration exterior. US377,9. M green-brown. Occasional small brown grits; vegetal temper. Light brown slip exterior. Black-brown painted decoration exterior. US377,12. M orange. Occasional white grits; vegetal temper. Red-brown slip exterior. Black painted decoration exterior. US377,13. M brown. Abundant small black and brown grits, occasional large white grits; vegetal temper. Light brown slip exterior and interior. Incised decoration exterior. US377,5. F orange-brown. Sparse vegetal temper. Orange slip exterior and interior. US377,20. M orange. Occasional white grits; vegetal temper. Light brown slip exterior, smoothed interior. US377,11. M light brown. Abundant small brown and white grits; sparse vegetal temper. Red-brown slip exterior and interior. US377,10. M light brown. Abundant small white grits; sparse vegetal temper. Dark brown slip exterior and interior. Incised decoration exterior. US377,6. M black (overfired). Abundant small and big white grits; sparse vegetal temper. US377,1. M orange. Abundant small black grits, occasional white grits; vegetal temper. Orange slip exterior and rim. US377,21.

3 . a complete sequence for the early iron age of central oman

-

157

plates

1 2

4

3

6 5

7

11 8

12

9

13

14

10

15

16 17

P late 41 – Burnt Building: US377. Scale 1:3.

158

chiara condoluci , michele degli esposti , carl phillips

P late 42 – Burnt Building: U377, 108. Very C orange-brown. Abundant large brown grits and small white grits. Black-brown slip exterior. Orange-brown slip interior. Incised decoration on raised cordon. US377,2. 2. M brown. Abundant small brown and white grits; vegetal temper. Orange-brown slip exterior and interior. US377,3. 3. C orange-brown. Large red, brown and white grits; abundant vegetal temper. Orange-brown slip exterior and interior. US377,7. 4. C orange-brown. Large red, brown and white grits; abundant vegetal temper. Orange-brown slip exterior and interior. US377,8. 5. C light brown with thick grey core. Brown and white grits; abundant vegetal temper. Orange-brown slip on upper surface and sides. Underside uneven. US377,4. 6. F brown. Black slip exterior and interior. US108,3. 7. F orange-brown. Orange slip on exterior and interior. Graffito on exterior. US108,2. 8. M brown. Sparse vegetal temper, sparse sand temper. Black brownish slip exterior and interior. Dark painted decoration exterior and interior. US108,4. 9. F light brown. Black-brownish slip exterior and interior. ‘Stained’ painted decoration exterior. US108,1. 10. M grey-black. Sand temper. Very dark brown slip exterior, black interior. Incised decoration on rim. US108,6. 11. F brown. Sparse white grits; vegetal temper. Orange-brown slip exterior and interior. US108,7. 12. C orange-brown. White grits and vegetal temper. Brown slip exterior. Graffito on exterior. US108,5. 1.

3 . a complete sequence for the early iron age of central oman

-

159

plates

2

1

3

4

5

6 7

8

9

10

11

12

P late 42 – Burnt Building: U377, 108. Scale 1:3, except nos 1-5: scale 1:4.

160

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

chiara condoluci , michele degli esposti , carl phillips

P late 43 – Burnt Building: US392, 389, 366. F orange. Abundant tiny white grits with sparse big red inclusions (exploded); sparse vegetal temper. Dark brown slip exterior and interior. Red painted decoration interior and exterior. US392,2. F orange-brown. Small brown and white grits; sparse vegetal temper. Brown slip exterior and interior. US389,3. M brown. Occasional white and brown grits; vegetal temper. Orange-brown slip exterior. Incised decoration exterior. US389,2. F orange-brown with thin grey core. Occasional large brown grits; sparse vegetal temper. Brown slip exterior and interior of neck. Incised decoration exterior. US389,6. F brown. Sparse tiny white grits; vegetal temper. Brown slip exterior and interior. Iron III. US389,4. M/C brown. Abundant brown grits; vegetal temper. Orange-brown slip exterior and interior. US389,1. M/C brown. Brown and white grits; vegetal temper. Brown slip exterior and interior. US389,5. F/M brown. Small black and white grits; sparse vegetal temper. Brown slip exterior and interior. US366,14. F brown. Small white and black grits. Light brown slip exterior and interior. US366,24. F light brown. Occasional white grits; sparse vegetal temper. Light orange-brown slip exterior and interior. String-cut base also shaved. US366,19. F light brown. Occasional small brown grits; sparse vegetal temper. Light brown slip exterior and interior. US366,20. F orange-brown. Sparse small brown grits, occasional large black grits. Light brown-whitish slip exterior and interior. US366,23. M brown. Brown and white grits; vegetal temper. Red-brown slip exterior and interior. String-cut base. US366,18. M brown. Brown and black grits; vegetal temper. Orange-brown slip exterior and interior. US366,21. M brown. Small black grits; vegetal temper. Light brown slip exterior and interior. US366,13. M orange-brown. Abundant small white grits, rare larger brown grits; vegetal temper. Orange-brown slip exterior and upper part of interior. US366,17. M light yellow-brown. Black grits; vegetal temper. Dark brown-black slip exterior. US366,22.

3 . a complete sequence for the early iron age of central oman

-

161

plates

2

1

4

3

5 6

8

9

10

7

12

11

15 14

13

17

16

P late 43 – Burnt Building: US392, 389, 366. Scale 1:3.

162

chiara condoluci , michele degli esposti , carl phillips

P late 44 – Burnt Building: US366. F/M yellow-brown. Rare tiny white grits; vegetal temper. Red-brown slip exterior. Black painted decoration exterior. US366,16. 2. M orange-brown. Occasional tiny white and brown grits; abundant vegetal temper. Brown slip exterior. Black painted decoration exterior. US366,15. 3. M light brown. Occasional small white grits; vegetal temper. Light brown slip exterior. US366,8. 4. M brown. Small black grits; vegetal temper. Brown slip exterior and interior. US366,7. 5. M brown. Small black and white grits; vegetal temper. Orange slip exterior and interior. US366,11. 6. M light brown. Small white grits; vegetal temper. Brown slip exterior and interior. US366,4. 7. M light brown. Occasional small black grits; vegetal temper. Light brown slip exterior and interior. Incised decoration exterior. US366,9. 8. M orange. Small black grits; vegetal temper. Red-brown slip exterior and interior. US366,6. 9. M light brown. Abundant small black grits; vegetal temper. Brown slip exterior and interior. US366,3. 10. F/M brown. Sparse small white grits; vegetal temper. Brown slip exterior and interior. US366,10. 11. M orange. Small black grits; vegetal temper. Red-brown slip exterior and interior. Incised decoration exterior. US366,5. 12. C orange-brown. Abundant brown grits; abundant vegetal temper. Light brown slip exterior and interior. Incised decoration on raised cordon. US366,12. 1.

3 . a complete sequence for the early iron age of central oman

-

163

plates

2

1

4

5

3

6

8

7

10

9

11

12

P late 44 – Burnt Building: US366. Scale 1:3.

164

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

chiara condoluci , michele degli esposti , carl phillips

P late 45 – Burnt Building: US407, 107, 393. F brown. Occasional small brown grits; sparse vegetal temper. Brown slip exterior and interior. US407,5. F orange. Occasional small brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. US407,7. F orange. Occasional small brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. US407,6. F light brown. Occasional small brown grits. Red-brown slip exterior and interior. US407,4. F orange-brown. Tiny white grits; occasional vegetal temper. Exterior fired white. Thin, streaky brown slip/wash exterior and interior. String-cut base, also shaved. US407,8. F orange-brown. Tiny white grits; occasional vegetal temper. Brown slip exterior and interior. String-cut base, also shaved. US407,9. F/M orange-brown. Small brown grits; vegetal temper. Orange slip exterior and interior. String-cut base, also shaved. US407,11. F/M orange-brown. Tiny white grits; vegetal temper. Brown slip exterior and interior. US407,10. F brown. Occasional brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Brown painted decoration on top of rim. US407,12. F brown. Tiny black grits; sparse vegetal temper. Light brown slip exterior and interior. US407,3. M orange-brown. Occasional black and brown grits; vegetal temper. Light brown slip exterior and interior. US407,2. M light brown. Abundant small brown grits; vegetal temper. Brown slip exterior and interior. US407,1. F brown. Sparse vegetal temper. Red-brawn slip on exterior and interior. Applied snake decoration. US107,1. F brown with thin grey core. Occasional tiny white grit; abundant vegetal temper. Orange-brown slip exterior and interior. Wheel-made. Iron III. US393,1. F brown. Tiny white grits; vegetal temper. Orange-brown slip exterior and interior. Burnished, with horizontal burnished streaks. Wheel-made. Iron III. US393,12. F orange. Sparse small brown grits. Dark brown slip exterior and top of rim. Iron III. US393,14. F brown. Small brown and white grits; vegetal temper. Brown slip exterior and interior. US393,10. F/M orange. Sparse brown grits; sparse vegetal temper. Orange-brown slip exterior and upper part of interior. US393,9. F red-brown. Small white grits; sparse vegetal temper. Brown slip exterior and interior. US393,11. M red-brown. Brown and white grits; vegetal temper. Orange-brown slip exterior and interior. US393,8. M light brown with grey core. Small brown and white grits; vegetal temper. Orange-brown slip exterior. US393,13.

3 . a complete sequence for the early iron age of central oman

-

165

plates

4 1

2

5

3

6

7

8

9 10

12

11

13

15

14

16 17

18 19

20 21

P late 45 – Burnt Building: US407, 107, 393. Scale 1:3.

166

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

chiara condoluci , michele degli esposti , carl phillips

P late 46 – Burnt Building: US393, 439. C brown with grey core. Large brown grits, occasional small white grits; vegetal temper. Brown slip on upper and underside. US393,5. M orange-brown with grey core. Vegetal temper with brown and white grits. Orange-brown slip on upper and underside. US393,4. Very C orange-brown with thick black core. Large brown and white grits; abundant vegetal temper. Orange-brown slip on upper and underside. US393,3. Vary C orange-brown. Large brown and white grits; abundant vegetal temper. Dark brown slip on upper and underside. US393,2. C light brown with thick grey core. Brown grits; abundant vegetal temper. Light brown slip exterior and interior. US393,6. Very C light brown with thick grey core. Large brown grits; abundant vegetal temper. Brown slip exterior and interior. US393,7. F/M brown. Abundant small brown and white grits; sparse vegetal temper. Light brown slip exterior and interior. US439,1. F brown. Rare vegetal temper. Light brown slip exterior and interior. Red-brown painted decoration exterior. US439,5. F brown. Occasional small brown grits; sparse vegetal temper. Brown slip exterior and interior. US439,6. F/M brown. Small brown grits; sparse vegetal temper. Brown slip exterior and interior. US439,2. M brown. Brown grits; vegetal temper. Orange-brown slip exterior and interior. US439,3. M brown. Abundant small black grits; vegetal temper. Light brown-whitish slip exterior and interior. US439,4. C orange-brown. Abundant large brown grits; vegetal temper. Light brown slip exterior and interior. US439,7.

3 . a complete sequence for the early iron age of central oman

-

167

plates

2

1

3

4

5

6

8

9

7

11 12

10

13

P late 46 – Burnt Building: US393, 439. Scale 1:3, except nos 1-6: scale 1:4.

168

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

chiara condoluci , michele degli esposti , carl phillips

P late 47 – Burnt Building: US385, 371, 370. F light brown. Tiny white grits; sparse vegetal temper. Light brown slip exterior and interior. Iron III. US385,4. F orange-brown. Small brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished, with horizontal burnished streaks. Wheel made. Iron III. US385,2. F orange-brown. Small brown and white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished, with horizontal burnished streaks. Wheel made. Iron III. US385,1. F orange. Occasional small brown grits; vegetal temper. Orange-red slip exterior. Wheel made. Iron III. US385,3. F orange-brown. Tiny brown grits; sparse vegetal temper. Orange-brown slip exterior and interior. Black painted decoration exterior. String-cut base also shaved. US371,4. F orange. Brown grits; sparse vegetal temper. Orange-brown slip exterior and interior. String-cut base, also shaved. US371,5. F/M orange-brown with thin black core. Brown and white grits; vegetal temper. Dark brown slip exterior and interior. US371,3. F/M brown with black interior. Abundant small brown grits; vegetal temper. Orange slip exterior. Wheel-made. Iron III. US371,2. F orange-brown with thin grey core. Brown and white grits; abundant vegetal temper. Brown slip exterior and interior. Wheel-made. Iron III. US370,8. F brown. Sparse small brown grits. Orange slip exterior and interior. Burnished. Iron III. US370,5. F brown with thin grey core. Tiny white grits; vegetal temper. Black-brown slip exterior and interior. Wheel-made. Iron III. US370,1. F orange-brown with thin grey core. Occasional brown grits; sparse vegetal temper. Black-brown slip exterior and interior. Iron III. US370,2. F brown. Brown and white grits, some large; vegetal temper. Brown slip exterior and interior. US370,3/buttress. M orange with thin grey core. Brown and white grits; vegetal temper. Orange-brown slip exterior and interior. US370,4.

3 . a complete sequence for the early iron age of central oman

-

169

plates

1

2

3 4

5

7

6

8

9

10

12 11

13

14

P late 47 – Burnt Building: US385, 371, 370. Scale 1:3.

170

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

chiara condoluci , michele degli esposti , carl phillips

P late 48 – Burnt Building: US370, 376, 390. F orange. Occasional brown grits; vegetal temper. Orange slip exterior and interior. Wheel-made. Iron III. US370,9. F orange-brown with thin grey core. Brown and white grits; abundant vegetal temper. Orange-brown slip exterior and interior. Wheel-made. Iron III. US370,7. F brown with black interior. Brown and white grits; vegetal temper. Orange-brown slip exterior. Wheel-made. Iron III. US370,6. M orange-brown. Small white grits; vegetal temper. Brown slip exterior. US376,3. M orange. Occasional brown and white grits; vegetal temper. Brown slip exterior and upper part of neck interior. Evidence for beginning of bridge-spout. US376,5. C brown. Abundant brown grits; vegetal temper. Light brown slip exterior and interior. US376,4. M light brown. Abundant brown grits; vegetal temper. Light brown slip exterior and interior. Incised decoration exterior. US376,1. F orange-brown. Brown and white grits; vegetal temper. Brown slip exterior and interior. Incised decoration exterior. US376,2. Very F brown. Occasional tiny white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished. Iron III. US390,2. Very F brown. Occasional tiny white grits; sparse vegetal temper. Orange-brown burnished slip exterior and interior. Iron III. US390,6. Very F brown. Occasional tiny white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished, with horizontal burnished streaks. Iron III. US390,4. Very F brown. Occasional tiny white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished, with horizontal burnished streaks. Iron III. US390,5. Medium orange. Small brown and white grits; vegetal temper. Light brown slip exterior and interior. US390,8. F orange-brown. Occasional tiny brown grits; vegetal temper. Light brown slip exterior. Black painted decoration exterior. US390,1.

3 . a complete sequence for the early iron age of central oman

-

171

plates

2

1

3

6

4 5

8

7

9

10

12

11

13 14

P late 48 – Burnt Building: US370, 376, 390. Scale 1:3, except nos 1-3: scale 1:4.

172

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

chiara condoluci , michele degli esposti , carl phillips

P late 49 – Burnt Building: US390, 99. F brown. Small brown and white grits; vegetal temper. Brown slip exterior and interior. US390,9. F orange. Small white grits temper. Black slip exterior and interior. US390,3. Very F brown. Occasional tiny white grits; sparse vegetal temper. Orange-brown slip exterior and interior. Burnished. Iron III. US390,10. F orange. Abundant small white grits, sparse vegetal temper. Light brown slip exterior and interior. US390,11. F brown. Vegetal temper with abundant small brown and white grits. Brown slip exterior and interior. US390,7. M/C orange with grey core. Orange-brown slip exterior and interior. US390,12. C orange reddish with grey core. Abundant red and grey grits and vegetal temper. Traces of light red slip exterior. US390,16. C light brown-whitish. Abundant small black and brown grits; vegetal temper. Red-brown slip exterior and interior. US390,14. C light brown. Abundant black and brown grits. Traces of brown slip. US390,13. C orange with thin grey core. Abundant red grit and sparse vegetal temper. Red orange slip on upper surface. US390,15. F buff orange. Tiny white grits and sand temper. Red slip exterior and interior. Iron III. US99,11. F buff orange with thin grey core. Rare white grits. Sparse sand and vegetal temper. Dark red slip exterior and interior. Iron III. US99,12. F/M brown orange with thin grey core. Grit inclusions and sand temper. Iron III. Red slip exterior. US99,13. F/M buff orange. Red and grey grits. Sparse white inclusions and vegetal temper. Reddish brown slip exterior and interior. Iron III. US99,14. F/M dark orange. Several grits, sparse vegetal temper. Dark red slip exterior. Iron III. US99,15.

3 . a complete sequence for the early iron age of central oman

-

173

plates

2

3

1

5

4

6

7

8

9 10

11 13 12

14

15

P late 49 – Burnt Building: US390, 99. Scale 1:3, except nos 6-10 scale 1:4.

174

chiara condoluci , michele degli esposti , carl phillips

P late 50 – Burnt Building: US99. C orange-brown with grey core. Brown grits; vegetal temper. Orange slip exterior and interior. US99,2. M green-yellowish. White grits. Dark slip exterior and interior. US99,9. C orange-brown with thick grey core. Red-brownish grits; vegetal temper. Red-black slip on upper surface. US99,7. M orange-brown. Sparse small white grits. Light brown slip exterior. Wheel made. US99,1. M orange. White grit; vegetal temper. Red slip exterior and interior. US99,8. C orange. Dark grey grits; vegetal temper. Reddish slip exterior. US99,10. C orange-brown with grey core. Brown grits; vegetal temper. Orange slip exterior and interior. Incised decoration exterior. US99,6. 8. C red-brown. Abundant red and white grits; vegetal temper. Red-brown exterior; incised decoration exterior. US99,4. 9. C orange-brown. Abundant red and white grits; vegetal temper. Orange-brown slip exterior and interior. US99,3. 10. C orange-brown with grey core. Brown grits; vegetal temper. Orange slip exterior and interior. Incised decoration exterior. US99,5. 1. 2. 3. 4. 5. 6. 7.

3 . a complete sequence for the early iron age of central oman

-

175

plates

2 1 3

5

4

6

7

8

9

10

P late 50 – Burnt Building: US99. Scale 1:4, except nos 2, 4-6: scale 1:3.

176

chiara condoluci , michele degli esposti , carl phillips

P late 51 – Burnt Building: US 87/86. M brown. Brownish-red grits; vegetal temper and sparse sand. Brown-black slip exterior and interior. US87,4. M brown. Brownish-red grits; vegetal temper and sparse sand. Brown-black slip exterior and interior. US87,2. M brown. Brownish-red grits; vegetal temper and sparse sand. Brownish slip exterior and interior. Shaved bottom. US87,1. F/M dark orange. Red grits and sand temper. Traces of brownish slip exterior and interior. Iron III. US86,1. M brown. Brownish-red grits; vegetal temper and sparse sand. Black slip exterior and interior. US87,3. M light brown. Brownish-red grits; vegetal temper and sparse sand. Smoothed exterior and interior. US87,5. M/F orange. Sparse grits and sparse vegetal temper. Red slip exterior and interior. Iron III. US86,2. F orange brown. Tiny red and grey grits, rare vegetal temper. Red slip exterior and upper part of interior. Iron III. US86,3. 9. F dark orange. Tiny white grits and sparse vegetal temper. Red slip exterior and interior. Iron III. US86,4. 10. F orange. Sand and vegetal temper. Traces of creamy slip on exterior. Iron III. US86,5. 11. Half complete shell button. US87,10 = Sh24. 12. C orange with grey core. Abundant red grits. Light brown slip exterior. Incised decoration exterior. US87,6. 13. C orange with thick grey core. Abundant grey grits (large size) and tiny white inclusions. Abundant vegetal temper. Light red slip on upper surface. Applied decoration with dot impressions (possible snake tail). US87,8. 14. C brown. Abundant red and grey grits. White slip exterior, black interior. Incised decoration exterior. US87,7. 1. 2. 3. 4. 5. 6. 7. 8.

3 . a complete sequence for the early iron age of central oman

1

177

plates

4

3

2

5

-

7

6

8

9

10

11

12 13

14

P late 51 – Burnt Building: US 87/86. Scale 1:3, except nos 12-14: scale 1:4.

178

1. 2. 3. 4. 5. 6. 7. 8.

chiara condoluci , michele degli esposti , carl phillips

P late 52 – Burnt Building: US 89, 88. M/C light brown. Abundant grey grits and sparse vegetal temper. Light brown slip on upper surface. US89,1. F brown. Small white grits; sparse sand temper. Orange-brown slip exterior and interior. Brown painted decoration interior. US88,1. F buff with thin grey core. Sparse tiny white grits. Dark red slip on interior and exterior. Wheel made. Iron III. US88, 4. F brown orange. Sparse tiny white grits. Red slip exterior and interior. Iron III. US88,5. F buff. Rare white inclusions and sand temper. Reddish slip exterior and interior. Iron III. US88,6. F/M dark orange. Red and grey grits. Sparse sand and vegetal temper. Reddish brown slip exterior. Iron III. US88,7. F brown. Small white grits. Orange-brownish slip exterior and interior. US88,2. C brown with grey core. Abundant large brownish-red grits, medium grey grits. Traces of red-brown slip on upper surfaces. Incised decoration exterior. US88,3.

3 . a complete sequence for the early iron age of central oman

-

179

plates

2

1

3

4

5

6

7

8

P late 52 – Burnt Building: US 89, 88. Scale 1:3.

4. Other areas of excavation enrica tagliamonte , michele degli esposti , carl phillips

T he

central part of the site

(A rea 4)

To the east of wall M26 and to the south of wall M67, both containing the massive mud-brick platform of Area 1 (Terrace I), more than a quarter of the site’s upper part was distinguished as Area 4 (pl. 53), connected to the lower tower by a stone zig-zag staircase, the construction of which is contemporary with the renewal and extension of the outer wall during the HSIII phase, although some late rearrangement of the upper flight of steps, if not its construction, occurred during the Islamic re-occupation of the site. Southeast of Area 4, Area 8 was investigated to a lower extent, mainly in connection with the study of the outer wall’s structure (see Chapter 5), but it is generally reputed to have hosted the extension of the same structures unearthed in Area 4 (see fig. 10). The precise layout of Area 4 remains rather difficult to reconstruct, largely due to massive ancient collapses1 and to the substantial impact on the ancient structures derived by the excavation of several deep pits and huts foundations, connected with the Islamic occupation that developed on the hill during at least two distinct periods (see Chapter 11). Nevertheless, several structures and features datable to the Iron Age, namely to the HSI phase, can be isolated. They comprise numerous walls that clearly followed the same axes as those discovered in the area covered by Terrace I. How these walls relate to Terrace I and what part they performed in the overall plan of the site's architecture is concisely discussed here, while more details on the construction technique of the platform system as a whole are given in Chapter 5. Against wall M26, at least three terraces can be outlined – Terraces II, III and IV –, not quite parallel to it, their eastern retaining walls being respectively M33, M35 and M59. Their inner structure, where recognizable (Terrace IV), comprised relatively well-laid mud-bricks, suggesting that inner compartments could have formed their supporting framework, consistently with Terrace I’s structure. Starting from walls M35 and M59 stands a series of walls that extend eastwards and follow the downward slope. This arrangement of walls does not seem to be meant to form rooms: the compartments that they define appear instead to be packed with mud-brick lumps and capped with well defined mud-bricks, thus forming one further series of terraces along the eastern side of the upper mud-brick platform (fig. 50). This is even more evident when the compartments are smaller and entirely defined by mud-brick walls, as is the case for walls M95, M96, M97 and M98.2 The bricks that form the upper surface of the terraces are also often laid along the same axes as those that form the main mud-brick platform, and in any case they follow the orientation of the walls that support the terrace to which they be-

As one moves eastward in Area 4, a considerable build-up (approximately 2.5 metres) of later deposits, collapsed walls and accumulated debris had to be removed during the excavation. 2 These features can thus be considered an example of the characteristic ‘compartment technique’ discussed in Chapter 5. 1

182

enrica tagliamonte , michele degli esposti , carl phillips

long. In some instances, below the upper well-laid capping layer, not just mud-brick lumps and clay were discovered, but further layers of complete mud-bricks, in these cases with a different orientation than the retaining walls of the compartment. The overall impression is that the wide mud-brick platform west of M26 (Terrace I) and the mud-brick terraces east of it are all part of a single building project (fig. 51). In exposing the plan of the Iron Age mud-brick walls in Area 4 a peculiar L-shaped void, approximately 4.25 x 1.90 metres in extent and 1.20 metres deep, was uncovered (fig. 52), delimited by walls M96, M98 and M120. This space was filled by the loose sand deposit US298 that notably contained no fewer than 20 complete, small carinated bowls and a few more pots (pls 54-55). Among them a globular jar is worth mentioning, distinguished by a tubular spout below the rim and by the presence of small horizontal lugs along the rim (pl. 55/2), for which parallels at Husn Salut are remarkably known from US454, one of the so-called foundation deposits that are part of the outer wall substructure (see Chapter 5). Remarkable is also a bowl’s flat rim that bears a bichrome (red and black) metopal decoration that comprises red or black, solid coloured panels and small, black dotted panels (fig. 53/a; F igure 50 – Mud-brick capping of Terrace II, immediately pl. 55/3). Two other bichrome sherds were discovered to the northeast of wall M26. in the adjacent layer US297, that actually covered the bottom, well-laid mud-brick surface of the compartment delimited by walls M95, M96, M97 and M98 (mentioned above) (fig. 53/b; pl. 54/31-32). From Husn Salut, the only comparable sherd comes unfortunately from a late, mixed context, US173 (fig. 54). Parallels for these sherds are not known from the Oman peninsula to the authors. In particular for the use of bichrome decoration and its general layout they can be compared with two sherds with black and maroon decora-

F igure 51 – Schematic profile across the upper mud-brick platform and the mud-brick terraces east of M26.

4 . other areas of excavation

183

F igure 52 – The L-shaped void discovered in Area 4.

tion collected from the surface at Sulaimanzai and Malazai in Baluchistan.3 This is particularly the case for sherd 297,1 (fig. 53/b) that comes from a comparable bowl and shows the same lozenge pattern of the mentioned Baluchi sherds, although it displays small dots as a filling pattern instead of a uniform surface. Significantly, these two Baluchi sites were dated, on the basis of comparisons with the cultural sequence of Pirak, to the mid-second millennium BC.4

a

b F igure 53 – Bichrome sherds discovered in US298 (a) and US297 (b). F igure 54 – Richly decorated potsherd from late, mixed context US173.



De Cardi 1983: fig. 8/2, 13. De Cardi 1983: 17; see also Jarrige and Enault 1976: pl. 9/9, showing a combination of dot-filled triangle, although monochrome.

3 4

184

enrica tagliamonte , michele degli esposti , carl phillips

To these same Baluchi sherds was compared a single fragment discovered at Tepe Yahya;5 however, similarities appear less strict between it and the Baluchi examples. Given the emphasised horizontality of its decoration’s layout it could indeed provide a better comparison for the small sherd 297,2 (fig. 53/b). On the other hand, its suggested date poses some problems: discovered in a context belonging to Tepe Yahya Period II, it therefore dates to after 500 BC.6 While this could somehow find an explanation in the case of US297, if one envisages that the corresponding ‘compartment’ or ‘premise’ could have remained open for a long time (as apparently happened for the space between walls M118 and M199, see below), such a late date is clearly conflicting with the radiocarbon determination obtained for US298. In fact, from a charred wood sample collected from US298 a radiocarbon date was obtained that ranges 2026-1701 BC at 95.4 % of confidence (fig. 55). When considered together with other dates that are available for contexts linked to the early construction of the site, this date stands out as a probable outlier, possibly deriving from the burning of an older piece of wood. At the same time, however, it provides further evidence pointing to a particularly early foundation of the site, as discussed in Chapter 10. Therefore, parallels from Baluchistan are to be reckoned as the most plausible so far, their date F igure 55 – Probability distribution for close to the middle of the second millennium BC also fitting with the US298 (Lab. code 14Fi0909, 14Fi0911). construction of Husn Salut’s monumental structures. From US298 also came the fragment of a soft stone spouted bowl (see Chapter 6, no. 17) and two long-handle bowl handles, both showing a horizontal groove at their distal end, that is supposed to represent the mouth of some animal (fig. 56). While one, Cl142, is decorated with an incised chevron pattern and can thus be compared to numerous specimens from the site that are meant to represent a stylized snake body, Cl144 is undecorated but has two broken features at both sides of its ‘mouth’. Without further details, it is not possible to determine whether it represents a horned mammal, possibly a ram like some other examples from Husn Salut, or a horned viper, as discussed in Chapter 2.

F igure 56 – Long-handle bowl fragments Cl142 (top) and Cl144 (bottom) from US298, both bearing horizontal groove at their distal end, representing a snake mouth.



Magee 2004: fig. 5/13. Magee 2004: 75.

5 6

4 . other areas of excavation

185

In addition to pottery vessels and the mentioned objects, US298 also contained a large quantity of animal bones. These included sheep/goat remains and some larger animal, notably camel (camelus dromedarius).7 It is evident that the nature of this deposit cannot be linked to domestic activities such as trash dumping, but is rather akin to the series of so-defined ‘foundations deposits’ that were recognized as the intentional fills of a number of compartments forming the core of the massive outer wall and other structures (see Chapter 5). This lends further, strong support to the interpretation of the different ‘rooms’ of Area 4 as structural elements sustaining upper terraces rather than actual premises. Some more doubts persist about the destination of the space delimited by mud-brick walls M118, M119, and M248. Removing the filling between these walls, a brown loam layer including Iron Age III pottery (US729) was discovered in the east half of the room, below two isolated, well preserved mud-bricks (US735). A more compact deposit (US730) occupied the other half of the room and was characterised by a more conspicuous presence of similarly dated pottery, including a complete tulip bowl (fig. 57). Further below, the washed deposit US731 (25 cm thick and mainly located in the southwestern corner), covers a well-laid floor made of mud-bricks arranged in a five by four pattern (US734). The presence of washed material above the floor seems to denote an exposition to weather for a given period, unlike from what appears to be the case for the other compartments in the area. This space could thus have been in use as an actual room for so remarkably long a time-span to still be in use during the HSIII phase, as indicated by the presence of additional Iron age III pottery in US369, another filling of this room (i.e. fig. 58). South of wall M248 a similar space/room probably existed, as suggested by the presence of mud-brick floor US737. However, this floor is only preserved in the western part of this space, the possible extension of which lies beyond the excavation’s limit. A similar pattern of terraces appears to exist on the southern side of the mud-brick platform, beyond wall M67. There, a number of walls also follow the main axis of the mud-brick compartment foundations of Terrace I, and

F igure 57 – Complete tulip-bowl discovered in US730, between walls M118, M119 and M248.



Rasile 2011; see also Chapter 8.

7

186

enrica tagliamonte , michele degli esposti , carl phillips

they are crossed by further walls built at right-angles to them. It is clear that these walls do not form rooms and are filled with mud-bricks that form terraces like those on the eastern side of the site. And once again it was seen that the orientation of the bricks is the same as found throughout the site, showing it to be a single, planned operation. Moreover, the further south one moves the lower the walls are founded, as they follow the underlying natural slope of the hill. This is consistent with the fact that the axes of the terraces on either side of the main mud-brick platform appear to be determined to some extent by the contours of the hillside, unlike the axis of the main Terrace I, whose intentional orientation might have some other significance. Approaching the outer wall that encloses the upper part of the site, it is also evident that the same walls defining the terraces become articulated with it, forming its supporting framework. This is particularly striking when the western end of the site is considered (Area 5, fig. 10), with substantial mud-brick walls arranged in a radial layout (see also Chapter 5). The situation appears to change in the easternmost part of the area, close to the main outer wall that separates it from the lower tower (fig. 59). Here, a northeast-southwest mud-brick wall, M101, seems to mark the eastern limit of the compartments and terraces system and a good, Iron Age surface, US727, was reached east of it, delimited by walls M102 and M190. US727, damaged by later excavations in F igure 58 – An almost complete Iron Age III jar from the area, was scattered with a few burnt traces and buried US369, characterized by hollow feet. beneath a remarkably thick accumulation of decayed and washed remains. Surface 727 was instead associated to the surviving lower part of the L shaped stone wall M190, in turn containing US742, likely to be recognized as the surviving part of an original mud deposit, largely removed by the excavation of the large, Early Islamic pit US741. This pit also dismantled most of M190 as well. The latter originally stood at a remarkably higher elevation, with a superstructure that likely included an upper mud-brick part. West of wall M190’s end, a passage was left between it and wall M101; the limited extension of the deep trench does not allow the reconstruction of the site’s earlier layout north and east of M190. However, some further deep excavation provided a few more details. Although the opposite end of M190 was not reached, due to the massive debris and later mud-brick structures under which it lay buried, it was anyhow possible to show that it is covered by M150, which has clearly to be assigned to a later phase. One can thus suppose that M190 actually abutted against the inner face of the original outer wall, later modified and remarkably enlarged. Together with M151, which abuts it at a right angle, and with M120, M150 was possibly defining a large, later phase squared area realized above the dismantled walls M190 and M102.8 Excavation against wall M150 also revealed that it was covered by a mud filling (US743) that can be dated to the HSIII phase on the basis of the materials retrieved inside it. US743 also covered a layer of sandy brown loam, in turn sitting on top of M190 – additional evidence for the early date of the latter.



No available evidence allows us to distinguish between a date for this re-arrangement into HSII or HSIII phase.

8

4 . other areas of excavation

187

F igure 59 – View of Area 4 and its easternmost part (looking northeast). In the background, L shaped wall M190 is visible, running below later constructions (M150 in particular).

T he

outer walls and the main tower

Linked to the investigation of Area 4’s original layout was also the excavation of two small trenches in its southern part, and of other trenches along the northern portion of the perimeter wall, the latter connected with each other (pl. 53). The results were indeed useful for the reconstruction and dating of the different building phases of the massive enclosure that encompasses the site, at least in its eastern and northern parts, and of its relation with the main tower that projects onto the plain. Excavation along the northern external wall (Trench 1 in pl. 53) revealed a sequence of massive intentional fills, part of a building technique that involved their use in connection with a framework of containment walls. These walls (M100, M232, M237, M239, M242) belong to the earliest phase of construction (HSI): the mentioned deposits were instead laid down in order to fill the space between them and M9, the outer stone wall currently marking the perimeter of the site and belonging to its last main building phase (HSIII). The earlier walls, stratigraphically covered by the mentioned fills, are the remains of the original surrounding walls of the site, remarkably well preserved and still surviving almost unscathed to a height of c. 3.5 metres, clearly as a beneficial effect of the protection provided by the later stone wall. Where they have been exposed down to their footings, a no good, structured surface was found. Indeed, their layout at right angles to each other,9 with limited spaces in between, could suggest that these areas were not meant for a proper occupation, but rather hint at the possible presence of recesses and buttresses along the outer perimeter of the site in its earlier phase. The excavation of a small trench along wall M242 showed that some additional mud-brick and stone features were leaning against the footings of this wall and of M239, directly on the bed-



M237 is perpendicularly leaning against M100; M239 is connected to M237 at a right angle to it.

9

188

enrica tagliamonte , michele degli esposti , carl phillips

rock (fig. 60); they could be interpreted as small buttresses, contributing to the reconstruction of the complex layout of the early outer walls. The majority of the mentioned fills consist of distinctive yellow-greenish stone chips (US601, US662, US664, US698) that can be associated in the different sequences with mud (US639, US690, US708, US747), collapsed mudbricks (US693), and washed deposits (US663, US602). Although the same type of chip fill was found to have been used in different periods, the collected materials allow a distinction between early and late deposits. In fact, apart from early Islamic sherds revealing the very latest re-arrangements, a chronological determination can be based on the presence in the remaining deposits of the distinct type of pottery known as Burnished Maroon Slipped Ware (BMSW), characteristic of the Iron Age III period, as well as a small but significant number of lathe-turned stonevessels.10 The date commonly accepted for the presence of this pottery in the Oman Peninsula,11 considered alongside the radiocarbon date obtained for an original floor level of Building 1 (US23, see fig. 22), allowed the confident distinction between HSII and HSIII architectural phases outlined in Chapter 2. Particularly significant was the discovery of wall M242, connected at right angle to M239 and going towards the northeast (fig. 60). In fact, the removal of the ubiquitous small green-yellowish chips (here distinguished as US698) revealed the presence of stone wall M247, cutting M242 and running parallel to the external face of wall M9 (fig. 61). M247 is founded directly on US733 (brown loam, small and medium sizes stones and pottery sherds), that in turn covers another layer (US738) characterised by the presence of a large amount of sherds. While they are mainly in a coarse ware belonging to the early Iron Age period (Iron Age II), several can be dated to the Iron Age III period (again, BMSW), consequently dating the construction of wall M247 into HSIII phase as well. A small trench was dug between M247 and the outside stone wall of Husn Salut, that is, M9, near the stairs leading to the main tower. The removed filling (US712) is composed of very hard mud, light grey in colour, poured between the two walls, that finally represent the two opposite faces of one and the same substantial wall: M9/M247. The fragments found inside this layer all belong to the Iron Age period. The excavation of these two trenches confirms that the later, perimeter stone walls of the site were built still during the Iron Age period, and marked a complete change in the external appearance of the northern part of Husn Salut; the earlier walls, made of mud-bricks above stone foundation, were then covered by several fills of mud and greenish-yellowish stone chips. However, extended excavation of the upper fill contained between the opposite stone skins of M247/M9 showed how it was all but homogeneous, and the compact deposit mentioned above was not present further to the northwest, in front of M239. Moreover, the collected material clearly indicates that in some places the upper part of the now massive and largely chaotic backfilling between M247 and M9 is related to Islamic period re-arrangements, which incorporated the lower surviving part of the ancient outer wall M9 as a foundation. Additional stratigraphic evidence of the relative chronology of these walls concerns the long stone wall M208. It is a single faced, low wall that runs parallel to M9/M247 and contains the upper fills to the opposite side of the earlier mud-brick walls, now buried. In fact, it is founded above fill US664, a deposit mainly composed of small stone chips, which covers the early wall M232, the former sitting atop the uneven surface US666 and the latter with its foundations cutting through it (fig. 62). As Trench 1 was extended southeast of wall M237, more of the green-yellowish stone chip fill was removed; this was shown to be covering two eroded mud-brick walls related to the first Iron Age phase (M121, M122), which delimitated another mud-filled compartment. Immediately to the southwest of mud-brick wall M102 the thick mud deposit US747 was also removed, and recognized as part of the construction fill; on this side, it covers US698 and seems to be delimited by M102, M121 and M237. US698 stood on top of US708, another mud fill partially excavated in a small trench along M237, which seems to cover US733, the foundation level for the outer wall M9/M247. On the basis of all these data, therefore, the stone outer wall that currently surrounds the site, and that actually is a double face wall whose opposite stone skins were given two different names (M9 the outer, M247 the inner), can



See also the discussion in Chapter 12 and Chapter 8. See Magee 2005a.

10

11

4 . other areas of excavation

F igure 60 – The original stone and mud-brick perimeter walls of the site, unearthed in Trench 1 of Area 4. Note the possible small reinforcement buttress built against M242’ footings, near the corner with M329 (to the left).

F igure 61 – View of the late stone wall M247, representing wall M9’s inner face and clearly cutting early phase’s M242.

189

190

enrica tagliamonte , michele degli esposti , carl phillips

clearly be dated to the Iron Age III period (c. 600-300 BC), thus providing a chronological threshold for the transition from the Husn Salut II to the Husn Salut III phase.12 The connection between the upper part of the site and the main tower projecting onto the plain was also investigated thanks to the trenches discussed here. In fact, the poor state of preservation of the uppermost (third) flight of the stone stairs coming up from the tower suggested that it could be the outcome of a late building (or rebuilding). This was confirmed by the removal of these steps and by the excavation of the underlying foundation deposit (US723), mostly made up of loose loam mixed with stones and lumps of mud. US723 also contained numerous sherds: although the majority was datable to the Iron Age period, fragments of Abbasid pottery and Late sgraffiato ware serve as a terminus post quem for the construction of the upper flight of stairs as it was preserved. It was also made clear how the latter and its foundation deposit had cut the previous layers and features belonging to the Iron Age phase, such as wall M102 and US698. What remained dubious was the relative chronology of the second flight of stairs, the one turning west at a right angle to the lower flight, the latter being one that physically crosses wall M9, starting inside the main tower (fig. 63). A trench cut below US723, in order to entirely expose the outer F igure 62 – A view of wall M242 and the later face of wall M247, revealed that this is just the continuation stratigraphy. Late wall M248 rests above US664. of the wall that supports the second flight of the staircase from the west. At the same time, this also proved that the construction of the main tower and its outer wall M39 has to be dated to the Husn Salut III phase, simultaneously with the construction of the new perimeter wall M9/M247 embracing the upper part of the site; this was actually already suggested by the apparent bonding of their footings, noted at the southern junction between the two walls. It is also worth noting that, while the most part of wall M39 was erected directly above the ancient plain surface, its south western portion, close to M145, was laid down inside a foundation trench whose backfilling was composed of the same stone chips repeatedly mentioned above. The opposite end of Trench 1, i.e. the one towards the east, allowed: 1) a better view on the substantial stone wall M4; 2) an investigation of its relations with M9; 3) the sequencing of a number of other walls (fig. 64). In particular, cleaning along the outer face of the perimeter wall M9/M247 showed the presence of US600, a deposit of loam and medium sized stones, clearly cutting US4-5, the massive fill laid between M4 and M9 and thus datable to the Husn Salut III phase, consistent with the technique described above. The medium compact fill of big and medium stones mixed with loam and lumps of mud US681, covered by US5, is also related to the same late Iron Age phase and the erection of wall M9/M247. Wall M4 was built partly on a compact mud layer (US677) and partly on the bedrock, that here outcrops as superimposed, multi-colour shale layers ranging from yellow to red. This geological formation is typical and clearly visible on the hills surrounding Salut (fig. 64 bottom left). Another fill laid above the earlier parts of the main outer wall is US601, again a thick deposit composed of small green-yellowish stone chips mixed with small and medium sized stones. To the northwest, US601 is delimited by the

12 It must be underlined here that this reconstruction is valid for the whole eastern part of the site, while no evidence for such a later phase was collected in its western portion, where the outer wall displays a supporting structure hinged on a pattern of radial walls.

4 . other areas of excavation

a

191

b

F igure 63 – The staircase leading upward from the main tower: first flight crossing M9 (a), and the corner with the perpendicular, second flight (b).

F igure 64 – The massive early wall M4, abutted by M8 while M9 runs northeastward (M9 higher rows, badly damaged, already removed). M8 and M9 are connected in their lower course. Bottom right, detail of M4 foundation above deteriorated bedrock levelled with a clayish deposit (US677) when needed.

192

enrica tagliamonte , michele degli esposti , carl phillips

long containment wall M208, already mentioned above; to the east, it is contained by the surviving part of the mudbrick filling US630, related to the Iron Age phases of the site and possibly representing the badly preserved continuation of wall M100. By virtue of the Islamic pottery sherds retrieved inside it, US601 can be dated to a late occupation of the site, and its deposition is clearly connected to the erection of the small containment walls M207, M225 and M66, laid above it. Significantly, the partial removal of US601 revealed two underlying, very compact, mud and stone layers, that are likely to be related to the Iron Age building phases (US602, US639). Furthermore, the cleaning of Terrace I near M4 showed that wall M233, the northwestern prosecution of wall M26, was partially cut and rebuilt using upright stone slabs in order to obtain a rectangular niche. The establishment of M227 took place after this re-arrangement, probably in the same phase as the construction of walls M230 and M226, distinguished by an orientation that is inconsistent with that of the adjacent, Husn Salut I phase walls. Between them, the well preserved floor-surface US667 was also unearthed, whih can be tentatively dated to the HSIII phase (fig. 65).13 Excavation in Trenches 2 and 3 was less extensive than in Trench 1. Nonetheless, the results go some way in corroborating the data discussed above. In fact, a long, low stone wall was discovered, that spans the whole distance from the eastern limit of Trench 2 to the western limit of Trench 3. This wall, M236, strongly resembles the late wall M20814 on the opposite side of the site, and is similar to wall US540 more to the southwest (see Chapter 5). Therefore, its more plausible date appears to be consistent with the late, Islamic re-arrangement of the outer walls. It contained the upper part of a sequence of deposits that also include the typical green-yellowish-stone chips (here distinguished as US685), that in turn covers deeper fills made of more compact mud, mud-brick lumps and stones (US 688, US696, US697, US700). US688 also serves as the foundation level for M236, and M244 as well, a massive mud-brick wall only partially exposed against the western limit of Trench 3. At a deepest elevation, still in Trench 3, US688 covers the stone wall M245, that testifies to the existence of an earlier building phase. This is also shown in Trench 2, where the massive mud-brick wall M240 stands above the compact mud layer US700, laid to bury an earlier stone wall, only seen for a very limited extension.

F igure 65 – Surface US667 between late Iron Age walls M226 (right) and M230 (left), delimited by perpendicular wall M231, in turn surmounted by a late Islamic wall. In the background the niche in wall M26 is visible.

13 These features are likely to be dated to the HSIII phase, on the basis of the Iron Age III pottery discovered in their proximity. However, the mixed and not-sealed nature of the excavated deposits make some caution necessary. 14 The wall is made of three to four rows of stones bound with thin layers of mud mortar, but it is not well structured.

plates

4 . other areas of excavation

-

plates

P late 53 – General plan of Area 4 with indication of the additional test trenches discussed in the text.

195

196

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

enrica tagliamonte , michele degli esposti , carl phillips

P late 54 – Carinated and simple cups from US298, and bichrome bowls’ sherds from US297. F light orange-brown. Occasional tiny white grits, sand temper; rare mica. Smoothed exterior. US298,1. F light orange-brown. Occasional tiny white grits, sand temper; rare mica. Light brown slip exterior and interior (traces). US298,2. F red-brown. Small black grits. Brown, smoothed exterior and interior. String-cut base. US298,3. F light orange-brown. Occasional tiny white grits, sand temper; vegetal temper and rare mica. Smoothed exterior. Stringcut base. US298,4. F light brown. Small black grits, sand temper. Brown slip exterior and interior. String-cut base. US298,5. F red. Small red grits; sparse vegetal temper. Red-brown slip exterior and interior. String-cut base. US298,6. F light brown-whitish. Sand temper. Light brown-whitish slip exterior and interior. US298,7. F light brown-whitish. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,8. F orange-brown. Occasional tiny white grits, sand temper; rare mica. Light brown slip exterior and interior. String-cut base. Shaped bottom. US298,9. F light brown with grey core. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,10. F/M orange-brown. Red grits, sand temper. Light brown slip exterior and interior. US298,11. F light brown. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,12. F light brown-whitish. Black and white grits. Smoothed exterior and interior. US298,13. F light brown-whitish. Black and white grits. Smoothed exterior and interior. US298,14. F light brown. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,15. F light brown-whitish. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,16. F light brown. Occasional tiny black grits, sand temper. Black slip exterior and interior. US298,17. F light red-brown. Sand temper. Red-brown slip exterior and interior. Shaped bottom with irregular base. US298,18. F light orange-brown. Small white grits, sand temper. Light orange-brown slip exterior and interior. US298,19. F/M brown. Black and white grits, sand temper. Black-orange rough exterior and interior. US298,20. F light brown-whitish. Sand temper. Light brown-whitish slip exterior and interior. String-cut base. US298,21. F brown, badly fired with black core. Tiny white grits, sand temper. Grey-brown exterior. US298,22. F light brown. Occasional tiny black grits, sand temper. Black slip exterior and interior. US298,23. F light brown. Occasional tiny black grits, sand temper. Black slip exterior and interior. US298,24. F light brown. Occasional tiny black grits, sand temper. Black slip exterior and interior. US298,25. F/M orange-brown. Abundant small red-brown grits. Light brown slip exterior and interior. US298,26. C grey-brown. White grits, sand temper. Blackened exterior and interior rim. US298,27. F orange. Orange-brown slip exterior and interior. String-cut base. Shaped bottom. US298,29. F/M orange. Red grits; vegetal temper. Bright orange slip exterior and interior. US298,28. F/M badly fired. Exploded white grits, sand temper. Blackened around exterior rim and interior. Mis-shaped during firing, brittle fabric. US298,30. F brown. Sparse sand temper. Pale brown smoothed exterior and interior. Dark red painted decoration exterior. Shaped exterior. US297,1. F brown. Sparse sand temper. Pale brown smoothed exterior and interior (possibly same bowl of 297,1). Black and dark red painted decoration exterior. US297,2.

4 . other areas of excavation

-

197

plates

2

1

4

3

5

6

9

7

10

11

18

17

25

16

20

19

23

22

21

12

15

14

13

8

24

27

26

28

31

29

32 30

P late 54 – Carinated and simple cups from US298, and bichrome bowls’ sherds from US297. Scale 1:3.

198

1. 2. 3.

enrica tagliamonte , michele degli esposti , carl phillips

P late 55 – US298: globular jars with lugs on the rim and spout, and bowl with bichrome decoration. F light brown. Sand temper; rare mica. Light orange-brown, smoothed exterior and interior. US298,32. F/M brown. White and red grits. Light olive green-brown slip exterior and interior, smoothed. US298,33. F light brown. Sand temper; rare mica. Light orange-brown, smoothed exterior and interior. Dark red and black painted decoration exterior. US298,31.

4 . other areas of excavation

-

199

plates

1

2

3

P late 55 – US298: globular jars with lugs on the rim and spout, and bowl with bichrome decoration. Scale 1:3. .

5. The Iron Age architecture of Husn Salut chiara condoluci

The uniqueness of Husn Salut is evident in many aspects of the site, such as its material culture and the results of the integrated analysis of ceramic materials, stratigraphy and absolute 14C dates (discussed in Chapters 2, 9 and 10). However, what undoubtedly makes it outstanding from the very first sight, is its architecture. In fact, it would be inappropriate to define Salut as a simple settlement, vis-à-vis its monumentality and the elaborated architectural project that forms it, which finds no comparisons among South East Arabia Iron Age sites (fig. 66). a

b

c

d

F igure 66 – A panoramic view of Husn Salut – respectively from N-NW (a), top N-NW (b), SE (c), NW sides (d) – gives the idea of the original building project and related terracing system, marked by the outer perimeter wall, which developed across the rocky slope.

202

chiara condoluci

Husn Salut was built on the most prominent point of an elongated rocky outcrop that emerges from the plain, isolated from other geographical features. Prior to human intervention, the hill would have appeared just like the other portions of the same outcrop that are still visible today, that is, a steeply inclined series of rock strata lying one above the other, sloping eastward and ending abruptly toward the west, where the hill side is in fact even steeper. This situation required an impressive amount of preparatory work, aimed at providing horizontal surfaces for the erection of its architectural features. To such a challenge, the builders’ response was the employment of a complex technique that seems not to be paralleled at other coeval sites, but for which ancient local roots can be traced, as discussed below. The construction appears to be the outcome of one single, coherent project from the beginning. In fact, the substantial archaeological stratification excavated by the IMTO team did not originate from a continuous superimposition of different, chaotic building phases, as may be the case for real archaeological tells, but it rather predominantly consisted of massive intentional fills that were deposited in order to allow the planned construction to be feasible. One main indication for this is provided by the fact that the impressive perimeter walls primarily served as containment and support for buildings and terraces erected on the top of the hill, more than for being real fortification walls.

A sophisticated

building technique : compartments foundations

The original project appears to have involved the creation of a prominent terrace on top of a hill (i.e., Terrace I, fig. 67), meant to host a few buildings erected above it and therefore being clearly visible from the distance, the outer wall being kept low enough not to hide their view. Consistently with such a project, the external walls do not embrace clusters of houses – as would be expected in the case of a fortified settlement –, but a system of platforms/ terraces that occupies an area exceeding 1500 m2. These platforms were arranged in a way that was functional to span a difference in altitude of 7 m between the lower (northeastern) and the higher (southwestern) parts of the site. The gap between the natural surface of the hill slope and the floors established atop these terraces, was filled by the mentioned, massive deposits. The platforms comprised single-faced perimeter stone walls, containing different kinds of intentional fills, and were sealed on top by mud-brick surfaces; the latter were probably visible and served as floor levels.1 The development of this terrace system followed a roughly radial pattern to the north and east of wall M26, the main northwestsoutheast wall containing Terrace I, founded directly above bedrock (see also Chapter 4). This terrace corresponds to the highest and widest area of the site, significantly the one where the actual buildings discovered at Husn Salut were concentrated – the Burnt Building, the Basement, and the Building 1 (see Chapter 2 and Chapter 3). This construction technique is not limited to the larger terraces. Indeed, it is found throughout the site, also used to build smaller compartments that are usually adjoining one another and that are delimited by mud-bricks rather than by stone walls: a single terrace can comprise several of such compartments. Also in the case of these compartments, their inner fills comprised layers of mud-bricks and mud lumps mixed with loam, topped with well-laid mud-brick rows. F igure 67 – View of Terrace I, from the northwest.

The existence of an original mud finishing on top of these bricks cannot be ruled out, although no surviving trace of it was found during the excavation. 1

5 . the iron age architecture of husn salut

203

Southwest of wall M26 for example, mud-brick walls were built directly above the rocky outcrop, delimiting quadrangular compartments averaging around 2.5x2 m in dimension. Although looking like false rooms, there is an absence of any recognizable floors and instead their inner space was filled with gravel soil, lumps of mud, loose bricks and fragments thereof, and finally topped with two or three rows of well laid mud-bricks. Thus, they actually constitute the supporting framework of Terrace I itself. The mud-bricks employed in the fill usually vary in dimension and shape, being either square (0.35x0.32x0.05 m) or rectangular (0.68x0.33x0.07 m or 0.45x0.33x0.07 m), as shown in one of the main trenches excavated through the platform. The lower surface of the mud-bricks often bears deep digital impressions that facilitates the bond with thick layers of mud mortar. Mud-bricks used for the upper surface of the platform are usually more homogeneous, with a rectangular shape. Two test trenches were excavated inside two of these compartments (pl. 56). Scanty Iron Age potsherds come from the sequence of superimposed intentional fills identified in Trench 1, below the mud-brick capping – US658, US659, US660.2 Shapes primarily belong to simple bowls with rounded rim – a basin is also present –, in fine or medium fabric, with traces of slip (pl. 62/1-4). Coarse ware is only testified by two wall fragments. No archaeological material was found in the strata filling the compartment investigated in Trench 2 (figs 68-69).3

F igure 68 – One of the compartments forming the inner structure of Terrace I (Trench 2), with retaining stone wall M26 in the foreground.

2 US658: medium soft, dark brown loam mixed with mud lumps and few fragmentary, disarticulated animal bones; US659: small stones mixed with brown-grey soft loam; US660 – thin layer above the bedrock: medium compact, stone chips and mud lumps with few fragmentary, disarticulated animal bones. 3 Below a capping made of three rows of mud-bricks, a thick layer of mud lumps mixed with mud-brick fragments, US679, covered a layer of mud lumps mixed with medium size stones, US680. The latter rested above the bedrock.

204

chiara condoluci

F igure 69 – US679, one of Trench 2’s fills, with retaining stone wall M26 in the background.

Moving upslope, where the fills’ thickness decreases as the bedrock gets closer to the occupational surfaces, a simple stone chip fill is found, covered by mud-bricks or directly underlying a trampled floor. On the highest point of the site, for example, the remains of two pre-existing Early Bronze Age graves (see pl. 1) were simply incorporated in the Iron Age construction and were found a few tens of centimetres below the Iron Age surface: they were covered by the described fill, once the majority of their stone had been removed and the already decayed structures were completely razed, leaving only little of the original contents untouched.4 In the Burnt Building area, where the original foundation layer of the four adjacent rooms 1b-4b has been exposed, the massive retaining stone wall M148 was unearthed (pl. 57 top). This contained thick, compact layers of mud lumps, as well as massive mud-brick fills that formed a protruding ‘buttress’. They developed along the southern slope of the site, toward the outer stone wall of the buttress (M154). A series of distinct mud-brick compartments of the kind described above were also built in this area, such as the one delimited by mud-brick walls M136 and M158, below room 4b (fig. 70; pl. 57). Both run parallel to the external wall, on a northwest-southeast direction. However, while wall M158 was a buried containment wall on which pillar M146 was erected above ground, wall M136 served both as a containment wall for the underground compartment and as a partition wall separating room 4b from the northwestern end of the corridor room 6b. The platform closing the Burnt Building from the northwest (SF49) was also composed of three adjacent mudbrick compartments and, immediately north of room 7b, a similar buried compartment was outlined, delimited by mud-brick walls M140 and M141. In the Basement, an area originally standing at a lower level than Terrace I and the Burnt Building, the main walls were built with mud-bricks above stone foundations, the latter sitting directly on the bedrock; beneath the original floors, layers of mud lumps and mud-bricks were equally laid to provide them with a stable preparation, at the same time levelling bedrock unevenness (fig. 71).



Condoluci and Degli Esposti 2015: 6-12; see also Degli Esposti and Phillips 2012.

4

5 . the iron age architecture of husn salut

F igure 70 – Burnt Building, room 4b, foundation mud-brick fills contained by mud-brick walls M136 and M158.

F igure 71 – Basement, room 1a, foundation mud-brick fill (US280).

205

206

chiara condoluci

However, the most extensive and impressive use of this construction method has been exposed in the southern part of the site, where a massive build up of mud-bricks came to light. This area covers around ¼ of the upper part of the site and is delimited by the curtain wall to the south (southeast and southwest), by wall M56 to the west, by wall M71/M46 to the north, and by the wall framing the entrance to the east (wall M54). Significantly, it connects with the compartment framed by walls M140 and M141, mentioned above. The difference in height that needed to be spanned between the outer wall foundation and the bedrock at the top of the hill – the latter limit corresponding to the stone wall M71/M46 – was above 5 m, implying the construction of a 7 m wide terrace (pl. 58). In the first place, the main outer wall was erected, obviously with only a single face toward the exterior, made of irregular rows of stone blocks. The framework supporting the inner structure of this massive wall/terrace structure was then created. It comprised a series of slightly curving retaining walls, made in stone or mud-bricks or the combination of the two, built following the contour lines of the hill (M82, M83, M75, M93, M169), interlocked with a series of usually larger, radial walls running perpendicular to the terrace/wall outer face, built in mud-bricks and reinforced with stone blocks (M81, M170, M167, M168, M255, M256, M257, M258 – e.g. fig. 72). This layout resulted in the formation of several adjacent compartments that were subsequently filled up with thick and compact deposits mainly comprising hard mud lumps fills, again capped with rows of well-laid mud-bricks. Among these fills, a certain

F igure 72 – Area 3, view of the curtain wall and location of the compartment delimited by mud-brick radial walls M168-M167, and containing a foundation deposit (see also fig. 78).

5 . the iron age architecture of husn salut

207

number were discovered that comprised a range of cultural materials that suggest a not merely functional interpretation (discussed below). Wall construction was likely carried out at the same pace as the progressive infilling of the compartments they defined; after a compartment was completed and sealed with mud-bricks, it could happen that the following one was arranged following a staggered pattern, so that its walls rested above the earlier compartment’s capping. Overall, it can be seen how this building method basically implies the juxtaposition of several compartments in a sort of cube construction. As mentioned above, these compartments can at first glance be mistaken for actual rooms, especially when they have slightly larger dimensions, as is the case for the one delimited by walls M80, M72, M73 and M71/M46, near the top of the area (fig. 73). However, it is worth repeating that no evidence that could be connected with a real occupation of these spaces was collected. It is highly probable that this massive terracing work provided a flat surface at its top, in all respects akin to that of Terrace I, which served as a foundation ground for the Burnt Building or, in the later phase of the site, for Building 1. Moreover, the fact that the stone retaining wall M71/M46 stands at an elevation that is almost identical to that of Terrace I’s surface seems to indicate that this area could originally constitute an extension of Terrace I itself. Whether it was a simple open area or hosted some upper building must remain a matter of speculation, given the high levels of disturbance due to ancient collapse, natural erosion, and later activities during the Islamic occupation of the site.

F igure 73 – Area 3 during excavation (2007). View of the compartment delimited by walls M71/M46-M80-M72, standing on a mud-brick massive fill. Retaining wall M75 in the background.

208

chiara condoluci

The compartment technique: carrying on a local tradition When possible parallels for the building technique described above are concerned, Husn Salut still represents a unique case in the Iron Age architecture of the Oman Peninsula. The domestic architecture of contemporaneous sites in the region shows in fact a general character of uniformity, with possible differences deriving from the type of subsistence, from building material availability, and from the necessity to conform to the morphology of the territory. It is, therefore, not unusual to find terracing and/or containment systems, both in settlements built on alluvial terraces – such as the site of Manal 1 on the western bank of Wadi Samail5 – and sites built on actual hill slopes, such as Lizq.6 Nevertheless, the complex system of retaining walls, compartments and platforms employed at Salut is something different and far more challenging, whose origins are to be tracked down in Early Bronze Age architecture. Admittedly, when the outstanding building technique witnessed at Husn Salut was first outlined and its study began, finding fitting comparisons in the local Iron Age revealed itself to be an arduous task. Therefore, looking back to a more ancient Oman peninsula appeared on the one hand necessary – as the transformations that take place during the Iron Age represents indeed the final manifestation of cultural developments started in earlier periods –, while on the other hand it meant making a leap back in time of almost one millennium, without any intermediate connection. Undoubtedly, however, this search proved worth undertaking, as essential information could be collected from the reports on a few Early Bronze Age sites excavated in the UAE, where platforms and retaining walls supporting them are witnessed. At the multi-period mound of Tell Abraq, on the west cost of the UAE, the third millennium monumental building comprised a circular stone facing ring wall, which served as containment for an interior fill, mainly constituted by mud-bricks. The wall was found still standing for a height of eight meters above ancient ground level. The internal construction comprised a mud-brick ring wall and two radial mud-brick walls with an intentional soft fill in-between.7 During the Wadi Suq period a massive stone wall two meters high, enclosing an oval area, was built on top of the earlier building. But what is more interesting for the present discussion is the construction, in the second half of the second millennium (Late Bronze Age), of a massive mud-brick platform, which covered the entire surface of the central area of the original third millennium building.8 At that time comparisons were put forward with areas outside the Oman Peninsula – namely the site of Tepe Yahya in the Iranian province of Kerman9 –, but closer parallels were soon to come to light. A similar situation was revealed at Kalba 4, on the East coast of the UAE, where the remains of an Early Bronze Age tower are surrounded by an Early Iron Age revetment wall that enclosed and contained a raised platform possibly occupied by an Iron Age I village.10 However, a more significant comparison for the compartment technique employed at Husn Salut has been more recently revealed at the prominent Early Bronze Age centre of Bat, not far from Salut, where a multi-year research programme also focused on the investigation of the internal structure of a few monumental stone towers. Specifically, the clearest instance for the use of this technique, showing a remarkable similarity to Husn Salut, is witnessed at the so-called Matariya tower, or tower 1147. Here in fact for the inner structure of the stone-made tower mud-brick architecture was primarily employed. Deep soundings unearthed a series of primary walls and perpendicular, connected or abutting walls, further intercepted by other smaller ones so as to form a structural grid that finally comprises mud-brick compartments of different size. Most importantly, these compartments were filled with rubble and/or mud-brick, resulting in the final construction of



Elmahi and Ibrahim 2003. Kroll 2013. 7 Potts 1990b: 28-33; see also Potts 1993. 8 Potts 1993. 9 Potts 1990b: 101. 10 Magee and Carter 1999: 167-168. 5 6

5 . the iron age architecture of husn salut

209

a platform.11 Two phases were identified in the construction of this mud-brick structure, the later certainly dated to the first half of the third millennium BC.12 Moreover, the use of this compartment substructure was not limited to the tower’s interior, but it was also employed in the construction of a wider platform above which the tower itself was erected, and that was delimited by an admittedly small stone wall.13 Other interesting details for a historical reconstruction of the building techniques employed at Husn Salut come from another tower excavated at Bat, tower 1146 (or Qasr al-Khafaji). Here, the main construction material is stone, including the compartment walls of its foundation framework, that were filled by hard packed, loamy mud.14 Apart from the compartments technique itself, the nature of the outer wall of the tower needs to be highlighted. In fact, it is essentially a single-faced, outer retaining wall, just as is the ‘fortification’ wall at Husn Salut, and as such suited to contain substructures and foundation fills meant to support some sort of floor/surfaces. Much closer to Husn Salut, the same is true for the ring wall of the Early Bronze Age tower Salut-ST1, investigated by the IMTO team between 2010 and 2015. Also in this case in fact, the outer stone wall only had an external good face, and contained a hardened silt and gravel fill that occupied the whole tower. Here, however, no inner partition wall was discovered, which might indicate the presence of compartment foundations.15 According to archaeologists, both Matariya and Qasr al-Khafaji towers were originally built as rounded raised platforms, and this was likely the case for Salut-ST1 as well. This should not appear far from the original aspect of Husn Salut, although at a much greater scale. Considering its overall architectonical configuration, the external stone wall mainly represented a monumentally impressive retaining wall for a series of platforms and correlated buildings atop them. Such a majestic layout is not surprising for Husn Salut, as it coalesces with other traits of the site to underline its peculiarity. A few, less strict comparisons for the compartment building technique of Husn Salut can be found also at other Iron Age sites in the Oman Peninsula. At Asima, in the Emirate of Ras al-Khaima, an Iron Age hillfort was discovered and partially investigated at site AS 97: the fortification comprised two parallel walls connected by shorter transversal walls that defined casemate-like compartments, interpreted as terracing foundations.16 Another fortified place, dated in this case after 300 BC, was partially investigated at Maisar 34, where a casemate structure comprising compartments or rooms encircles a natural rise.17 However, these later examples lack the substantial mud-brick work that conversely distinguishes Husn Salut and other earlier sites mentioned above. As shown above, the Husn Salut builders have drawn on an ancient local tradition and, at the same time, they adapted it to specific topographic constraints, with the intent of placing Husn Salut precisely in the spot that it occupies, that is, in the centre of a fertile plain and in an elevated, strategically favourable position that allows it to overlook its whole surroundings. In order to achieve this result and realise this project on top of a small hill, a wellstructured programme had to be implemented, and this suggests some form of centralised decision. Besides, the construction of the site required a substantial amount of manpower, considering that stone and mud were the main building materials. The work involved such things as: mud preparation (digging out, mixing) and transport; mud-brick production and transport; water-supply management; stone block procurement (including quarrying?) and transport; all activities that clearly required a large number of masons and workers.

11 Cable 2016: 60-63, figs 4.12, 4.13. 12 Ibid.: 63. 13 Cable 2016: 72, 75-77, figs 4.19, 4.24. In the light of this results, a re-evaluation of the structural layout of other Early Bronze Age towers first excavated more than 40 years ago would seem worthwhile, especially when considered together with the nature of the correlated deposits, as their interpretation may be susceptible to significant changes. Although this would be beyond the scope of the present work, a relevant example that can be mentioned is Hili 8, where the internal stratigraphy of the so-defined ‘rooms’ excavated inside Period I’s tower consisted of an earth and gravel fill and no floor surface (Cleuziou 1989: 63). 14 Thornton, Cable and Possehl 2013: 256. 15 Degli Esposti 2016: 668-669. 16 Vogt 1984: 139. 17 Tillmann 1981; see also Yule 2016: 55.

210

chiara condoluci

In a seemingly egalitarian culture, where social differentiation is reflected neither in domestic architecture nor in burial practices, it seems more than plausible to think that Husn Salut was a public place – most probably a cultic place – the construction of which represented a shared interest for a whole community. Finally, what is also of relevance concerning the complex compartments and terraces system of Husn Salut – and in particular of its southern area – is its explicit homogeneity in chronological terms. No material was discovered from the inner fillings of the compartments that was not datable to the early Iron Age period. On the contrary, these deposits often contained sherds of the most diagnostic shapes, such as carinated cups and spouted jars (as detailed below). From two such contexts connected with the earliest construction of the site also came two calibrated 14C dates which indicate it being established at least as early as 1300 BC, consistently with the data obtained from other spots at the site (see below and Chapters 3 and 10). However, the use of this well-structured compartment technique was not limited to the HSI phase, but it was replicated in the following main architectural period of the site, when the pre-existing buildings were backfilled and their walls used as compartment containments and partitions themselves, and the interior of the rooms was intentionally filled and capped with mud-bricks to form new foundations (see below).

Early phase foundation deposits and associated materials (Husn Salut I) Among the structural compartments described above, a certain number appear to be characterised by a sequence of internal, deliberate deposits whose nature links them to a ceremonial sphere; as such, they can be fairly referred to as foundation deposits. The southern and southeastern portion of the main outer wall (Area 3 and Area 8, see fig. 10) has revealed not only the most complex layout of the compartments technique at Husn Salut, but also the highest concentration of foundation deposits: six of them were found to the west of the site’s entrance (Area 3) and one to the east (Area 8), the latter displaying the most sizeable amount of materials (pl. 59). Extremely relevant is also the assemblage recovered from the northernmost compartment among the three that were forming the later structure of platform SF49, closing the Burnt Building from the west. The arrangement of the first group of foundation deposits almost follows one and the same contour line, which develops along the eastern part of the area, rather close to the outer wall’s face. Given their location deep in the construction layer sequence, and thus consistent with the initial phase of the work, they can be considered roughly contemporary with each other. These deposits all show a common stratigraphic configuration, encased within irregularly rectangular compartments that are delimited by retaining mud-brick walls or simple rows of mud-bricks, when not roughly shaped mud ridges, possible remains of compacted mud walls. On the inside, a series of superimposed intentional deposits – comprising loam mixed with ashes, charcoals, organic (animal bones) and/or inorganic materials (pottery, bronze) – are separated by thin layers of mud lumps and sealed with one or more rows of mud-bricks, when not simple mud. Although only one was found still almost completely sealed, all the deposits considered here were clearly distinguished from the surrounding contexts thanks to their organic component. Another similar foundation deposit was discovered in Area 4, primarily containing animal bones and pottery (see Chapter 4) Given the substantial collapse that damaged the outer wall, with a massive accumulation of washed deposits through which, during the Islamic period, huts and other features were excavated, it is possible that other foundation deposits were originally located in the higher part of the wall substructure as well, now disappeared due to erosion. Indeed, some contexts suggest this possibility, by virtue of their characteristics or their content (in terms of specific materials that could be linked to the ritual sphere). A rectangular hollow18 located in this upper part of the wall, immediately north of wall M75 and obtained in the



0.60x0.64 m, 0.30 m deep.

18

5 . the iron age architecture of husn salut

211

mud-brick capping of one of the mentioned compartments, could be tentatively connected to some ritual activity. Its fill, different from any other excavated at the site, consists of almost clean loam; an elongated, unhewn stone was also found, vertically set into this fill. Given these details and considering its location, it can be suggested that this cavity was used for pouring liquids, possibly as offerings. Interestingly, just a little further to the east, and at a slightly upper level, a soft, reddish-brown silty layer (US185) with scarce animal bones, potsherds and rodent bones (see Chapter 5) were found, that could possibly be recognized as the bad preserved remains of another deposition, possibly related to the one described above. Immediately to the east of wall M205 the discovery of a fine, half complete painted cup inside US584 (fig. 74; pl. 62/5), a compact layer comprising mud lumps, loam and mud-brick fragments, may hint at the presence of another compartment, likely delimited by walls M73, M92, M71/M46 and M205 itself. In fact, although not rare within the ceramic assemblage of the South East Arabian Iron Age, the shape and the associated painted motifs of the cup (three consecutive bands of diagonal cross-hatching, ‘running dog’ and oblique lines on the exterior, hanging rectangles on interior) are uncommon at Husn Salut. Another compartment was possibly located further east, partly delimited by stone walls M47 (west side), M259 (north side) and M209 (east side). A miniature copper/bronze axe (MB167, see chapter 7, no. 18) was in fact discovered inside a badly preserved mud lump and stone chip layer, US355, close to the corner formed by walls M47 and M259. Some general remarks can be made about the pottery collected from these foundation deposits. It mainly comprises open forms such as cups and carinated cups, bowls and large bowls, basin and large jars. Necked jars – with or without spout – constitute a minority but they are represented in all the contexts, with the exception of US451 (below). Compared to the ceramic sequence exposed for the Basement, the pottery assemblage shows remarkably little differences, considering the restricted scope of the foundation deposits context. Basins with flat triangular or slightly in-turned rim with straight wall are not represented in the Basement. The majority of sherds are in either fine or coarse fabric, often with grey core, and with a widespread use of vegetal temper. Slip ranges from brown to dark-red and orange colour, while few self-slipped sherds are present, and burnishing is rare. Painted sherds amount to 4% of the sample. A discussion on its own, presented below, is necessary for the pottery coming from the deposit located east of the entrance (US454) and from those connected with the erection of platform SF49. The description of the foundation deposits exca5 cm vated in Area 3 is given below, along a northeast – southwest direction, followed by the foundation deposit from Area 8 and by those related to the raised F igure 74 – Painted cup from US584 (Area 3). platform SF49. south of wall m 54

This compartment (fig. 75) – 1x0.80 m, 0.50 m deep – was delimited to the east by the stone clad of the outer wall, by wall M54 (north), wall M213 (east), and the large radial mud-brick wall M258 (south). The related thick deposit US451 was composed of medium soft, brown loam mixed with mud lumps and scanty traces of ashes and charcoals. Few potsherds, disarticulated animal bones, wood fragments (including several twigs), small bronze prills and slag pieces were collected. US451 was resting above a mud-brick surface. The scanty potsherds mainly comprised non-diagnostic fragments of coarse and medium fabric (pl. 62/6).

212

chiara condoluci

between walls m 93 and m 89

This compartment had an irregularly rectangular plan of 0.80x1.20 m, 1.19 m deep. It was delimited by the retaining stone walls M93 (east) and M89 (west) – which rest on the bedrock – and by other filling deposits that are part of the main construction substructure (fig. 76). These latter deposits are capped by a rectangular mud-brick feature of 0.60x0.37 m, 0.05 m thick, standing at the same elevation of the stratum sealing the deposit US284, which consisted of medium compact mud lumps mixed with sandy loam. Underneath it, the other superimposed fills, that show similar organic compositions, rest on the bedrock, and had been laid down in the following order:

F igure 75 – Area 3, small compartment southwest of wall M54.

US293. Medium soft, brown-greyish layer, composed of scattered broken mud-bricks (lying flat and showing finger impressions) mixed with loam, ashes, disarticulated animal bones, Iron Age potsherds. US285. Very soft, brown-greyish, 2-5 cm thick layer, composed of loam mixed with ashes and burnt vegetal materials. US295. Very soft, brown-greyish loam, mixed with disarticulated animal bones and Iron Age potsherds (non diagnostic). US296. Medium compact, brown-greyish, 8 to 20 cm thick layer composed of mud mixed with ashy loam and scattered medium size stones, Iron Age potsherds and disarticulated animal bones. A sample from this layer has been radiocarbon dated between 1387 BC – 1056 BC at 95% of probability (1308 – 1130 at 68.2 % of probability) (fig. 77). US299. Medium compact, brown greyish layer, composed of loam mixed with ashes, disarticulated animal bones, and Iron Age potsherds.

Pottery from these layers comprised a range of shapes including small cups, spouted jars, storage jars and lids, in fabrics that consistently vary from fine to coarse (pls 62/7-20; 63; 64/1-8). a

b

F igure 76 – Area 3, foundation deposit between walls M93-M89 (US285) during excavation (a) and at the end of the excavation, with exposed bedrock (b).

5 . the iron age architecture of husn salut

213

south of wall m 93

Located immediately south of wall M93, this deposit was badly preserved, the upper levels and its eastern part having been washed away. It was originally sealed by a massive mud lump fill partly still in situ. To the north it was delimited by the stone wall M93 while to the south by the retaining mudbrick wall M256. An original rectangular plan with a 1 m long main side can be suggested, on the basis of the layout of the surviving western side. The fills probably did not exceed a thickness of 20/25 cm. The only preserved layer, US612, was covered by a thin layer of mud lumps. It was mainly composed of ashes mixed with a low percentage of loam and scattered lumps of F igure 77 – Probability distribution for US296. mud and charcoals. Iron Age potsherds, small size bronze slag fragments, few small and unshaped bronze fragments and disarticulated animal bones were found inside it. The presence of not joined epiphysis clearly indicated that the bones belonged to young animals. US612 rested on a flat mud surface. The pottery comprised few fragments of fine fabric cups and spouted jars (pl. 64/9-11). north of wall m 168

This area was lying in a very bad state of preservation, due to the intense washout that affected the whole southeastern part of Area 3. The surviving deposit, US476, was distinguished along a narrow band – about 1 meter long – north of wall M168. It was identified as a foundation deposit on the basis of its position – originally below a retaining mudbrick wall – and composition. No potsherd was collected. between m 168 and m 167

This compartment was set up in phase with the construction of the parallel, retaining mud-brick walls M168 and M167, running along a northwest-southeast direction, that also were delimiting the compartment itself. This was covered by the massive mud and stone fills laid down between the two walls. Its east side was eroded by washed materials. The preserved perimeter shows a rectangular plan of almost 1x0.80 m. The uppermost layer, US465, was a 0.40 m thick deposit of medium compact, grey sandy loam and mud, with scattered small size charcoals. The stratum composition – with an evident laminated layout – is very similar to US72, a substantial waterlaid layer covering a ritual deposit related to the Husn Salut II phase (see below, earlier room 5b). US465 covered US464, a thin layer of soft, dark brown loam mixed with Iron Age sherds, bronze slag and several disarticulated animal bones, including those of small rodents. At the bottom stood a surface of hard lumps of mud (fig. 78). Pottery from US464 included a fine fabric storage jar and scanty sherds related to craters (pl. 65/1-3). Furthermore, one half of a dagger or tool handle made from animal bone was also discovered (pl. 65/4).19 south of wall m 170

The foundation deposit was originally contained by the retaining walls M170 and M169. Only a small strip – 1.0x0.30 m – was preserved along the southeast face of M170 and the northeast face of M169, the remaining part being completely washed away. On the top stood a compact layer of mud and loam, also badly preserved and washed. This covered the reddish brown layer US471, composed of friable loam mixed with mud lumps, ashes, wood frag-

19 Semicircular in section, the external surface is burnished. Two couples of holes were probably related to the insertion of rivets. (Length: 9.2; width: 2.16; thickness: 0.76 cm).

214

chiara condoluci

ments, disarticulated animal bones, Iron Age potsherds and a little bronze slag. A copper/bronze spearhead was also found that had been intentionally bent (pl. 65/16). Considering this bending alongside other features – i.e., the rounded point and the thin blade – it can be suggested that this represented a non-utilitarian object. Pottery comprised a range of shapes including carinated cups and bowls, storage jars and large open vessels like craters, in fabrics that consistently vary from fine to coarse (pl. 65/5-15). east of entrance

To begin with, a different building technique employed for the inner curtain on this part of the site needs to be pointed out if one is to compare it to the one exposed south of the entrance and previously described. In fact, the use of the compartment technique appears to be restricted to the central part of Area 4 (see Chapter 4), while immediately east of the upper entrance and along the adjacent stretch of the outer wall no compartment was revealed, with the notable exception of the one discussed here, which is connected to another foundation F igure 78 – Area 3, compartment between radial deposit. walls M168 and M167 (view from south; see fig. This difference in building techniques appears to be con69): detail of the laminated layer US465, visible in sistent with a slightly later phase of construction, as the rasection, and of the hard packed fill at the bottom. diocarbon analysis of a sample collected from a deposit excavated in Area 8 (US454, mentioned below) seems to suggest. Calibrated dates range in fact 1367 BC – 916 BC at 95.4% of probability, (although 95.1% is given by the range 1314 – 916) and 1252 BC – 1008 BC at 68.2% of probability (fig. 79). This small temporal shift between the construction of the two halves of the early site reasonably finds its raison d’être in the logistic constraints that are intrinsic to such a demanding enterprise as the construction of Husn Salut was. In fact, moving from the southern to the northern part of Area 4, there is a considerable increase in the gap that needs to be overcome between the natural slope of the hill and the final, projected elevation of the different platforms of the site. This required an enormous amount of building and filling materials, hence of time as well: it seems thus likely that the builders opted for the completion of the uppermost part of the site first, instead of carrying out the whole construction at the same time. Therefore, this shift does not mean that the eastern part of the outer wall has to be attributed to a different building phase; rather, this shows how the initial, unitary project, was realized into at least two main steps for logistic needs, with the westernmost part of the site, west of wall M26 and including the entrance, completed first. In Area 8, the outer curtain wall still shows an outer face made of irregular rows of roughly hewn blocks. Inward, this exterior revetment is backed by a first containment feature that consists in a fill of large and medium size stones bound with simple loam (US552), more than 1 m wide. The series of massive and compact deposits that constituted the filling of the wall structure were composed either of loam mixed with a large amount of small blocks, mud and rare sandstone F igure 79 – Probability distribution for US454. chips (such as US548=US538), or by simple mud (US541).

5 . the iron age architecture of husn salut

215

The foundation deposit discussed here, like the others located west of the entrance, was part of the original construction of the massive outer wall substructure. Despite an Islamic cut that impacted its eastern portion, most of it was still lying undisturbed. The rectangular compartment, 2.8x1.8 m wide, was realized against the retaining mud-brick wall M183 and was delimited by the other mud-brick walls M194, M189, M188. A compact fill comprising irregular rows of scattered stone slabs and blocks bound by thick layers of mud, US531, also sealed its northeast side. This compartment contained a single, homogeneous deposit sloping eastward, US454.20 This medium compact fill was composed of light brown loam mixed with mud lumps and small sized stone chips, scattered charcoal flecks, botanic remains (concentrated on the upper part), bronze slags, disarticulated animal bones, and a huge amount of Iron Age sherds. The quantity of potsherds and animal bones greatly exceeds similar evidence from the other foundation deposits. At the bottom of the compartment, US454 covered two identical mud surfaces, US535 and US536, which were separated by mud-brick wall US534, only exposed in its upper crest, which clearly represents an inner feature of the outer wall substructure (fig. 80). However, it must be underlined that US454 exceeded the compartment limits and, together with the layers covering it, must be interpreted as the outcome of some sort of larger scale foundation ritual. Some of these layers (US492, US497, US502), in fact, while certainly being part of the massive fill of the outer wall, show features that are more similar to a foundation deposit than a simple fill: traces of burning, scattered stones, the presence of disarticulated animal bones and Iron Age sherds. On the top they were partly sealed by a sort of poorly made wall, US498, made up of seven rows of stones bound by mud, parallel to M175. To their east stood a friable stone chip fill US493, dark brown in colour due to the traces of burnt materials, mixed with mud-brick fragments, Iron Age sherds and a large amount of small size stones.

F igure 80 – Area 8, the compartment that partly contained US454 after its removal, with the underlying compact fillings exposed (view from top of wall M175).



Its thickness decreases from 0.75 to 0.40 m.

20

216

chiara condoluci

- Pottery from US454 Despite being only preliminary, the following general description21 is deemed to be necessary until further detailed analyses and comparative study is completed, given the exceptional nature of the assemblage. This in fact includes a wide selection of shapes, made in fine to medium and coarse fabrics, several bearing painted decorations.22 Some categories, such as carinated cups, spouted jars, large basins and long handled bowls, are well represented in the early layers of the site. However, what is striking is the appearance in US454 of specific shapes some of which find really little if no comparisons in South East Arabian Iron Age. Among them the high percentage and variety of spouted vessels used for pouring liquids has to be emphasised, such as hemispherical bridge-spouted bowls, spouted large bowls/basins, spouted hole-mouth globular jars, spouted jars with strainers. Another significant feature of US454, in consideration of the rites that had likely been accomplished and lead to its deposition, is the presence of several complete or almost complete vessels, found smashed and mixed with other sherds inside the fill. This is not the case for the other described foundation deposits, where only small portions of the different vessels were thrown. Cups and carinated cups from US454 (fig. 81; pl. 66/1-5) are all in a fine fabric – ranging from pale to dark orange – with medium or small size grits, both painted (in red or brownish colours) and unpainted. Rosettes, stars, suns, crosses, hanging rectangles represent the assorted decoration motifs generally found on the interior surface, while oblique cross-hatching, ‘running dog’ and wavy horizontal or vertical lines are generally on the exterior.23 Cups at Husn Salut are a small group, represented both in late Basement contexts (such as US473, US7, US9, US12) and in early and late Burnt Building contexts (US381, US385, US390, US407); some of the shapes seem indeed more a reminiscence of the Late Bronze Age assemblage,24 being not so popular on Iron Age II sites, which would be consistent with the proposed early foundation of the site (see Chapter 10). Except for the painted motifs – less frequent among the Burnt Building pottery and almost absent in the Basement, this high frequency being characteristic of US454 assemblage as a whole (fig. 82) – the carinated cups don’t

F igure 81 – Spouted bowls and carinated cups from US454.

F igure 82 – Painted sherds from US454.

21 Individual descriptions are not given in the plate captions for contexts that have not been fully studied. 22 A large number of large, coarse storage jar sherds will not be included in the discussed sample since they still await proper graphic documentation. 23 For similar decoration motifs see, among others, examples from nearby Izki in Schreiber 2007: pls 2-3. 24 See Velde 2003: 106, fig. 4/4-5.

5 . the iron age architecture of husn salut

217

have other distinguishing features compared to the early assemblage of the site. However, some observation can be made: the manifest asymmetry of the sherds (definitely handmade), and an accentuation of the carination – from which resulted slightly larger and lower shapes and more convex bases. There is also a larger use of string-cutting on the lower part of the walls. Curiously, these features find some echoes in carinated cups from late layers in the Basement’s sequence, such as those from US35. Bowls are mainly carinated with simple rounded or flat rim, some with brown painted decoration comprising stars, hanging rectangles and ‘running dog’ motifs. Bowls with undulating walls above the shoulder (pl. 66/16-17), absent from Basement contexts, can be compared with examples from Hili25 and Rumeila.26 Bowls with flat, slightly rounded or bevelled rim (pl. 66/19-22) find comparisons both in US75 and US322, two layers related to foundation deposit contexts. Although these bowls are all handmade, it is interesting to note a certain degree of uniformity in their production; the majority has a diameter of 20 cm and a depth of 8 cm. Deeper bowls are also present (pl. 67/3-10). Among the large bowls/basins noteworthy is a spouted one with grooved rim (pl. 68/2), made in an orange medium-fine fabric with grey core, some vegetal temper and white grits, red slipped on outside and brown on the inside. This shape finds no precise published parallel among South East Arabian Iron Age sites, although some examples were recently discovered at Mudhmar near Adam and in a long underground grave at Daba (Oman).27 The hemispherical bridge-spouted bowls (pl. 69/1-3) were made in fine pale orange to dark orange fabric, with scattered presence of white grits, and dark brown to red slip. Only one example is painted with a black decoration comprising a ‘running dog’ motifs and hanging triangles (pl. 69/4). Indeed, the latter is more a transitional shape between the mentioned hemispherical bowls and a squatted spouted jar, as shown by the short neck and the flat, out-turned rim. General comparisons can be found at Maisar, Rumeila, Shimal28 but also from nearby Wadi Bahla.29 Bridge-spouted jars (pl. 69/6-7) are largely attested in early levels of the Basement (US16, US35), although not painted. Handled jars (pl. 70/1-4), although known in one example from the Burnt Building (US366, see pl. 43/17), are far more representative within this context. It is possible that they were also equipped with a spout. Some sherds are painted in brown, both on the rim and on the walls, with ‘running dog’ and vertical line motifs. One possible painted comparison is known from Nizwa,30 few other unpainted ones from Izki and Rumeila.31 Jars with short neck and rounded rim (pl. 70/6) are rare at Husn Salut32 and neighbouring Iron Age contexts as well.33 Without comparisons is a jar with flat out-turned rim and wavy applied ridge on its shoulder (pl. 70/5), further distinguished by its painted decoration. The fabric is fine with scattered white big inclusions and a pale yellow slip on the exterior. A brown net-pattern band decorates the neck, while the upper part of what can be recognized as a frieze of hanging triangles – filled with a pattern of wavy vertical lines – is below the ridge. Globular bridge-spouted jars (pl. 71), are also unusual at Husn Salut and marginally testified on other sites. The illustrated sherds from US454 were made in an orange fine fabric with scattered white grits and rare vegetal temper. Red slip is on interior and exterior surfaces. The dark red painted decoration – mainly occupying the half upper part of the vessels, including the spout and the upper part of the rim – comprises patterns of assorted motifs such as ‘running dog’, dot in circles, hanging rectangles (in-filled with vertical or horizontal wavy lines), cross-hatchings, hanging circles with vertical wavy lines, and simple vertical, transversal or wavy lines that were used to fill the narrow spaces of the vessels.



al-Tikriti 1998: pls 7-8; Benoist, Magee and al-Tikriti 1998: fig. 5/A-B. Boucharlat and Lombard 1985: pl. 45/5, 7. 27 Jean, Pellegrino and Gernez in press; M. P. Pellegrinno, pers. comm. 28 Yule 1998: fig. 1/10; Boucharlat and Lombard 1985: pl. 50/1-3; De Cardi 1998: fig. 2/37. 29 Condoluci and Degli Esposti 2015: pl. 4/39. 30 Schreiber 2007: pl. 66/1-3. 31 Schreiber 2007: pl. 31/1; Benoist 1998a: fig. 5/10. 32 One example comes from US377 in the Burnt Building. 33 See resembling examples from Izki (Schreiber 2007: pl. 26/4) and Lizq (Kroll 1998: fig. 7/54-55). 25 26

218

chiara condoluci

Their globular, in some cases slightly squat shape finds comparison with un-spouted jars from Lizq34 and from Bithna35 although with less elaborated decorations.36 Probably a more fitting example is one of the spouted jars found in the pillared building of Muweila,37 bearing a geometric painted decoration. In a similarly early context at Husn Salut (US16) a globular but less squat jar with tubular spout on the shoulder was also found (see pl. 23/6). Definitely the most intriguing shape within US454 assemblage is a tubular-spouted globular jar (pl. 70/11). The spout is inserted almost at right angles with the wall, on the shoulder, and is also equipped with a strainer near its mouth. The vessel’s rim also shows a broken inner ledge that indicates the original presence of a strainer. Two or more probably four horizontal lugs were on the shoulder. This globular jar was probably not unique inside the stratum. Other sherds (pl. 70/8-10) show a similar inner ledge below the rim, with evidence of the original strainer, and could belong to the same kind of jar, despite the absence of spout traces. Another similarly shaped jar shows the presence of a handle (pl. 70/12). Jars with more or less out-turned rim and inner ledge are attested on Iron Age sites, although of a type other than those found in US454.38 Besides, in those cases, the inner ledge is interpreted as serving to support the lid, and the presence of a strainer is never attested. Several sherds belong to hole-mouth globular jars. Their rims are usually in-turned, simple or nail headed, with the presence of small triangular shaped lugs on the rim. The fabric is generally medium/fine, orange in colour, with scattered vegetal temper and red grits; pale-red slip is largely used. Some vessels show a blackened exterior. Another characteristic of this form at Husn Salut is the possible presence of spouts. Some are tubular-shaped and almost vertical, inserted on the shoulder (pl. 72/1-4); others are U-shaped, open and horizontally inserted on the rim (pl. 72/5). Interestingly, both the shapes with tubular spout and those with triangular lugs on the rim find good comparisons at Husn Salut from another context likely connected with ceremonial activities (US298, see Chapter 4). At other Iron Age sites hole-mouth globular jars are widely attested, but they lack the spout. General comparison comes from Bithna,39 Hili 17,40 Nizwa and Izki.41 More specific examples of hole-mouth globular jars with lugs along the rim and with in-turned rim are attested at Izki42 and in the Salut area.43 Deep and large basins with almost straight walls and lugs on the rim (pl. 73/1-5) are generally made in medium/coarse dark orange fabric with a black slip on both surfaces, sometimes also blackened. One comparison comes from Izki.44 Pans with short, slightly incurved walls (pl. 73/9) are well attested in US400 and also in US473 of the Basement. Fabric is coarse, generally orange in colour and blackened on the outside, with big and medium red, grey and white inclusions. Similar examples are testified at Shimal, Rumeila Periods I and II contexts, and Nizwa.45 Fragments of jar lids with incised herringbone decoration or simple, short transversal lines along the edges are also present (pl. 73/6-8). Long handle bowls are widely attested in the stratum, both fragmentary and complete (fig. 83; pl. 74/4-8). The reservoir can be considerably flared, up to 20 cm in diameter. Inside, two shallow digital impressions deriving from the fixing of a handle are usually present. The handle is constantly decorated with a herringbone pattern, occasionally comprising a central line. Decoration is sometimes simplified with short transverse lines along the edges. The front

34 See Kroll 2013: fig. 21/17, 19-20. 35 Benoist 2013a: fig. 39/1. 36 Extremely fragmentary sherds possibly belonging to these kinds of jars and also with painted decorations of hanging triangles are testified at Rumeila and Husn Madhab. See Benoist 1998a: fig. 7/9; Benoist 1998b: fig. 6/13. 37 Magee 2007a: 51, M1046. 38 The best comparisons come from Shimal (Velde 1998: fig. 3/1-2;). Other kinds of jars with inner ledge are known for example from Lizq (Kroll 1998: fig. 5/44-46;), Tell Abraq (Potts 1991: fig. 120/15-16), Bithna (Benoist 2013a: fig. 106/1) and Husn Madhab (Benoist 1998b: fig. 4/2, 4, 7). 39 Benoist and Corboud 1998: fig. 3/2; Benoist 2013a: fig. 124/6. 40 Benoist, Magee and al-Tikriti 1998: fig. 11. 41 Schreiber 2007: pls 8/1-2; 22/4; 53/9; 59/7-8. 42 Schreiber 2007: pls 37/7; 58/7; 59/9. 43 Condoluci, Degli Esposti and Phillips 2014: fig.7/82; Phillips, Condoluci & Degli Esposti 2012: pl. 5/52. However, recent excavation at Mudhmar, near Adam, has finally revealed similar shapes, also comprising tube-spouted examples (Jean, Pellegrino and Gernez in press). 44 Schreiber 2007: pl. 20/6. 45 Velde 1998: fig. 4; Benoist 1998a: fig. 16/7-8; Schreiber 2007: pl. 56/7-8.

5 . the iron age architecture of husn salut

F igure 83 – Long handled bowls from US454.

10 cm

F igure 84 – Incised (graffito) snake silhouette on a storage jar rim from US454.

46

E.g. Benoist 2013a: figs 112-114; Benoist et al. 2012a: fig. 14; Taha 1982-83: fig. 15/A-B. Schreiber 2007: pls 67; 79/9-10.

47

219

section of the handles has a deep groove, a common feature used on long handle bowls from the site (see Chapter 2). The fabric is generally medium/coarse with sometimes grey core. Slip ranges from pale red and pale orange to brown. Scanty traces of burnt matter are often found inside the reservoir and several examples were blackened on their interior. Apart from the well know examples from Bithna, Masafi and al-Qusais,46 comparisons come also from Nizwa and the Jabal alAkhdar survey.47 The snake motif is also attested within US454 assemblage, although only present on two coarse ware fragments. The complete silhouette of a crawling snake with triangular shaped head is incised on the rim of a storage jar (fig. 84). Another triangular-headed snake is applied in relief on a wall sherd, where only part of the body with impressed dot decoration is preserved (fig. 85).

5 cm

F igure 85 – Applied snake decoration on a storage jar wall from US454.

220

chiara condoluci

10 cm

Among the characteristics of US454’s assemblage are a few sherds of grey ware (fig. 86; pl. 74/1-3) and other fragments bearing a bizarre red painted decoration. Two, not joining grey ware sherds belong to a small oval-shaped jar with slightly out-turned rim, incised vertical line decoration on the walls and a ring base (pl. 74/1-2). The fragment of a jar neck with incised wavy line decorations is also present (pl. 74/3). The fabric is fine, with rare white grits and rare vegetal temper. Grey wares are reported from Wadi al-Qawr and other sites in the region, although there’s no comparison with the shapes from US454.48 The mentioned sherds of red painted pottery are not attributable to specific shapes (fig. 87). Their buff-orange fine fabric contains several small, medium and large white and red grits, and vegetal temper. The decoration is rather unusual and actually without comparisons. The most evident pattern is apparently a series of red ‘dot-in-circle’ motifs, but a more accurate observation allows tracing another, possibly brownish motif, now almost disappeared. The resulting decoration comprises a pattern of adjoining ‘running dog’, probably delimited by monochrome bands and wavy lines.

F igure 86 – Grey ware sherds from US454.

10 cm

F igure 87 – Red-painted sherds from US454 (left), and a drawing reproducing the original bichrome decoration (right).



Phillips 1987: figs 19-20; see also Barker 2002: 51. See also Chapter 2.

48

5 . the iron age architecture of husn salut

221

- US454 small finds A few fragmented objects were also discovered within this deposit. One copper-alloy fragment, cast, possibly reproduces the wavy body of a snake with scattered dots (see Chapter 7, no. 10). Despite several copper/bronze snakes being known form Husn Salut (see Chapter 7), the only one of such tiny dimensions comes from a Burnt Building fire-related context (Chapter 7, no. 9). More numerous are instead similarly sized examples from Masafi 3 in the Emirate of Fujairah.49 A reshaped arrowhead was found nearby (Chapter 7, no. 69). The tang is bent, and part of the blade folded, possibly as a result of intentional spoiling and also possibly reflecting the intention of reproducing a sketchy snake body. An incomplete animal figurine, possibly the representation of a camel, was also present – Cl163 (pl. 74/9).50 The figurine is headless, while the back half of the animal’s torso with the insertion of its hind legs is preserved. An irregularly squared bulge on the upper part of the torso presumably corresponded to a saddle rather than to a hump. Animal figurines are not rare at Husn Salut,51 and camel figurines are widely present in the South East Arabian Iron Age, both in dwellings and in collective building contexts.52 The fragmentary piece Cl163 can be better compared with examples from Muweila.53 Rather bizarre is a fragmentary ceramic item, Cl217: 22 cm long, it has a curved silhouette and a flat, thicker base.54 The object could be part of a clay terracotta pipe for water drainage from a roof, like those testified at the Iron Age site of Hili 2, Hili 17 and Muweila,55 which are, however, complete objects and usually present a ‘U’ shaped section with a flat base. - Archaeobotanical analysis The analyses of botanical samples from US454 gave a meaningful contribution to the understanding of the ritual practices on the site. The presence of a thin layer of unburned plant remains mixed with charcoals at the top of the stratum, has already been highlighted in a specific archaeobotanical study concerning Salut.56 This layer is almost totally composed of leaves, twigs, branches, fruits and seeds of Nerium oleander. Few parts of the date palm fibrous leaf sheath and a fragment of monocotyledon leaf were also present, while the charcoal fragments were attributed to Juniperus sp. A high percentage of Cerealia pollen grains, Sesamum and Ocimum grains have also been revealed by palynological analysis. Nerium oleander L. grows in periodically inundated wadis and thus it should be readily available nearby Salut. However, the high toxicity of the entire plant and the feeble consistence of its wood, make it unsuitable both for fuel and for architectural use. Although considered an ornamental plant, pollen analysis excluded that flowers were placed in the deposit. The presence of oleander is rather singular and difficult to explain, considering also the lack of comparisons with records from other Iron Age sites in the Gulf. On the other hand, it is important to underline the scarcity of archaeobotanical studies in Iron Age Oman, specifically for likely ritual contexts, and the absence of contemporary written sources. Hence, the possibility that oleander was offered in ceremonial practices must not be ruled out. Interestingly, there are many mentions of its medical use in Roman57 and Arab58 written sources as an effective snakebite cure. This



Benoist et al. 2012a: fig. 12. M/F orange fabric. Medium grey grits, and vegetal temper. Traces of red slips and black painted decoration. 51 Cf. Condoluci 2015a. See also Chapter 2. 52 See Boucharlat and Lombard 1985: pl. 65/4, 6, 10; Benoist et al. 2012a: fig. 18. 53 Magee 2007a: figs 25, 30. 54 M/C pink fabric with black core. Red and black grits. Thickness 2.3/1.2 cm. 55 See Lombard 1985: 154; al-Tikriti and Haddou 2001. See also Chapter 2. 56 Bellini et al. 2011. 57 Pliny, Naturalis Historia 16.33, 24.53. 58 See Levey 1966. 49 50

222

chiara condoluci

aspect cannot be without significance in a foundation deposit of one of the most important Iron Age sites related to ‘snake worship’. Juniperus (juniper) currently grows on the summit of the al-Hajjar Mountains (c. 1500-3000 m a.s.l.), and does not belong to the same vegetation belt of the other plants identified at the site. Juniper charcoals have also been found in US75, another foundation deposit (see below). Therefore, its presence in the US454 seems to testify the deliberate choice of that specific wood for a precise purpose. Concerning date palm remains, they may indeed represent an important component of the ritual acts throughout the site, considering that the archaeobotanical results coming from US75 strongly suggest the offering of palm dates and palm flowers (see below). The comparison of the pollen sample from US454 with those from other areas of Husn Salut has shown that, whereas the pollen records of cultivated plants – i.e. Sesamum, Ocimum, Cerealia, and also Phoenix – generally recalls farming activities in the surroundings of the site, and specifically date-palm oasis agriculture, their pollen percentage in some cases is so high that they more likely represent anthropic accumulation from activities within the site rather than the pollen rain from nearby oasis farming. In this respect it is worth mentioning that Ocimum sanctum L. (tulsi or tulasi) is currently cultivated in India for medicinal and religious purposes, being considered one of the most sacred plants for offerings in Hindu temples. between m 143 and m 201 ( platform sf 49 )

The northernmost of three compartments that form the original structure of this platform was excavated down to the bedrock, revealing two adjacent sequences of intentional fills, separated for half their depth by the reinforcement wall M181 (fig. 88). While the central one hosted on its top an hollow tank/niche, the southernmost was almost completely dismantled by later works (pls 60-61). Northeast of M181, the compact mud-base fills US510, US509 (well-laid mud-bricks), and US508 were laid down, reaching the elevation of M181’s crest. To the opposite side of this same wall, the fills’ sequence comprises (bottom to top) US580, US581, US579, US468 and US349, the latter covering M181’s crest and US508 as well. Above this level, US337 represents the essence of the ritual offerings, and comprises the remains of organic materials burnt elsewhere and intentionally laid down here: charcoals, date stones, animal bones, bronze slag and potsherds are abundant, together with peculiar bronze objects. A radiocarbon date was also obtained for US337, which ranges between 1497-1030 at 95.4% of confidence and between 1414-1132 at 68.2% of confidence (see fig. 40/e). Further above, the thick, compact mud-bricks and mud lump fill US751 was laid down to fill more than half of the remaining space; against M117 the sequence instead comprises US336, US335, US332 and US327, the latter comprising a small pile of stones that sealed the compartment (fig. 89). Although protracted over a long time, and with changing technique and materials between one step and the other (discussed in Chapter 3), the construction and in-filling of this compartment can be interpreted as a single action. - Pottery The pottery assemblage from US510 is for the most part represented by large basins with flat rim, with the exception of a single sherd corresponding to a small hole-mouth globular jar. The fabric is generally medium, brown or red in colour, with large use of vegetal temper and grits. Brown and pale red slip is commonly used. Few coarse and one fine, non-diagnostic sherds were also present. Black painted horizontal and vertical bands or wavy line decoration is usually attested on the upper flat part of the rim. One globular jar (pl. 75/7) has a black painted decoration of hanging concentric semicircles surmounted by a horizontal wavy line, an almost identical sherd having been found at Husn Salut in late reshuffled deposits. Interestingly, this assemblage seems to bear some archaic (i.e., pre-Iron Age) features, which are however not inconsistent with its stratigraphic location in earliest layers of the site. Large basins, on the one hand, are widely present on Iron Age settlements. However, flat thick rims (pl. 75/2-6) are closer to examples from the

5 . the iron age architecture of husn salut

F igure 88 – Burnt Building, room 5b, excavation of the space delimited by walls M143 and M201.

F igure 89 – Burnt Building, room 5b, a stone heap (US327) on top of the compartment southeast of M181.

223

224

chiara condoluci

late second millennium (i.e., Late Bronze Age) contexts of Tell Abraq59 and Shimal60 than to those from Iron Age sites, such as Lizq.61 The globular jar is also singular both for shape and decoration. An identical example comes from Husn Salut, unfortunately from a mixed context (US40) located on Terrace II. Concerning the shape, which is not so common, the best comparison comes from Iron Age contexts at Tell Abraq.62 On the other hand, the hanging concentric semicircle motif is quite rare in the Iron Age pottery assemblage,63 while it is more used during the Wadi Suq or Late Bronze Age period.64 From US509 and US349 only a few non-diagnostic sherds were collected, all in medium to coarse fabrics, with abundant grits, some vegetal temper and often a grey core. Pottery from US337 comprises few diagnostic fragments of medium and fine fabric,65 mainly hard fired, with vegetal temper and occasional grits. Slip ranges from dark grey to brownish-red. This small number of sherds included a selection of shapes: one carinated bowl with painted wavy lines motif, one bowl or jar flat base and one possible jar with rounded and beaked rim (pl. 75/8-10). This last jar is quite singular. The shape of the rim and the flaring wall are more remindful of a type of second millennium jars, such as those from Tell Abraq,66 than of typical Iron Age II jars. The same features would also be rather unusual if compared to typical Iron Age bowls, and only a vaguely similar example comes from an early foundation fill of the Burnt Building (US427). Given the context of discovery, it would be tempting to recognize this sherd as the fragment of the upper part of a cylindrical foot brazier like those attested at Bithna,67 but the rim differs. Potsherds from the uppermost levels US336 and US332 comprise a storage jar with incised criss-cross decoration on the cordon, a spouted jar spout, a carinated bowl with bevelled rim and a jar with rounded everted rim (pl. 75/1314, 16-17), as well as few non-diagnostic fragments in coarse fabric. Diagnostic sherds are mainly in a medium/fine fabric with a wide use of vegetal temper and grits, with slip ranging from brown to pale red and buff-yellowish. Storage jars and spouted jars of this type are largely attested on Iron Age II sites. The carinated bowl finds few comparisons with bowls and large basins from the early layers of the Basement (cf. US16) and one example from Nasla in the Emirates.68 However, examples of carinated bowls with bevelled rim are more common on Iron Age III contexts.69 The jar from US332 (pl. 75/16) is also slightly unusual. Similar but not equal examples can be found at the Iron Age site of Shimal70 or in Iron Age contexts of Tell Abraq71. At Husn Salut a comparable jar was found in Burnt Building contexts of the HSIII phase (US99). - Stone finds Three stone objects were discovered from this stratigraphic sequence. Two are uncompleted soft stone lids, one circular and the other rectangular, meant for a typical compartment box, coming respectively from US327 and US335 (see Chapter 6, nos 40, 43). The third is a stone tool, a fragmentary anvil stone from US336, with a cavity on the upper side.72

59 Potts 1990b: fig. 91/8,10. 60 See for example Velde 1998: figs 5/Eso 02.1-2; 8/Ess 02.7,9. 61 Painted decoration on rim is witnessed on large basins from Lizq, although the rim shape is quite different from those found in US510, being in fact flat triangular (Kroll 1998: fig. 2/24-25). 62 Potts 1990b: fig. 136/10. 63 One unusual example is shown on a bowl from Maisar (Yule 1998: fig. 1/3). 64 See for example Potts 1991: fig. 43/9. 65 Scanty, non-diagnostic sherds in coarse fabric were also found. 66 Potts 1990b: fig. 97/8,10. 67 Benoist 2013a: fig. 110/1. 68 Phillips 1998a: fig. 20/4. 69 Benoist and Del Cerro Linares 1998: figs 4/17; 6/16; see also Phillips 1998b: fig. 1/17. 70 Velde 1998: fig. 3/Efl 01.2-3. 71 Potts 1991: fig. 112/8. 72 Not illustrated. Dimensions in cm: l. 8.8; w. 9.3; th. 3.6; d. of the cavity 2.0.

5 . the iron age architecture of husn salut

225

- Metal finds Definitely, among the bronze objects from ceremonial contexts at Husn Salut, one of the best preserved and also one of the most meaningful is the shaft-hole axe from US337 (MB196, see Chapter 7, no. 83). A vertical rib is shown on the back side of the tubular shaft while the scanty remains of the wooden handle were still present inside the shaft. The chronological implication of such a kind of artefact in relation to early Iron Age levels from other settlement in the Oman Peninsula have been discussed elsewhere,73 while its possible interpretation is discussed in Chapter 5. A small copper-base arrowhead comes from US337 (see Chapter 7, no. 67). The shape of the object, with rounded edges and rounded point is not functional, suggesting a re-working or an original useless shape related to its destination (ritual deposition). From the same context come also few bits of scattered bronze slag.

C ompartment

substructures

( and

superstructures ) in

H usn S alut II

phase

At some point during its Iron Age history the site of Husn Salut – or at least part of it – underwent substantial architectural re-arrangement, probably as a consequence of some dramatic event. This latter circumstance is evidenced by a line of collapsed mud-bricks mixed with large sized stones in the central corridor of the Basement, a layer that is covered by new floor surfaces (see fig. 46). Even more significant are the thick layers rich in burnt matter that are found in the Burnt Building and in several areas of the Basement, and that bear witness to a strong fire that hit this whole area of the site (e.g. US318, US367, US394). Doubtless, these events mark the transition to a new and significant architectural phase (HSII), that is also characterized by the use of the compartment technique described above. This has been specifically highlighted in the eastern part of the Basement, and in the northern part of the Burnt Building. In the Basement, the mud-brick walls of its eastern extension, i.e. room 2a, were now turned into containment walls delimiting new compartments, with the addition of a newly built, smaller cross wall such as M6a/M6 and M11. These compartments were then intentionally filled with thick, compact muddy material, laid down directly above the burnt layers that constitute the trace of the abovementioned large fire (US400/US408). These fills were then capped with one or two mud-brick rows that levelled the compartments with the surviving crest of the adjacent, early phase walls (M149/M161); this operation thus provided an even surface to serve as the foundation ground for the erection of the so-called Building 1, whose plan has been detailed in Chapter 2. With the exception of walls M173 and M174, built near the northern corner of the building, where massive deposits were being laid down to backfill room 5a and the space between M187 and M8, the remaining part of the Basement did not require the construction of additional walls, likely because the spaces between the existing ones were already reduced enough to provide sufficient solidity while the massive deposits finally obliterating the rooms were being laid down (US12, US13). In the northern portion of the Burnt Building significant structural changes concerned room 5b. Another smaller mud-brick compartment was in fact built, or better completed, northwest of wall M143, laying down a series of deposits that sealed the earlier room 5b.74 The final result was the enlargement of the raised platform SF49, now L shaped and reaching a northwest-southeast extension of 11.8 m, with a maximum perpendicular width of 5.7 m at the northwestern end. All in all, however, these changes can be read as the essential part of some ceremony75 of which what can be seen is only the final outcome. Indeed, the whole of the room must be considered as a ‘secondary’ compartment, in the sense that walls not originally meant to delimit such a feature were later used in this way, just like the originary walls

73 Phillips 2010: 73. 74 See also Chapter 2 and detailed stratigraphy in Chapter 3. 75 The possible nature of these ceremonies is discussed in Chapter 9, as they are necessarily connected to the broader issue of the site’s function.

226

chiara condoluci

of the Basement mentioned above. This compartment was then intentionally in-filled within the context of a specific sort of ritual act that also included the deposition of peculiar materials in the lower layers, which are in all instances comparable with the foundation deposits previously discussed. As the rite proceeded after the deposition of these specific materials, the compartment was initially left open in its eastern part, where water-laid, laminated deposits sedimented (US72). In a later moment, it was sealed with thick mud fills and then capped with mud-brick rows to reach the level of the surrounding walls’ crest. One of the main concerns related to this late phase in the use of the compartments technique was clearly that of structural solidity. With this goal in mind, the fills inside the new compartments never exceeded the retaining walls’ height. Probably for the same reason several walls were reinforced by flanking them with new ones, finally increasing their thickness (M30-M41, M42/M28, M78/M201). Among the consequences, it has already been underlined that the L-shaped space between the Burnt Building and Building 1 became completely occupied by these reinforcement walls.76 Overall, what has to be underlined is the long continuity in the use of the compartment technique, with the only difference that during the later phases these largely re-used pre-existing walls; moreover, this building technique continues to be inclusive of the presence of characteristic foundation deposits, always encased inside some selected compartments.

Late foundation deposits and associated materials (Husn Salut II) southeast of wall m 78 ( platform sf 49 )

Close to room 5b, platform SF49 witnessed the deposition of ceremonial/foundation deposits that were meant to obliterate and level the earlier central niche. Here, three deposits were distinguished – US569, US568 and US567, bottom to top (fig. 90). Although they contained only a few potsherds, not illustrated here,77 the discovery of a complete copper-base spouted cup in US568 (fig. 91; see Chapter 7, no. 22) was exceptional.

a

b

F igure 90 – Husn Salut II, Area 1, compartment east of wall M78: a) during excavation (mud-brick fill US567). In the foreground oven SF21 and a recess for the bronze cauldron delimited by stone slabs; b) at the end of the excavation.



To the extent that an even larger plan may be envisaged for SF49. Substantial erosion however precludes certainty on this issue. Fabric is mainly fine, with few examples of it being medium.

76 77

5 . the iron age architecture of husn salut

227

between walls m 41 b and m 143 ( earlier room 5 b )

The most significant foundation deposits of this phase are those discovered inside room 5b of the Burnt Building, now turned into a ‘secondary’ compartment. There, deposit US75 and US74, laid down above the ancient surface US100,78 represents the actual core of the ceremonial offerings (pls 60-61). The origin of these two layers is similar to the other foundation deposits described above, in the sense that they likely represent the remains of organic materials burnt elsewhere and intentionally laid down here: charcoal, date stones, animal bones, together with bronze slag and potsherds are abundant. However, US75 is undoubtedly even more representative than most of the other deposits (except US454), F igure 91 – Bronze spouted cup (MB 214) from the as it is a remarkably substantial layer mainly consisting compartment east of wall M78, before conservation. of combusted vegetal materials. The distinct organic material – charcoal, wood fragments, burnt date stones, fragmentary palm-leaf baskets and disarticulated animal bones (including those of small rodents) – is supplemented by the presence of objects the majority of which seems to be endowed with a ritual nature, such as two copper-base ladles, a miniature copper-base axe, some iron fragments and a few potsherds, among which is a long handle bowl handle with applied snake decoration. A more unusual find, actually a unicum so far at Husn Salut, is a fragmentary, red paint-coated brick (see general remarks below). Part of the ritual apparently was the fact of leaving an open space directly above these two layers, where water-laid deposits could sediment over an un-specifiable time span. In fact, US72 was distinguished here, consisting of a generally compact, grey-brownish sandy layer that comprises a substantial series of thin laminations (total thickness around 0.5 m). Given the radiocarbon date available for US75,79 calibrated at 784-542 BC (95.4% of confidence, see fig. 27), this addition to SF49 was realized some time after the beginning of the HSII phase. Indeed, this could also have happened during the HSIII phase, when it was shown that a large part of the Burnt Building was re-arranged. This uncertainty remains when the material from the connected deposit is examined. In fact, although for no sherd a date in the Iron Age III can be safely assumed, it could at the same time be suggested for a few ones which slightly differ from the overall assemblage (pl. 76/5-7). - Pottery from US74 and US75 A reduced number of shapes was witnessed in these two layers (pl. 76): necked jars (with out-turned rounded rim), bowls (with simple pointed or in-turned flat rim), carinated cups (one with painted decoration, the other with traces of a graffito design). The fragmentary handle of a long handle bowl with applied snake decoration was also found (pl. 76/9). All, except one complete bowl from US75, are fragmentary. The fabric is usually fine or medium with sparse grits and rare vegetal temper, although some coarse, non-diagnostic sherds were also present. Slipped sherds range from red-brownish to orange-brownish colour. Eight non-diagnostic Iron Age sherds in coarse fabric come from US72, while several non-diagnostic sherds, in either coarse or medium fabric, come from US75. Pottery assemblage from US74/75 is not so distinguished as to be attributable to a specific period within the Iron Age; conversely, some shapes are well represented both in Iron Age II and Iron Age III assemblages from South East Arabia settlement contexts. However, tentative parallels can be suggested. Bowl US75,1 for example (pl. 76/3) finds comparisons at Rumeila Period II80 (that is, Iron Age III period), as well as in Iron Age II contexts of Shimal



See footnote 10 in Chapter 3. 784-542 BC at 95.4% of confidence (see fig. 27). 80 Boucharlat and Lombard 1985: pl. 53/3; Benoist 1998a: fig. 10/23. 78 79

228

chiara condoluci

settlement SX.81 Bowl US75,6 (pl. 76/4) can be compared again with shapes from Rumeila Period II,82 but also from al-Madam at the end of the Iron Age II period83 and also from Lizq 1.84 At Husn Salut, bowls with in-turned slightly rounded rim are mostly represented in contexts of HSIII phase, although basins with flat in-turned rim already appear in levels of the earliest phase of the site. Regarding the jars, examples with out-turned and slightly triangular rim come from HSII levels (US23); however, one example has been found also in one of the later contexts of the Basement (US14), still within the HSI phase. Comparisons with other sites are not so strictly fitting, except some examples from the early phases at Lizq and Tell Abraq.85 However, some better correspondence for jar US75,2 (pl. 76/8) seems to be found in Iron Age III contexts, such as a spouted jar from Rumeila and a jar from al-Madam.86 This concise overview is thus in agreement with the chronological data provided by the radiocarbon determination reported above. - Metal finds Definitely the most specific objects that can be linked with ceremonies and related offers come from US74/75: two ladles, a miniature axe, a snake, some iron fragments. The ladles were found together (pl. 77/2-3; Chapter 7, nos 12, 13). They differ both in size and in manufacture. At the present state of research,87 the rarity of comparable finds in Iron Age South Eastern Arabia underlines the specificity of this type of object, which has to be beyond ordinary use. Chronologically corresponding comparisons come from the pillared Building II of Muweila,88 similarly interpreted in relation to elite activities. Centuries later, the presence of bronze ladles is attested in a 5th century BC tomb in Bahrain and in 1st century AD funerary contexts at ed-Dur.89 Here, where they were found as part of drinking sets inside the tombs, the ladles have been connected to banqueting and funerary feasting. The miniature axe found in US75 (pl. 77/4; Chapter 7, no. 17) conforms in size with another two coming from archaeological strata possibly related to or adjacent to ritual contexts (Chapter 7, nos 16, 18). These objects are clearly non-functional, but they retain the symbolic value of real axes, which they duplicate in smaller dimensions. They may have had an amuletic, apotropaic purpose, apart from a possible decorative one. The manufacture of axes during the Iron Age appears to be a local craft in the Oman Peninsula and in a few cases they are associated with ritual or funerary contexts, as shown by the axes from the funerary areas at Hili 8, al-Qusais, and Qarn Bint Saʿud90 or from the pillared buildings at Rumeila.91 This recurrent link is also supported by five nonutilitarian battle-axes found at the new discovered site of Mudhmar est.92 Moreover, at Husn Salut an axe is also found represented on the handle of a long handle bowl (see pl. 33/3). At Muweila, a socketed shaft-hole axe and another splaying axe-head were discovered, together with at least another fragmentary example.93 The two published axes, on the basis of their proportions, can be interpreted as non-

81 Velde 1998: fig. 6/Eso 05. 82 Benoist 1998a: fig. 11/6. 83 Benoist and Del Cerro Linares 1998: fig. 2/13, 18. 84 Weisgerber 1981: fig. 69/9. 85 Kroll 2013: figs 23/1, 11, 17; 24/17; Potts 1990b: fig. 132/4, 6, 8. 86 Boucharlat and Lombard 1985: pl. 53/1; Benoist and Del Cerro Linares 1998: fig. 3/14. 87 Apart from several other unpublished tombs on display in several museums of the region, a huge quantity of metal finds were recently discovered at the archaeological site of Saruq al-Hadid (Nashef 2010; Contreras et al. 2017; Weeks et al. 2017), as well as at al-Safa near ʿIbri (Sultanate of Oman – al-Bakri, Genchi and Tosi n.d.), still under publication. Moreover, the recent discovery of a cultic building at Mudhmar – near the village of Adam, not far from Salut – where votive deposition of bronze objects has been revealed, will certainly shed new light on these peculiar contexts. It seems however that weapons represented the bulk of the metal objects discovered at Mudhmar (Gernez 2016). 88 Magee 2003a: 186. 89 Haerinck 1994. 90 Lombard 1985: 212-213, figs 108-109; for Hili 8 see also Cleuziou 1978-79: 19-29. 91 Bourchalat and Lombard 1985: 61, pl. 62/15-16. 92 Gernez 2016. 93 Magee 1996a: fig. 20; Magee 2007a: 42-45, fig. 47.

5 . the iron age architecture of husn salut

229

utilitarian tools,94 and as such clearly recall a possible ceremonial destination like the specimen here discussed form Husn Salut. The snake from US74 (pl. 77/1; Chapter 7, no. 8) is indeed the only one coming from a sealed ritual context at Husn Salut. Five complete copper-base representations of a snake and two fragmentary ones have been in fact found, the majority of which come from non-diagnostic contexts in the Terrace I area.95 While the remaining ones are cast in full relief (see Chapter 7), MB143 was obtained by a single, flattened and partly folded metal band. The first evidence of a snake cult96 – in the Iron Age Oman Peninsula – came to light in the small squared building of al-Qusais,97 where plenty of bronze snakes were found. More recently, the existence of this cult has been highlighted at the well known sites of Bithna,98 Masafi 3,99 Saruq al-Hadid, 100 as well as at Husn Salut101 itself. Iron fragments were also recovered from US74 and US75.102 Despite their bad state of preservation, both groups of fragments seem to belong to handles, possibly dagger handles. One of the fragments (MI7 from US75, see Chapter 7, no. 147) preserves two projections for the insertion of the handle, which was probably in perishable material. Strikingly scarcely attested during the Iron Age, iron was considered a prestige good, uncommon and related only to distinctive archaeological contexts.103 However, also in this case recent discoveries may change the picture, as Iron working is being documented at Saruq al-Hadid.104 - Basketry A large amount of vegetal strips was recovered from both US74 and US75; these can be recognized to be elements of woven baskets, as large portions of the latter were discovered in US75. The baskets were at least partially burnt, as shown by the fact that the upper part of the surviving weft was charred. This, together with the presence of animal bones mixed with the basket weft has further complicated the recovery of the remains. However, they provide a clear indication of how they were made, and different manufacture actually indicates that they can be related to at least two different baskets. At least one of them contained dates. As well as informing about the probable oasis environment near to the site, the preserved date stones provided ideal samples for the 14C dating mentioned above (see fig. 27). The basket remains show that it was woven using a technique whereby the strips of palm create a distinctive pattern known as twill plaiting, one of the most diffused and fastest techniques of basketry production both in the past and nowadays.105

94 The blade length of the socketed shaft-hole axe is approximately 7.4 cm and its maximum width is 6 cm; the splaying axe has a length of 11 cm for a width of 4 cm. 95 Despite being secondary, the location of these finds is worth pointing out. Some metal snakes were in fact found between the stone foundations of the Islamic huts built on top of the site, seemingly a deliberate placing. Abundant evidence in fact shows that during the Islamic period copper was extensively mined and worked in the Oman Peninsula (see for example Costa and Wilkinson 1987: 93-131), and the re-melting of useless or broken objects was common; the location of ancient bronze snakes within a living space instead of their re-use as scrap metal could thus suggest an apotropaic use of the object. 96 See Benoist 2007; Potts 2007. 97 Taha 1982-1983. 98 Benoist 2010b; see also Benoist et al. 2015. 99 Benoist et al. 2012a. 100 Nashef 2010. 101 Condoluci 2015b: 190-191; see also Avanzini, Sedov and Condoluci 2005. 102 MI7 from US75 comprises two elongated iron fragments, one of which showing two outgrowths at the opposite edges (dimensions of the larger fragment, in cm : l. 5; w. 2; th. 1.1); MI15 from US74 comprises several fragments of an elongated object, flat in section (dimensions of the larger fragment, in cm: l. 3.8; w. 1.5; th. 0.4). 103 See Magee 1998a. 104 See footnote 86. 105 Sentence 2001: 64-67. This technique involves arranging two sets of strips, one interlaced at right angles to the other. The typological nomenclature of plaiting is based on the spacing of interlaced elements, and reflects the number of elements or strips in each set that are crossed by strips of the perpendicular set. On this basis two main class may be distinguished: simple and twill plaiting (see also Adovasio 1977).

230

chiara condoluci

One of the basket fragments shows in fact a diagonal pattern, where the vegetal elements are perpendicularly interwoven (fig. 92; pl. 77/5). The strips are flat and thin and have the same dimensions (0.3 cm wide). They are woven ‘two up, two down’ (twill plaiting, while the base technique is ‘one down, one up’, also called simple wattle). In this fragment the rim of the basket is clearly visible. The original shape cannot be determined, but the dimensions of the preserved fragment (40 cm long) suggest a flat elongated sack. Also, associated with the basket remains were the fragments of an S-spun fibre (probably date fibre and likely part of a second basket) that has been used to make a loosely woven strap in which the strings of fibres are grouped in fours and plaited in the same manner as the other basket, so as to form a more elaborated ‘two up, two down’ twill pattern (fig. 93; pl. 77/6). Macro-remains analysis106 has shown that the actual basket was made from date-palm and gives a unique indication of the kind of containers that were made from organic materials and thus rarely preserved in the archaeological record. The second fragment also looks very similar, although its preservation did not allow a precise identification. However, the date palm leaf is the raw material primarily used for basketry production in Arabia, by virtue of its ready availability:107 it is thus more than plausible that it was used for both baskets. Ancient basketry is known from a few Bronze Age sites in the Gulf, where it survived thanks to the protection provided by bitumen coating, as is the case at the third and second millennium necropolis of Sar in Bahrain.108 Along with pottery, basketry has formed a major component in the range of domestic objects used by people throughout the world. Unlike pottery, however, studies of ancient basketry are few, mainly due to the dearth of the surviving remains that is intrinsic to their organic nature. Therefore, evidence for basketry often consists only in impressions on more durable materials such as pottery, mud-bricks, or clay surfaces, as in the few cases documented at Husn Salut and mentioned in Chapter 2. Almost three thousand years later, these kinds of baskets can still be seen being used in present day Oman for several purposes, such as food containers (dates in particular) or collecting incense.

F igure 92 – Basket remains in situ, US75.



Bellini et al. 2011: 9. Miller and Morris 1988: 224-226. 108 Højlund 1995. 106 107

F igure 93 – Detail of S-spun fibres from US75, likely part of a second basket or mat.

5 . the iron age architecture of husn salut

231

- Archaeobotanical study Data coming from palynological analysis109 mainly highlight the significant presence of Palmae cf. Phoenix pollen in US75, found inside one of the baskets together with abundant date stones (fig. 94). It has been noted that this percentage, apart from representing the highest percentage throughout the site, is generally recorded only in close vicinity of the male inflorescence. Therefore, the most plausible hypothesis appears to be the offering of date palm flowers in connection with ritual practices, consistently with the interpretation of this deposit and with the mentioned, similar evidence form US454. High percentages of Gramineae pollen were also recorded, including Cerealia (Hordeum and Triticum groups), and the occurrence of Ocimum and Asphodelus pollen. Besides the mentioned basket remains, the collected macro-remains from US75 comprise large charcoal fragments belonging to Juniperus sp., Ziziphus sp., Acacia sp. (this also from US72), and a long wood fragment belonging to Ziziphus sp. Among the numerous burnt date stones some were found clumped in groups together with date remains, which were in all likelihood part of the offering. Important considerations on the presence of these specific plants remains and their use in ritual contexts have been reported elsewhere.110 What is interesting to highlight here is the concomitant presence of the same plants – Palmae cf. Phoenix, Cerealia, Ocimum and juniper micro- and macro-remains – in the two examined ritual deposits, US75 and US454, suggesting that they represented an essential component of the rites. In fact, practical uses for juniper wood are limited and generally restricted to furniture and decorative panel making, since despite its great durability, it is not suited for example for roofing larger spaces. Although its use as construction timber is known in antiquity,111 juniper wood was most commonly burnt for its highly aromatic smoke and in connection with special ceremonies.112 F igure 94 – Charred date stones from US75. - Faunal remains from US75113 Faunal remains were recovered from US75 in a fair state of preservation, which is all the more remarkable as they largely belong to young – in some cases very young – animals, and are as such more prone to deterioration. The remains include largely complete individuals, mainly sheep and goats;114 this, together with the mentioned immature age, supports the ceremonial interpretation of the deposit. Traces of burning at relatively low temperatures are frequent, as are rodent marks, both kinds of evidence identified on ovicaprid remains.115 The bones of some of these same rodents, that most likely entered the context in antiquity when the deposited animal remains still maintained some nutritional interest, have also been discovered

Bellini et al. 2011. Ibid.: 11-13. 111 See for example Marston 2009. 112 See Middeke-Conlin 2014: 23; see also Godbey 1930. 113 This paragraph summarises a preliminary report submitted to the IMTO by B.Wilkens (n.d), who has therefore to be acknowledged for this and all deriving considerations. 114 The detailed figures and related tables are reported in Chapter 8. 115 Wilkens n.d.



109

110

232

chiara condoluci

inside US75 (fig. 95). A confirmation of the fact that living rats were entering this context is provided by the identification of three coprolites that, given their small dimensions, have been attributed to them. Three species were distinguished: the well-represented black rat (Rattus rattus) and one individual each of gerbil (Gerbillus dasyurus) and spiny mouse (Acomys).116 The presence of at least 13 individuals of black rat is of particular relevance. This species, which is of eastern origin, is not autochthonous to the Arabian Peninsula and probably indicates the region’s regular contacts with India and other areas of central-southern Asia. The Arabian Peninsula could therefore represent a stage of its diffusion towards Europe, where it starts to be recorded in the Roman Imperial Age. Among human-laid remains, bovines are witnessed by two fragments only, indicative of one single individual – young or sub-adult, less than three years of age. This scarcity suggests that only portions of the head and of a hind paw were laid down into the deposit. Ovicaprids overall represent 99.11% of the mammals’ remains that have nutritional interest. At least nine individuals could be identified: three infants; one around 3/5 months; one between 15 months and 3 and a half years; one adult generically above two years; one adult above three and a half years; one middle-age adult (wear stage IV), and one old individual (wear stage VII). Although both sheep and goats were distinguished, a large part of the remains could not be differentiated due to the young age (see tables 5 and 6). Some size determination has been possible for sheep and goats, which showed that they comprised notably small breeds, likely as a response to an environment that was not optimally suited for grazing. Small fragments of cuttlefish, an edible mollusc, were also identified, as well as one sweet water (Melanoides tuberculata) and one terrestrial (Enidae) mollusc that are of no nutritional interest due to their small size. Scanty fish remains were present, among which a catfish dermal plate. Avifaunal remains were collected, but their deterioration prevents any identification. This was also impossible for the numerous reptile bones, for which no reference comparison was available.

F igure 95 – Rodent remains from US75.

116 Other rodent remains, generically indicated as Rodentia, are likely to be attributed to these last two species that, due to the absence of other comparable elements, were only identified on the bases of teeth.

5 . the iron age architecture of husn salut

B eyond

materials and techniques .

G eneral

233

remarks

It was discussed above that the compartment technique integrated by massive containment walls, employed in the foundations of substantial architectural works, has antecedents in the South East Arabian Bronze Age and can be considered, after two millennia of use, as a deeply rooted, local architectural tradition. At Husn Salut, or at least in some significant areas thereof, the implementation of this technique appears to be supplemented by ritual acts that are mirrored in the deriving foundation deposits. The expression ‘foundation deposit’ is used here in accordance with R. Ellis,117 to indicate a ‘building deposit’ located in the foundations of a building beneath the floor. The coalescence of the two actions – new construction and ritual deposition – is evident both in relation to the early foundation of the site (HSI phase) and to the large re-structuring following the obliteration of the Basement, almost a ‘re-foundation’ of the site (HSII phase). So far, no foundation deposit is attested in connection with Bronze Age tower construction; therefore, these particular ceremonials seem to be exclusive to the Iron Age period.118 The foundation deposits at Husn Salut are always encased between surrounding walls and subsequently sealed: they can occupy actual compartments appositely built as part of the original foundations of the site, or can be obtained by re-adapting spaces delimited by earlier, re-arranged or buried walls. The ceremonies involved animals sacrifice or at least meals that included meat consumption, whose remains were partially deposited together with potsherds, various objects (generally but not exclusively copper-base items), copper-base prills and slag, palm date fruits or flowers, and possible other plants. Deposition occurred in multiple layers, alternated with fine layers of clay mortar and/or charcoal and ashes, as evidenced from the best preserved examples. Another practice possibly related to a ritual sphere seems to be that of leaving the deposit open for an indefinable period of time, during which the accumulation of water originated a grey, sand-silty laminated deposit. While the stratigraphic evidence for this practice in only known for two foundation deposits,119 it can not be ruled out for others, of which only the base levels survived. What is also of relevance is that a confirmation of this practice is provided by the widespread presence of Rattus rattus bones in the foundation deposits.120 Also know as ship rat, roof rat or black rat, this commensal mammal – not endemic to Arabia but introduced from port areas on the Indian Ocean probably around the third/second millennium – has been completely dependent on humans for food and water. This is the reason why the finding of its remains in archaeological layers has been generally considered a crucial indication of a settled lifestyle vis-à-vis nomadism.121 The Rattus rattus is an arboreal species, and as such it can nest in palm trees, where it feeds on dates: archaeological remains and archaeobotanical studies have undoubtedly shown that palm groves were present in the area surrounding Husn Salut during the Iron Age.122 Moreover, rodents’ gnaw marks have been identified on the studied animal bones from US454 and US75, among which some have been attributed with certainty to the Rattus rattus (US75, see also Chapter 8). This is of the greatest importance in confirming the presence of unsealed deposits, since ‘the black rat is not a burrower by nature and the often assumed intrusive taphonomic status of the species is therefore not valid’.123 Thence, it can be assumed that at Husn Salut a few rats died inside some of these open foundation deposits, attracted by the presence of meal and/or sacrifice remains. In addition, these well-recognizable waterlaid layers, documented both for the early and late phase foundation deposits may address another less obvious issue, that is, the deliberate choice of the builders of a specific period of

Cf. Ellis 1968: 1. Despite the absence of actual foundation deposits of nature of those discussed here, the presence of bronze weapons below the floors or inside the walls of second millennium tombs has been interpreted as possible evidence of such deposits (cf. Carter 1997: 49, 105). 119 US465 (HIS phase) and US72 (HSII phase, see above). 120 These are known foundation deposits in Area 3 (US464), Area 1 (US75), Area 2 (US568), Area 4 (US298). 121 On the evidence for Rattus rattus in South East Arabia settlements, see Mosseri-Marlio 2003. 122 Cf. Bellini et al. 2011. 123 Cf. Mosseri-Marlio 2003: 310 (quoting A. Ervynk). 117 118

234

chiara condoluci

the year for these construction activities: during the rainy season.124 This is not surprising, since in the Ancient World (and more specifically in the regions surrounding the Persian Gulf) the construction or re-construction of temples was often subject to a specific calendar indicating the propitious moments125 and always, of course, followed the acceptance by a divinity of the building place. The only information about foundation ceremonies at Husn Salut is provided by the materials intentionally deposited. Despite the similarity among them being undeniable and defining a sort of standard in the ritual experience, a more detailed evaluation reveals different practices. As mentioned above, some dissimilarities are manifest among deposits belonging to the earlier phases (Area 3, Area 8, Area 4). Those from Area 3 always contain only one fragment of each vessels, pars pro toto. The vessels are supposed to be used during the ritual ceremony and then deliberately broken, possibly as an exorcist act.126 Although the presence of vessel shapes directly related to banqueting is attested (carinated cups and spouted jars), no long-handle bowl shard has been documented. This may indicate a different ritual compared to those evidenced in other foundation deposits where a long-handle bowl is always present. Despite animal bone analyses being still under way, it is possible to say with confidence that, as in the case of the pottery, only limited parts of the animals were laid inside the deposits. Among these, one contains disarticulated bones of young animals less than one year old, as indicated by the presence of non-welded epiphyses (US612), supposed to belong to sheep/goat (see Chapter 8). Whether the bones were burnt or not is uncertain. In Area 8, US454 includes larger parts of vessels deliberately broken, but not complete, still bearing the pars pro toto concept. However, what is striking is the large amount of sherds and the great diversification in shape, a fair number of which, as mentioned above, find no comparison elsewhere at the site. Actually, US454 represents the largest foundation deposit discovered at Husn Salut. If the presence of such a number of vessels really derives from their use during the ceremony prior to the deposition, it would suggest the participation of a large community. A fundamental part of these ceremonies was the sharing practices of banqueting and libations, testified by the variety of spouted vessels and carinated cups for drinking liquids and by the remains of meat consumption. The presence of several long-handle bowls is also suggestive of the combination of this feasting with ritual processions that could also take place at night time. Sheep and goats have been the predominant sacrificed animals for the great meal, again comprising less than one year old individuals and with a predominance of females. In contrast to the rite performed in the foundation deposits from Area 3, the bone remains are representative of most of the anatomical districts, attesting that they were sacrificed and prepared on the site. While slaughter marks have been identified, also in this case it is not possible to determine the method of meat cooking and consumption, as no burning trace survives. Other animals were identified in the deposits and, although representing a small proportion of the assemblage, deserve mention. They include scarce remains of wild fauna – gazelles, foxes, partridges, chelonids – and domesticated fauna such as cattle, camels and cats. The presence of (domesticated) feline remains is somewhat eccentric and unparalleled in Arabia to the author’s knowledge, and its significance remains elusive. Conversely, the presence of the sea turtle attests to contacts with coastal communities, as do fish and mollusc remains (see Chapter 8). It is tempting to relate this evidence to possible visits of members of these distant communities, possibly intended for participating in the ceremonies held at the site. Such an explanation would also justify the presence inside US454 of a few grey ware sherds, not witnessed elsewhere at the site and almost certainly coming from outside. Further significant information emerging from US454 is the early date it provides for the attestation of a ‘naturalistic’ snake figure, represented on two sherds, opposed to more stylized ones that are known from other, equally early contexts (e.g. US298, see Chapter 4). This is of relevance in underlining the profound and original link of the site with the snake cult, blatantly manifest for the HSII phase in the materials deposited inside US75. In the Basement,

124 Such a choice could also facilitate the progress of the work, beginning with clay mortar production so needed at Husn Salut. 125 See for example Ellis 1968: 17. 126 Similarly regarded as a magic act for analogy: breaking the pots in order to ‘break’ the evil in same way. See for example Ambos 2007: 312.

5 . the iron age architecture of husn salut

235

this type of snake representation only appears after the great fire (US35), while in the Burnt Building one single small bronze snake is attested in the fire-related layers (US394). The results of the archaeozoological study on bones from US297 and US298,127 all disarticulated, are in strong concordance with those from US454. Adult goats continued to be the favoured sacrificed animals, with the presence of young individuals, followed by other domestic and wild animals. Goat bones (disarticulated) from US297 include most of the anatomic districts excluding limbs, while in US298 all parts are represented and the presence of at least 11 individuals has been revealed. Conversely, only specific anatomic parts, mainly limbs (and sometimes only a single find), seem to be representative of other animals, such as camels, cattle, donkeys, birds, foxes and wolves. Camel remains from US298 are supposed to belong to 3 individuals. Rattus rattus was also largely attested in US298, its significance discussed above. The main foundation deposit belonging to the HSII phase is the one laid down inside former room 5b of the Burnt Building, characterized by an elaborately stratified layout within which the main ritual deposition is represented by US74+75. Ceremonial practices evinced from the material there buried show general affinity with previous foundation deposits, although the snake cult appears now more strongly attested by the presence of a new category of cultic objects, the bronze snakes, associated with the more common long-handle bowls bearing applied snake decorations. The scanty potsherds still represented a pars pro toto deposition, and banqueting still is a praxis during the celebration, indicated by several animal bone remains, large amounts of burnt stone dates, and by specific objects such as bronze ladles and carinated cups (see above). Sheep and goats represented 99% of the animal remains (at least 9 individuals including 3 newborns have been identified), of which all the anatomic parts were deposited, although dislocated. Only a low percentage of their bones showed burning traces,128 an occurrence that may suggest the use of two different ways of preparing the ‘ritual meal’ – grilling and boiling? –, or the mixed deposition of cooked parts and fleshless bones.129 Also used during these meals was cattle, of which only the bones of the skull and legs were found. In concordance with the results from other foundation deposits (US454, US298) it seems that the low percentage of mammals other than sheep/goats – i.e. the deposition of only some parts of the animal – was intentional in these ritual contexts, as if an explicit formulary, also enumerating the ‘right equipment’, was followed.130 Also intriguing is the presence of edible molluscs; while the presence among them of remains of sepias sp. attest once more to relations with the coastal regions (see also chelonid and fish remains from US454), the presence of fresh water (Melanoides tuberculata) and terrestrial molluscs (Enidae), devoid of food interest, still need to be explained, although the latter can be related to the upper water-laid accumulation layer US72. A catfish dermal plate was also found in the stratum. Rattus rattus is attested by a minimum of 13 individuals and also by some coprolithes. The prestige nature of the metal and pottery artefacts from US75 has been already discussed (Chapter 5); likewise, the significance of the snake figure for Iron Age society has been the subject of several papers over the recent years.131 However, to the countless symbolic meanings this chthonic animal possesses – related as they are to fresh water emerging from the ground, keeper of knowledge, immortality, renewal, magic, including a connection with metallurgy, just to enumerate the most known ones – a more practical consideration can be added, connected with the repercussion of the Rattus rattus appearance in the Oman peninsula. Rat or rodent remains have been documented and studied from second millennium sites in South East Arabia, such

Cf. Chapter 8. Wilkens n.d.: table 3. 129 In this regard it is interesting to note that gnaw marks indicate a predilection of rodents for unburned bones (boiled or uncooked? – see Wilkens n.d.: tables 3, 4), possibly because the burnt ones were devoid of any food nutrition, having been consumed before being placed in the deposits. 130 By way of example, see the list of needs for exorcism procedures to be followed when the Mesopotamian god Kulla has to leave a new building (Ambos 2007: 311). Concerning the god Kulla and his possible implication with Husn Salut see below. 131 E.g. Benoist 2010b, also including mention of Husn Salut (there still indicated as Salut); Benoist et al. 2015. 127



128

236

chiara condoluci

as Kalba,132 Tell Abraq133 and Ras al-Jinz.134 To these can now be added the evidence for their considerable presence at Husn Salut. As pointed out above135 rats are dependent on humans for food and water, especially in marginal and arid environments such as South East Arabia, which could never have been a natural habitat for the species. The shattering impact of rat colonies on human food reserves needs to be highlighted – especially on an emerging economy such as that of the Iron Age society was –, together with the consequent epidemics that may have led to a population lacking immunity for certain disease. Therefore, the hypothesis of worshipping of one of the main predators of the rats seems not a flight of fantasy at all. Similarly, one might wonder if the bizarre presence of one or two cat remains in US454 could be related to an analogous meaning. The other foundation deposit dated to a late phase of the site, US568 (see above), is linked to the levelling of SF49 – and thus probably in phase with US74+75 deposition, which appears to represent some sort of platform – or altar (?). Although their scarcity, the materials found in US568, reflect the same ritual practices repeatedly mentioned above, excluding the presence of animal bones but including a complete copper-base spouted bowl. To sum up, it has been shown that some kinds of ceremonies took place at Husn Salut, in connection with significant foundation (or re-foundation) works, and they mainly comprised partaking of banquets, these consisting of meat, drinks, and date fruits. A precise selection was made for the animals to be eaten and/or sacrificed: first of all sheep/ goats, mainly adults and preferably female, followed by a lower percentage of other domestic and wild animals. Portions of the sacrificed animals became part of the deposition, with sheep/goats generally stored with most of their anatomic parts, while only one or few portions of other animals were deposited. This evidence recalls ritual sacrifices in Ancient Mesopotamia and beyond, where sheep and also goats were the most sacrificed animal to any god.136 The scale and the importance of these ceremonies are such that they suggest a communal relevance, and possibly a participation restricted to small groups of ‘notables’. Foundation deposits at Husn Salut also comprised part of the pottery sets supposedly used during the feast, some symbolic objects and scattered copper-base scraps. Date fruits, Cerealia and Sesamum were also offered, together with plants bearing a symbolic significance, such as date palm inflorescences and probably Ocimum and Nerium oleander. Juniperus wood (documented both in US75 and US454) suggests the use of fumigations with aromatics, testified also by the presence of long-handle bowls.137 Several materials from Husn Salut’s foundation deposits directly recall those documented at the cultic sites of Bithna and Masafi 3, where offerings and cultic objects – snakes and bronze weapons, incense burners, long-handle bowls, bronze scraps, meat – were found both inside small shrines and in the adjacent external area. At Bithna, they were discovered above or in the vicinity of mud-brick platform altars (including open-air ones), or in specially excavated pits outside the shrines, which contained dislocated animal bones and/or bronze scraps.138 At Masafi 3, the concentration of votive objects in the centre of the building suggests the presence of altars/structures made of perishable materials. Analogous practices have also been documented in pillared buildings. For instance, the deposition of bronze scraps and prills inside distinct pits or scattered on the floor is attested at Masafi1 (stocked inside jars)139 and Muweila (near

Mosseri-Marlio 2003; Phillips and Mosseri-Marlio 2002: 206. Cf. Stephan 1995: table 2 (Rodentia). 134 Cf. Cleuziou and Tosi 2000: 43. 135 See also Mosseri-Marlio 2003. 136 Cf. Scurlock 2006: 18-20. The same author suggests that this may have to do with a purely economic reason, as the male of the species was considered a luxury and generally more valued than the female, a consideration also extended to bull sacrifices in Ancient Israel, generally offered by the entire community (ibid.). 137 GC-MS analysis on burnt organic remains found inside one long-handle bowl’s reservoir revealed the presence of a substantial quantity of fatty acids retraceable to the use of animal fat or vegetable oils (M.P. Colombini and E. Ribechini pers. comm. 2008). However, this result does not imply that censer-lumps were not used for fumigation as well as lamps, given the shape of the object that provides both functions and given also the high number of censer-lumps recovered on the site. 138 Benoist 2010b: 114. 139 Benoist et al. 2012a: 153. 132 133

5 . the iron age architecture of husn salut

237

a fireplace).140 These particular offerings related to copper working seem a direct reference to metallurgical skills and are probably connected to the snake cult,141 more than to the possible social relevance of the metallurgists themselves. However, all the ritual practices just mentioned express common devotional acts carried out by the (single or groups of) worshippers at cultic sites, and have no connection to foundation ceremonies, which may be assumed to be more usually attended by the ‘authorities’. At Husn Salut, similar ‘private’ cultic structures and related offerings are not clearly manifest, most probably due to the bad preservation of the upper area of the site near Building 1, where it is presumable they were located. In fact, the presence of a large platform-altar, modified during the HSII phase (SF49), seems to be rather certain (see Chapter 2, Chapter 3). Ritual practices possibly connected to structure SF49142 seem to be testified by the presence of an oven with a bronze cauldron find still in situ, SF21 (see Chapter 2), and by a series of unfortunately badly preserved, shallow circular pits scattered in the area immediately north of it and of SF49. Some of these pits still contained scanty remains of animal bones, charcoal, potsherds and copper-base scraps. From one in particular (US28) come a non-utilitarian copper/bronze axe (Chapter 7, no. 16) and some iron fragments. It is evident that the material from these pits closely corresponds to that from the foundation deposits at Husn Salut; moreover, the latter shows great uniformity in the amount and typology of the objects as well as in the sacrificed animals and, in conclusion, this seems to mirror the existence of some ‘protocol’ that had to be followed during the ceremonies, of whatever nature they were – private or communal. Last but not least, mention is due to the retrieval of a brick fragment from US75, which still showed the presence of a thin veil of purple powder on one of its surfaces. This unique find could in fact be of considerable symbolic value, similar to, if not greater than, all the other materials found in the stratum. To try to unveil its possible meaning it is necessary to refer to areas where abundant written documentation about foundation rituals enlighten their interpretation, as is the case for Ancient Mesopotamia – a totally opposite situation from South East Arabia and its well known lack of textual or iconographic sources. This approach to ritual practices described in first millennium BC Mesopotamian texts,143 has the purpose of suggesting a ‘standard’ – therefore providing its better understanding – for that part of the foundation ceremonies carried out at Husn Salut that remain invisible to the archaeologist’s eye, and of which the materials discovered in the deposits described above only represent the durable material outcome.144 To begin with, it is necessary to say that in ancient Mesopotamia the rituals associated with both construction and reconstruction of a temple – a significant part of royal ideology – were reputed to be strictly necessary for the protection of the builders (also referred to as customers), as any of the innumerable edification acts, whether done badly or carried out at a not propitious moment, could trigger the fury of the deities, a fury which could be hurled upon the whole community. Consequently, building rituals were of vital importance for the successful fulfilment of the building process and the well-being of the builders. Besides, when the building was completed, it had to be ritually purified before it could be used. Some ancient manuals exist, which explain the rituals related to the erection of a temple or a sanctuary, including

Magee et al. 2002: 144. Cf. Benoist et al. 2015. Copper working-related items are numerous and widely spread at Husn Salut, especially in the late accumulation levels above Building 1’s central rooms, and were at first interpreted as stored material for re-melting (see also Chapter 9). However, no feature clearly related to such activity has been unearthed, and the discussed parallels from other sites such as Bithna and Masafi 3 may indicate that also these items could have been part of votive offerings. 142 This is curiously located to the south of Building 1, like the platform-altars in Bithna temples J and K are located in the southern part of those small buildings. 143 In the cuneiform tradition of the first millennium BC, a considerable corpus of building rituals has been preserved on tablets from Nineveh, Assur, Dûr-Sharrukîn, Huzîrîna, Uruk, Sippar, and Babylon. See Ambos 2007, Ambos 2010. 144 This approach finds justification in the evidence for substantial exchanges between the Oman peninsula and the surrounding areas, specifically those bordering the Persian Gulf but also beyond, occurring at least from the third millennium BC onward. In fact, these links, based on the exchange of valuable goods and raw materials, certainly also allowed the transmission of ideas, making plausible the diffusion of a common perception of the divine over a wide area (see the interesting discussion about ‘borrowing’ in Scurlock 2006: 14-17).

140

141

238

chiara condoluci

its construction, the introduction of divine images in the temple and its consecration – rituals that were outlined of competence of specific figures such as the exorcist, the singer of lamentations, the musician, and the aruspice. More specifically, from a catalogue of exorcist rituals, one is informed that Kulla – the name of the Mesopotamian brick-god145 – is also the term used for a series of rituals for the construction of a sanctuary.146 Kulla, son of the god of wisdom Enki/Ea, represents the personification of the brick and the divine architect who oversees the building construction (along with Mušdama). Different kinds of rituals were performed depending on whether the foundation of new temples or re-construction of old temples, even when still in use, was concerned.147 In the latter circumstance (either following damages/collapse or related to restoration), it was essential that the new foundations kept in consideration the original plan of the temple – supposedly to avoid any misunderstandings with the god because his house was being demolished and his services interrupted. At least theoretically and ideologically, the customer should necessarily find the original foundations to preserve the architectural integrity of the temple and ensure its continuity. However, in the texts there is no mention of old foundations’ discovery and exposure, but another ritual is described, according to which ‘the first brick’ or ‘ancient brick’ should be removed from the ruins of the old temple and be set aside, in order to begin the act of renewal.148 Consequently, standing in front of the ‘brick’ – another case of pars pro toto, the brick as the essence of the sanctuary – the lamentation singer (kalû) tries to appease the possible rage of the god with his lamentations, which are accompanied by libations and animal sacrifices. A combination of two of these text fragments related to this specific ritual149 is reported below as an example of a series of acts, which might have been accomplished during the foundation ceremonies at Husn Salut. It is worth underlining some correspondences between the described ritual and some features related to HSI and HSII phases’ foundation deposits (US337 and US74+75): the intrinsic value of the axes (MB196 and the miniature axe MB122), possibly in connection with the ‘cutting’ of the old mud-brick; the presence of a purple substance on the mud-brick, possibly testifying a libation performed on it; and, not less significant, the performing of the ritual in an isolated room – as the location of room 5b was, inside the Burnt Building. The builder of that house puts on a pure garment (and) he puts a ring of tin on his hand. He takes an axe of lead and removes a ‘former brick’ (libittu maḫrītu) (from the debris/the collapsed wall) and places it in a restricted room. You place an offering table in front of the brick for the god of the foundations (var.: the goddess Bêletilî). You perform an offering, scatter (aromatic) seed(s) of all kind, (and) make libations of beer, wine, and milk. … They offer honey, ghee, milk, beer, wine and fine oil on (the brick) and the lamentation singer recites the incantation “When Anu created the heavens” before the brick.150

145 It is worth underlining the value of the ‘mud-brick’ in Ancient Mesopotamia, a raw material associated with mythical concepts, generally reserved to the construction of temples and not to public buildings, usually built with baked bricks (cf. Ellis 1968: 17). 146 Ambos 2007: 299. 147 Ambos 2010: 227-228. 148 Ellis 1968: 13; Ambos 2007: 304. 149 O 174: 13-19 // BE 13987: 14-23. 150 Ambos 2010: 227. See also Ellis 1968: 27, and Appendix A, footnote 43, p. 184. For a more detailed discussion of the ‘first brick’ topic and related written sources, see Ellis 1968: 17-34.

plates

240

chiara condoluci

P late 56 – Area 2, Terrace I: general plan with location of the two test trenches and section of Trench 2.

5 . the iron age architecture of husn salut

-

plates

P late 57 – Area 1, Burnt Building: room 4b, plan and section showing the foundations’ levels.

241

242

chiara condoluci

P late 58 – General plan of Area 3, with main section transversal to the outer wall structure. Grey shades highlight the different material employed in the construction.

5 . the iron age architecture of husn salut

-

plates

P late 59 – Area 1, Area 3 and Area 8: location of the investigated foundations deposits.

243

244

chiara condoluci

P late 60 – Area 1, Burnt Building: room 5b, plan and section lines.

5 . the iron age architecture of husn salut

P late 61 – Area 1, Burnt Building: sections through room 5b.

-

plates

245

246

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

chiara condoluci

P late 62 – Area 2, Terrace I: materials from early phase compartments; Area 3, phase HSI foundation deposits. F orange. Small white and black grits, occasional large ones. Traces of dark red slip exterior. US658,1. M light brown, overfired. Small, medium and large red and grey inclusions; vegetal temper. US659,1. F light orange. Sparse, small grey grits, small and tiny white grits; vegetal temper. Red slip exterior, dark red interior. US659,2. F/M light brown with grey core. Small black grits, occasional white grits. Red slip exterior, light orange interior. US660,1. F orange. Tiny white and grey grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US584,1. M/F orange with light grey core. Small and medium red grits; vegetal temper. Red-orange, self-slipped exterior and interior. US451,1. F light brown-creamy. Tiny white and red grits; occasional large red and white grits. Graffito decoration interior. US293,7. F orange-brown. Occasional tiny white and red inclusions. Red slip exterior and interior. US293,6. M pale brown. Small grey grits, occasional tiny white grits; sparse vegetal temper. Cream, self-slipped exterior and interior. US293,5. F dark orange. Abundant small white grits, occasional small grey grits. Dark red slip exterior and interior. US293,8. F brown buff with light grey core. Tiny white grits. Dark red slip exterior, brown interior. US293,4. M orange-brown with grey-black core. Rare medium black grits, occasional white grits. Brown slip exterior and interior. Burnished interior. US293,1. C orange brown with black core. Large and medium red and grey grits, occasional white grits. Light red slip exterior and upper rim, brown interior. US293,2. C pale brown with grey core. Abundant very large (> 5 mm) and medium red and grey grits; occasional tiny white grits. Pale orange slip exterior and interior. Burnt traces on surfaces. US293,3. F pinkish-orange. Tiny red grits, occasional white grits; vegetal temper. Cream slip exterior. US296,7. F orange. Abundant tiny black grits, occasional tiny white grits. Self-slip exterior. Dark painted exterior. US296,5. F pale orange. Small white grits. Creamy slip exterior and interior. US296,6. F orange-brown. Small red grits, tiny white grits. Pale red slip exterior and interior. US296,8. F orange-brown. Abundant tiny white grits. Orang-brown, self-slipped exterior, pale brown slip interior. US296,9. F orange. Abundant small and tiny white grits, grey grits. Pale red-brown exterior, orange-brown, self-slipped interior. US296,10.

5 . the iron age architecture of husn salut

-

247

plates

1 2

3

4 5

8 7

6

9 10

11

12

14 13

15

18

17

16

19

20

P late 62 – Area 2, Terrace I: materials from early phase compartments (1-4); Area 3, phase HSI foundation deposits (5-20). Scale 1:3.

248

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci

Plate 63 – Area 3: materials from phase HSI’s foundation deposits. M/F orange with grey core. Small grey grits, occasional small white grits; vegetal temper. Light brown slip exterior, dark brown on rim, pale brown interior. US296,3. M brown with grey core. Medium brown grits, small grey grits, rare white grits; vegetal temper. Pale red slip exterior. Traces of burnt exterior and spout. US296,4. M light orange. Abundant medium and small, red and white grits, occasional large grey grits. Dark brown slip exterior, pale red interior. US296,1. M orange with grey core. Small and large white grits; vegetal temper. Dark slip exterior and rim, pale brown interior. US296,2. C red-brown with grey core. Small, medium and big grey grits, occasional white inclusions; vegetal temper. Dark red slip interior, brown slip exterior. Burnished interior. US296,17. M orange-brown fabric. Big and medium red grits, very small white inclusions. Self-slip exterior. Overfired interior. US296,16. M brown with light grey core. Small and medium red grits. Pale red slip exterior. Smoothed exterior. US296,15. C dark orange with light grey core. Small and medium white grits, large red and grey grits. Red slip exterior and interior. Burnished interior and rim. US296,13. C orange-brown with grey core. Red and grey grits, occasional tiny white grits. Red slip exterior and interior. Burnished interior and rim. US296,12.

5 . the iron age architecture of husn salut

-

249

plates

2 1

4

3

5

6

7

8

9

P late 63 – Area 3: materials from phase HSI’s foundation deposits. Scale 1:3.

250

chiara condoluci

Plate 64 – Area 3: materials from phase HSI’s foundation deposits. C orange-brown with dark grey core. Occasional large grey grits; vegetal temper. Reddish-brown slip exterior, brown interior. Burnished interior. US296,11. 2. C orange with grey core. Large grey grits, occasional small white grits. Pale orange slip exterior. US296,14. 3. F brown-orange. Tiny white and grey grits, occasional large red grits; vegetal temper. Pale red slip interior. US299,2. 4. F brown-orange. Small white and grey grits. US299,1. 5. M/F orange. Tiny white grits, occasional large ones. Red slip exterior, dark red interior. US299,4. 6. F brown. Tiny white grits, occasional grey grits. Red slip exterior, brown interior. Burnished interior. US299,3. 7. C orange-brown. White and red grits, occasional large red grits. Orange brown, self-slipped exterior and interior. US299,5. 8. F orange. White and grey grits, occasionally large white grits. Cream slip exterior and interior of neck. US299,6. 9. F orange-brown. Medium red and grey grits, small white grits. Dark slip exterior, red on spout, cream interior. US612,1. 10. F brown with light grey core. Tiny white grits, red and grey grits. Brown, self-slipped exterior and interior. US612,2. 11. F brown. Tiny grey grits, occasional medium white grits. Brown slip exterior. US612,3. 1.

5 . the iron age architecture of husn salut

-

251

plates

1

2

4

3

5

6 8 7

10 9

P late 64 – Area 3: materials from phase HSI’s foundation deposits. Scale 1:3.

11

252

chiara condoluci

P late 65 – Area 3: materials from phase HSI’s foundation deposits. F light brown. Occasional medium grey and red grits. Dark brown slip exterior and interior. US464,3. C light orange with dark grey core. Medium red and white grits; vegetal temper. Light red slip exterior, brown interior. Impressed decoration exterior. US464,2. 3. C orange with dark grey core. Medium grey grits, occasional tiny white grits; vegetal temper. Light brown-reddish slip exterior, pale brown interior. Impressed decoration exterior. US464,1. 4. Half of a possible tool/knife handle made from animal bone. US464,4 = B11. 5. F dark orange with grey core. Small white grits; sparse vegetal temper. Dark red slip exterior, brown interior. Dark painted decoration exterior. US471,3. 6. F dark orange with light grey core. Tiny grey and white grits. Dark red-brownish slip exterior, red interior. Blistering on exterior (overheating?). US471,1. 7. F dark orange with grey core. Small white grits; sparse vegetal temper. Red slip exterior, dark red interior. US471,2. 8. F orange. Small red and medium grits, occasional tiny white grits. Dark slip exterior and interior. US471,7. 9. F orange-brownish. Occasionally tiny white and grey grits. Dark slip exterior and interior. US471,6. 10. F orange-brownish. Small white grits, occasional medium ones. Red slip exterior and interior. US471,11. 11. F light brown. Small and medium grey grits, occasional small white grits. Pale red slip exterior. US471,5. 12. F pale yellow. Occasional small grey grits; vegetal temper. Pale yellowish, self-slipped exterior and interior. US471,4. 13. F orange-brownish. Red grits. Dark red slip exterior and interior. Incised decoration exterior. US471,8. 14. C light orange with dark grey core. Large and medium white grits; vegetal temper. Pale red slip exterior, red interior. US471,9. 15. C orange-brown with dark grey core. Large and medium red grits. Traces of red slip exterior. US471,10. 16. Bent spearhead. US471,12 = MB204 (see Chapter 7, no.76). 1. 2.

5 . the iron age architecture of husn salut

-

253

plates

1 2

3

4

8

5 7

6

9

10

11

12 13

14

15

P late 65 – Area 5: materials from phase HSI’s foundation deposits. Scale 1:3, except no. 16: scale 1:4.

16

254

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

chiara condoluci

P late 66 – Area 8, sample from US454: cups; carinated cups; bowls. M brown-pinkish. Rare white grits. Self-slip exterior? Dark red painted decoration exterior and bottom interior. US454,123. F pale brown. Small white grits, occasional grey grits. Red slip exterior and interior. US454,69. M brown with grey core. Occasional tiny white grits, rare medium red grits; rare vegetal temper. Red slip exterior and interior. US454,94. F brown. No visible grit. Dark red painted decoration exterior and bottom interior. US454,31. M pink. Rare medium and large white grits. Purple-red slip exterior and interior. Dark red painted decoration exterior and interior. US454,32. M pink. Occasional large white grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,33. M brown. Abundant tiny white grits, sparse small grey grits. Red slip exterior and interior. Dark red painted decoration exterior and bottom interior. US454,9. F light brown. Small white grits, rare grey grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,127. F brown. No visible grit. Red slip exterior and interior. Dark red painted decoration exterior. US454,125. M pink. Rare tiny white grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,124. F pink. Rare white and small red grits. Red slip exterior and interior. Dark red painted decoration exterior. US454,110. M pink. Rare white grits. Red-purplish slip exterior and interior. Dark red painted decoration exterior and interior. US454,122. M dark brown-greyish. Small white and red grits. Buff slip exterior and interior. US454,105. F brown-pinkish. No visible grit. Buff slip exterior and interior. US454,102. M light brown. Small and medium red grits, some large, occasional small grey grits, tiny white grits. Red slip exterior and interior. US454,6. M light brown-pinkish. Occasional small and medium grey grits. Red slip exterior and interior. Dark red painted decoration exterior. US454,49. M pink with grey core. Sparse grey grits, rare white grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,47. F brown. Rare tiny white grits, occasional small red grits; occasional vegetal temper. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,48. M light brown. Medium red grits; sparse vegetal temper. Traces of buff slip? S454,68. M pink. Small and tiny white grits, some large, red grits. Red slip exterior and interior. US454,70. M brown. Abundant small red grits, rare tiny white grits. Red slip exterior and interior. US454,67. F red-pinkish. No visible grit. Red slip exterior and interior, burnished. Dark red painted decoration on rim. US454,128. M pink with grey core. Small and medium red grits; sparse small white grits. Red slip exterior and interior. US454,77.

5 . the iron age architecture of husn salut

6

8

7

12

11

255

plates

3

2

1

-

5

4

10

9

13

15

14

16

17

19 18

20

21

22 23

P late 66 – Area 8, sample from US454: cups; carinated cups; bowls. Scale 1:3.

256

chiara condoluci

P late 67 – Area 8, sample from US454: bowls. M light pink. Abundant white grits. Red slip exterior and interior. US454,85. M brown-pinkish. Red and grey grits, sparse white grits. Red slip exterior and interior. US454,88. C brown-orange with grey core. Abundant medium and large red grits, sparse white grits. Abundant vegetal temper. Red slip exterior and interior. US454,116. 4. M pale brown. Abundant red grits. Red slip exterior and interior. Dark red painted decoration exterior and interior. US454,14. 5. M red-pinkish. White and red grits. Red slip exterior and interior. US454,18. 6. M light brown. Red grits, some large. Red slip exterior and interior. US454,11. 7. M brown. Abundant red grits. Red slip exterior and interior. US454,5. 8. M brown-reddish. Small and medium red grits. Dark grey-blackish slip exterior and interior. US454,74. 9. M pinkish. White and red grits. Red slip exterior and interior. US454,171. 10. M pale brown. Small and medium red grits. Red slip exterior and interior. US454,72. 1. 2. 3.

5 . the iron age architecture of husn salut

-

257

plates

1

2

3

5

4

6

7 8

9 10

P late 67 – Area 8, sample from US454: bowls. Scale 1:3.

258

1. 2. 3. 4. 5. 6.

chiara condoluci

P late 68 – Area 8, sample from US454: large bowls. M pinkish with grey core. Occasional small grey grits, tiny white grits. Sparse vegetal temper. Red slip exterior and interior. Incised decoration on rim. Paint stains exterior. US454,79. C brown-pinkish with grey core. Abundant white and red grits. Red slip exterior and interior. US454,12. C light brown. Abundant medium red grits, occasional small white grits. US454,2. C light brown-reddish with grey core. Abundant medium and large red grits, sparse small white grits. Red slip exterior and interior. US454,111. M brown-violet. Abundant small white grits, abundant medium grey grits. Dark brown-greyish slip exterior and interior. US454,65. M light brown. Sparse medium grey grits. Dark brown-greyish slip exterior and interior. US454,90.

5 . the iron age architecture of husn salut

-

259

plates

1

2

3

4

5

6

P late 68 – Area 8, sample from US454: large bowls. Scale 1:3.

260

1. 2. 3. 4. 5. 6. 7.

chiara condoluci

P late 69 – Area 8, sample from US454: bridge-spouted bowls and bridge-spouted jars. F/M brown. Sparse tiny white grits. Red slip exterior and interior. US454,27. F/M brown. Rare tiny white grits. Red slip exterior and interior. US454,25. F buff. Occasional large red grits, rare tiny white grits. Red slip exterior and interior. US454,26. M brown. Abundant small red grits, sparse white ones; abundant mica. Red slip exterior and interior. Red-purplish painted decoration exterior and rim. US454,40. M pink. Sparse small white grits. Red slip exterior and interior. Bordeaux painted decoration exterior and rim. US454,118. M pale pink. White and red grits. Red slip exterior and interior. US454,13. M pink. White and red grits. Red slip on exterior, darker neck interior. US454,39.

5 . the iron age architecture of husn salut

-

261

plates

2

1

3 4

5

6 7

P late 69 – Area 8, sample from US454: bridge-spouted bowls and bridge-spouted jars. Scale 1:4.

262

chiara condoluci

P late 70 – Area 8, sample from US454: jars; globular jars with strainer and/or tubular spout. M brown. Sparse grey grits. Dark grey slip exterior and interior. US454,113. M light brown with grey core. Red grits. Red slip exterior and interior. US454,57. M pink. Red and white grits. Red slip exterior and interior neck. Dark red painted decoration exterior and rim.US454,120. M pale brown. Abundant red grits. Red slip exterior and interior. Dark red painted decoration exterior and rim.US454,121. M dark brown with grey core. Sparse red grits, occasional tiny white grits. Red slip exterior, neck interior and interior. Black painted decoration exterior. Incised hatching on cordon.US454,55. 6. M pink-reddish. Red and white grits. Dark brown-greyish slip exterior and interior. US454,59. 7. M red. Sparse small white grits. Red slip exterior and interior. Incised decoration exterior. US454,63. 8. M brown-reddish with grey core. Sparse red grits, occasional white grits. Red slip exterior and interior. US454,44. 9. M brown-reddish with grey core. Sparse red grits, occasional white grits; abundant vegetal temper. Red slip exterior and interior. US454,43. 10. M brown-pinkish. White and red grits. Red-purplish slip exterior and interior. US454,180. 11. M light brown. Red grits, some large. Red slip exterior and interior. US454,38. 12. C brown-pinkish with grey core. Abundant red grits, some large, sparse small white grits. M brown-reddish with grey core. Sparse red grits, occasional white grits. Red slip exterior and interior. US454,12. 1. 2. 3. 4. 5.

5 . the iron age architecture of husn salut

-

plates

2

263

4

3

1

6

5

7

8

9

10

11

P late 70 – Area 8, sample from US454: jars; globular jars with strainer and/or tubular spout. Scale 1:4.

12

264

1. 2. 3.

chiara condoluci

P late 71 – Area 8, sample from US454: globular bridge-spouted jars. M light brown-pinkish. Abundant tiny white grits. Red slip exterior and interior. Red-purplish painted decoration exterior and rim. US454,37. M brown with grey core. Small and medium red grits, occasional tiny white grits. Red slip exterior and interior. Red-purplish painted decoration exterior and rim. US454,66. M brown-pinkish. Sparse grey grits. Red-purplish slip exterior and interior. Dark red painted decoration exterior and rim. US454,15.

5 . the iron age architecture of husn salut

-

265

plates

1

2

3

P late 71 – Area 8, sample from US454: globular bridge-spouted jars. Scale 1:5.

266

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci

P late 72 – Area 8, sample from US454: hole-mouth globular jars. M brown with grey core. Red and white grits. Red slip exterior and interior. US454,181. C brown. Abundant red grits, some large, sparse white grits. Self-slip exterior. US454,112. M brown-pinkish. Abundant grey grits, sparse white grits. Red slip exterior and interior. US454,179. C brown. Abundant red and grey grits. Red slip exterior and interior. US454,183. M red-orange. Abundant white grits, sparse red grits. Light red slip exterior and interior. US454,130. C brown-reddish. Abundant red and grey grits. Red slip exterior and interior. US454,184. M brown with pinkish core. Red and white grits; occasional vegetal temper. Red slip exterior and interior. US454,50. M light brown-pinkish. Red grits. Red slip exterior and interior.US454,53. M brown-pinkish. Sparse red grits, sparse small white grits; occasional vegetal temper. Red slip exterior and interior. US454,185.

5 . the iron age architecture of husn salut

-

267

plates

1

2

3

4 5

6 7

8

P late 72 – Area 8, sample from US454: hole-mouth globular jars. Scale 1:5.

9

268

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci

P late 73 – Area 8, sample from US454: deep and large basins; jar lids; pan. M pink. Sparse grey grits, sparse white grits, some large. Red slip exterior and interior. US454,132. M brown. Grey and white grits, some large. Red slip exterior and interior. US454,36. C light brown with grey core. Abundant red grits, sparse small white grits, US454,1. C light brown-pinkish. Abundant red and white grits. US454,10. M light brown-pinkish. Red grits, sparse small white grits. Red slip exterior and interior. US454,170. C buff. Abundant red and white grits. Red slip exterior and interior. Incised decoration exterior. US454,96. C pale brown-pinkish with grey core. Abundant red grits; sparse vegetal temper. Red slip exterior and interior. Incised decoration exterior. US454,95. C light brown with grey core. Abundant red grits; vegetal temper. Red slip exterior and interior. Incised decoration exterior. US454,97. C brown. Red grits, sparse white grits. Red slip exterior and interior. US454,115.

5 . the iron age architecture of husn salut

-

269

plates

1

2 3

4 5

6

8

P late 73 – Area 8, sample from US454: deep and large basins (1-5); jar lids (6-8); pan (9). Scale 1:5.

7

9

270

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci

P late 74 – Area 8, sample from US454: grey ware; long handled bowls; camel figurine. Grey ware. F. Tiny white grits; rare vegetable temper. Incised decoration exterior. Probably same vessel as no. 2. US454,29. Grey ware. F. Tiny white grits; rare vegetable temper. Incised decoration exterior. Probably same vessel as no. 1. US454,28. Grey ware. Very F grey with brownish core. Rare tiny white grits. Wheel made. US454,175. M/C red with grey core. Medium and small white grits. Red slip. Incised herringbone decoration. US454,218. M red with grey core. Sparse small white grits. Red slip. Incised herringbone decoration. US454,219. M/C light brown-greyish. Sparse white grits, occasional large red grits. Light buff slip. Incised decoration. US454,217. M red with black core. Sparse white grits. Red slip. US454,137. M red with black core. Sparse white grits. Red slip. US454,138. M red. Sparse small white grits. Red slip. US454,9 = Cl163.

5 . the iron age architecture of husn salut

-

271

plates

1 2

3

4

5

7 6

9 8

P late 74 – Area 8, sample from US454: grey ware (1-3); long handled bowls (4-8); camel figurine (9). Scale 1:3, except nos 4-8: scale: 1:4.

272

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

chiara condoluci

P late 75 – Area 1: materials from phase HSI’s foundation deposits. M red. Black grits, sparse red grits; abundant vegetal temper. Pale brown-beige slip exterior and rim, brown interior. Black painted decoration on rim. US510,5. M brown. Abundant white grits, larger red grits, occasional grey grits; vegetal temper. Pale red-pinkish slip exterior and interior. Brown slip on rim with dark brown painted decoration. US510,1. F/M brown. Tiny white grits, grey grits, occasional red grits; vegetal temper. Brown-reddish slip exterior and interior. Dark brown painted decoration on rim. US510,2. M brown. Abundant red and white grits, occasional mica; vegetal temper. Red slip exterior and interior. Black painted decoration on rim. US510,4. M red. Small white and grey grits, rare large black grits; vegetal temper. Red slip exterior and interior. Black painted decoration on rim. US510,6. M brown. Abundant dark grey and red grits, sparse white grits; vegetal temper. Pale red slip exterior. US510,3. M red. Abundant white grits, some very large (>5mm), some exploded; occasional vegetal temper. Red slip exterior, smoothed interior. Black painted decoration on rim. US510,7. M brown, overfired. Small white grits, several exploded; vegetal temper. Red slip exterior and interior. Dark red painted decoration exterior. US337,2. M/F red, hard fired. White, red and black grits. Dark grey slip exterior. US337,1. F brown. Small white grits, occasional red grits; vegetal temper. Light brown, smoothed slip exterior and interior. Rotation marks on interior. US337,3. Leaf-shaped arrowhead. US337,4 = MB202 (see Chapter 7, no.67). Bronze shaft-hole axe. US337, 70 = MB196 (see Chapter 7, no.83). C brown. White and grey grits, large red grits; abundant vegetal temper. Light brown slip exterior and upper interior. US336,1. F/M pale red. Sparse small white and red grits; vegetal temper. Dark brown slip exterior, red interior of spout. US336,2. Soft-stone lid US335,1 = S193 (see Chapter 6, no. 43). M brown. Small red grits, occasional white grits; abundant vegetal temper. Red slip exterior, pale red-pinkish interior. US332,1. M/F brown-dark greyish. Tiny white grits, grey grits, occasional larger red grits; vegetal temper. Pale buff-yellowish slip exterior, light brown interior. US332,2. Soft-stone lid US327,1 = S192 (see Chapter 6, no. 40).

5 . the iron age architecture of husn salut

1

-

273

plates

3

2

4

6

5

9

8 7

10

11 12

13

14

15

16

P late 75 – Area 1: materials from phase HSI’s foundation deposits. Scale 1:3.

17

18

274

1. 2. 3. 4. 5. 6. 7. 8. 9.

chiara condoluci

P late 76 – Area 1, Burnt Building, room 5b: materials from US74/75 (phase HSII). F red. White grits, sand temper. Black painted decoration exterior and interior. US74,1. M brown. White grits, sand temper; vegetal temper. Brown slip exterior and interior. Graffito decoration on exterior, slight burnished exterior. US74,2. M dark brown. Sparse red and white grits, occasional large white grits. Red-brownish, blackened slip exterior and interior. US75,1. F light brown. Sparse vegetal temper. Whitish slip exterior and interior. US75,6. F brown. Orange slip exterior and interior – possibly Iron Age III. US75,5. F brown. Orange-brown slip exterior and interior. Shallow impressed decoration on neck – possibly Iron Age III. US75,4. F brown. Orange slip exterior and interior – possibly Iron Age III. US75,3. M cream. Small white grits; vegetal temper. Incised decoration exterior. US75,2. M/C orange. Sparse large red grits, occasional large grey grits. Red-brownish slip. Long handle bowl. US75,7 = Cl61.

5 . the iron age architecture of husn salut

-

275

plates

2

1

4

3

5

6

8

7

9

P late 76 – Area 1, Burnt Building, room 5b: materials from US74/75 (phase HSII). Scale 1:3.

276

1. 2. 3. 4. 5. 6.

chiara condoluci

P late 77 – Area 1, Burnt Building, room 5b: materials from US74/75 (phase HSII). Copper-base snake; copper-base ladles; copper-base miniature axe; drawings of the two baskets’s weft. Copper-base snake figurine. US74,3 = MB143 (see Chapter 7, no.8). Copper-base ladle. US75,8 = MB107 (see Chapter 7, no.11). Copper-base ladle. US75,9 = MB112 (see Chapter 7, no.12). Copper-base miniature axe. US75,10 = MB122 (see Chapter 7, no.17). Drawing of a date-palm basket fragment showing flat and thin stripes, woven using the twill plaiting ‘two up, two down’. US75. Drawing of a basket or mat fragment showing S-spun fibres grouped in four and plaited ‘two up, two down’. US75.

5 . the iron age architecture of husn salut

-

plates

277

1

3

2

5

4

6

P late 77 – Area 1, Burnt Building, room 5b: materials from US74/75 (phase HSII). Copper-base snake (1); copper-base ladles (2-3); copper-base miniature axe (4); drawings of the two baskets’s weft (5-6). Scale 1:3, except nos 4-6: scale 1:1.

6. Stone vessels enrica tagliamonte

The stratigraphic sequence revealed at Husn Salut stands like a sort of epitome of the prolonged occupation of the area through the centuries, also reflected in the finds. About one hundred chlorite vessels were found during the excavations, in addition to a few specimens made of other stones. The vessels presented here were selected according to two different criteria: the first was the context of provenance, that is, vessels coming from undoubtedly pre-Islamic contexts (as indicated by the stratigraphic relations) have been studied. However, most vessels come from disturbed, secondary contexts; therefore, exceptions were made for special and unusual pots or fragments, and for those vessels that, on a typological ground, could undoubtedly be identified as Iron Age productions despite their provenance. The second criterion was the exclusion of too small and undecorated fragments, for which a chronological and typological subdivision is not possible.

B ronze A ge

stone vessels

The occupation of the hill of Husn Salut during the Early Bronze Age, as evidenced by the tombs revealed on the top of the site,1 is mirrored by the discovery of eight fragments of chlorite vessels that, on the basis of the color of the stone, the shapes, and the decorations, can be dated to a period that covers the entire Bronze Age. During the Umm an-Nar period, three subgroups can be distinguished within a new style of vessels that has been defined as série récente by S. De Miroschedji:2 engraved vessels, undecorated vessels, and jars of Iranian origin. Three open vessel fragments were found on the site, that are indicative of the different typologies characteristic of this period. Two of them (nos 1-2) are datable by stylistic and morphological comparisons with many contemporaneous sites such as Hili:3 they are rather common shapes characterized by one or more incised horizontal lines just below the moulded rim. These are usually the only decoration on vessels that would otherwise be plain. This detail is typical of Umm an-Nar vessels; similar specimens were for example discovered also at Tell Abraq, in a mixed context.4 Other examples of bowls with moulded rim and without decoration come from Tomb A at Hili North and Tomb N at Hili Garden,5 although with very different diameters and profiles, and they have been attributed to local production.6 Frag-



Condoluci and Degli Esposti 2015: 6-12; see pl. 1. De Miroschedji 1973: 9-80. 3 David 2002a: 178-181. 4 Potts 1991: 34, fig. 22. 5 David 2002a: 178, fig. 7/1-3, and 181, fig. 12. 6 Ibid.: 175-178. 1 2

280

enrica tagliamonte

ment no.1 shows a peculiar detail: the bottom of the fragment seems to shrink and suggests the development of a flattened base. Hemispherical bowls, another typical Umm an-Nar shape, are attested by one single, undecorated fragment, no. 3. All the types of bowls found at Husn Salut not only have close parallels with the findings from Hili but are commonly found in the Umm an-Nar tombs spread all over the Oman peninsula, where soft-stone vessels are frequently attested, as exemplified at Abraq.7 The diffusion of these vessels reaches more distant sites in the Gulf area, as in notably the case for the Bahrain cemeteries; their presence as far afield as Susa and Ur in southern Mesopotamia further shows how their appreciation went beyond cultural borders.8 Fragment no. 4 deserves a particular mention. Made of red chlorite, it bears an excised zig-zag decoration on the exterior. Although it certainly belongs in the Early Bronze Age, this type of decoration is typical of murgabo-bactrian productions during the Bronze Age.9 The discovery of vessels of this type illustrate the international trade relations that developed at this early stage and in the first half of the second millennium BC, at the same time of the construction of the tombs on the hill of Husn Salut and in its surroundings. Objects with excised decoration have been found throughout the Near East and of course the discovery of one single fragment cannot indicate direct contact with central Asia, but at the same time it provides evidence for the inclusion of Salut in long-distance commercial routes, even if only as a terminal of a regional, South East Arabian re-distribution of foreign goods. The intrinsic value of the object, reckoned as a luxury good, is undoubtedly underlined by the fact that it found its way into Iron Age contexts: it was common practice that precious goods were kept for a long time and repaired when needed. Other finds can be dated to the final phase of the Bronze Age, into the Wadi Suq/Late Bronze Age periods. They testify to the continuity of occupation of the plain, as confirmed by the investigation of a few burials located on the hill facing Husn Salut from the northeast and overlooking the plain.10 The fragments from Husn Salut exhibit the main features of the period, such as is the case for biconical or globular vessels without lateral handles bearing a complex decoration comprising a series of metopes and strips with different motifs – usually oblique lines, triangles and rows of dot-in-circle motifs. It is probably thanks to the contacts with South East Iran that Oman developed a different production of stone vessels after the Umm an-Nar period. While in the latter period in fact manufacture focused on plain and simply decorated bowls, the introduction of new ideas and objects through trade led to a more varied production in the following Wadi Suq period. This is also evident in the fragments found at Husn Salut: the typological and decorative differences are clear and show a greater mastery of the production technique. However, this development is linked to an increase in local production and not to massive imports from outside areas, the opposite rather being true, that is, Omani products are in this period still found scattered over a wide area of the Near East. Among the four fragments belonging to the last part of the Bronze Age, no. 7 is difficult to identify, both for its limited size and for its type of workmanship. The dark color of the stone and the polishing derived from lathe-turning would in fact suggest a later chronology, but comparisons with third millennium finds from Mesopotamia11 allow us to hypothesize a Bronze Age dating. The remaining fragments belong to conical or globular vessels of Wadi Suq/Late Bronze Age date. No. 5, despite the small size of the fragment, is identified as a globular vessel bearing a dot-in-circle decoration enclosed between two horizontal lines. The rest of the decoration seems to be formed by groups of oblique lines. This type of vessel finds comparisons, although not very close, with objects from Shimal funerary contexts.12 The only example of a lid dated to this phase (no. 8) has a typical decoration of the period, comprising a double dot-in-circle motif enriched by a series of oblique lines running along its edge.



Potts 2000: 55. David 2002a: 175-178; Reade and Searight 2001. 9 During Caspers 1994; 1996; Potts 2003: figs 1-2; Degli Esposti 2015d. 10 Condoluci and Degli Esposti 2015: 13-37. 11 C. Phillips, pers. comm. 12 Velde 2003: 107, fig. 5. 7 8

6 . stone vessels

I ron A ge

281

stone vessels

Most of the Iron Age vessels have been found in mixed contexts or superficial layers. However, the picture that emerges is that of a considerable presence that reflects the main period of occupation of the site. Morphology-wise, five main types can be distinguished: 1. bowls with flattened, rounded or pointed rims. The base, when present, is flat. These can be subdivided into - small (diameter between 7 and 10 cm). In rare cases with a horizontal spout - large (diameter above 10 cm and with very deep body) 2. conical vessels with convex base; 3. beakers; 4. small compartmented boxes; 5. lids, that can be subdivided into two categories: - circular lids - rectangular lids – in turn subdivided into, singles and doubles. This mirrors the typologies established at other sites, both residential such as Rumeila13 and funerary, such as a number of sites located in Wadi al-Qawr, like Nasla, Fashga and Waʿab.14

Bowls This group of open vessels can be divided into two categories which differ principally over the diameter. The smaller fragments have a diameter between 7 and 10 cm, while the larger ones can reach a diameter of 22 cm. Their decorations are varied but comprise the main motifs that already P. Lombard identified in his analysis of the stone vessels from Rumeila.15 Specifically, they are characterized by the occurrence of two different motifs: rows of dot-in-circles, and patterns composed of two or more oblique lines forming a triangle motif. Only on bowls can a gadroons motif decoration be found. This consists of a series of vertical engravings overlaid by small arches connecting the upper ends of two adjacent vertical lines, giving a final appearance that resembles petals. This type of decoration, which is common on Iron Age vessels, seems to be borrowed from bowls manufactured in different materials, such as gold and silver, originating in the Mesopotamian area.16 Since stone vessels are probably to be interpreted as elite products, though less precious than metals, it is possible to assume that the shapes and decorations of the vessels themselves simply represent a cheaper imitation (skeuopmorph) of metallurgical products. All of Husn Salut’s bowls are characterized, in their lower portion, by vertical engravings that follow their contour. While no. 9 is decorated with simple lines, the pattern becomes more complex on the other two vessels: nos 10-11. On the former, the motif is topped by a saw-teeth engraving; on no. 11, it includes gadroons and the saw-teeth pattern is delimited by two parallel lines. The decoration on the surface of no. 11 is even more complex: in addition to the exterior wall, in fact, the rim is decorated with a herringbone pattern that matches part of the engraved decoration on the wall. Small bowls with a horizontal spout may be distinguished as a sub-group within small bowls. The half preserved and the complete vessel found at Husn Salut both bear a chevron decoration on the spout. While no. 17 has an alternate triangular decoration with central metope under the spout, no. 19 features a dot-in-circle decoration, and a sawteeth motif enclosed between oblique lines.



Boucharlat and Lombard 1985. C. Phillips, pers. comm. See also Phillips 1987. 15 Lombard 1982; 1985: 192, figs 304-314. 16 Potts 1990a: 380. 13 14

282

enrica tagliamonte

Fragments belonging to bigger bowls vary in shape and decoration; in particular, the shape and the rim vary, characterising vases with more or less wide openings and with different depths. All are however distinguished by very deep bodies. Due to the fragmentation of the finds, it is not possible to determine what kind of bases they had, as is the case for the remarkably tall no. 21. The latter also bears a very complex metopal decoration; the presence of a similarly decorated piece at Kalba suggests a possible dating to the Early Iron Age.17 Other bowls have simpler decoration or are plain like no. 22, except for three secondary graffiti on the exterior surface. Comparable bowls of varying sizes and with similar decorations have been found in large numbers at contemporaneous sites, such as Qarn Bint Saʿud, Nasla 1, Fashga 1 and Fashga 2,18 but also among funerary goods in burials I and II in Sharm19 and Grave AS100 of Asima,20 just to list a few examples. The shapes may be slightly different, with greater opening and depth, but this reflects the usual variety of the characteristic Iron Age types.

Conical vases This closed shape is typical of South East Arabian Iron Age production, distinguished by strongly in-curving walls and convex bases. Vessels of this type are characterized by a remarkable variety in both shape and decoration; they are engraved on the outer surface and may comprise different motifs. The diameter can vary significantly: the two fragments no. 25 and no. 26 belongs to medium-large vessels (with a maximum diameter between 20-30 cm), but not all the collected fragments allow a reconstruct of their size. No. 27 is the only complete conical vase found at Husn Salut. The decoration on the walls summarizes the main elements of the Iron Age, with oblique lines forming triangular patterns and saw-teeth motif. Also for conical vessels comparisons can be found throughout the whole of South East Arabia, both in funeral complexes such as Qarnt Bint Saʿud and al-Qusais,21 and from settlements such as Rumeila.22 Excavations of single graves for example at Dibba, Shimal, Fashga,23 Qidfaʿ,24 and also in the al-ʿAin oasis, attest the variance of this type in terms of size and decoration.

Beakers Only two fragments belonging to this type were discovered at Husn Salut, both with a very small diameter, less than 6 cm. They come with different characteristics: one, in fact, has an open shape with a flared profile (no. 30), while the other (no. 31) is of the cylindrical type. In both cases the decoration realized with engravings defines a clearly phytomorphic motif. Comparisons for this type can be found at Waʿab 4 and Fashga 2 in the Wadi al-Qawr.25 The use of the saw-teeth motif to trace lines that in turn define a figurative decoration also find parallels at other sites, such as Rumeila and Jabal al-Buhais, significantly comprising phytomorphic motifs as well.26 No. 31 deserves further description: in fact, it bears a saw-teeth phytomorphic decoration on the outside and shows symmetrically positioned holes, possibly used for the insertion of bronze plates completing the decoration.



Unpublished; C. Phillips, pers. comm. C. Phillips, pers. comm. See also Phillips 1987. 19 Ziolkowsky 2001: figs 30, 49, 50, 54, 55, 61. 20 Vogt 1994: fig. 42/8-14. 21 Zutterman 2004: 109, fig. 9. 22 Lombard 1985: figs 315-322. 23 Phillips 1987: figs 24/7-8; 26/13-14; 27/15; 29/22-23; 30/26. 24 Unpublished, on display in the Fujaira Museum. 25 Unpublished; C. Phillips, pers. comm. 26 Boucharlat and Lombard 1985: pl. 61/10; Jasim 2012: fig. 162/11. 17 18

6 . stone vessels

283

Compartmented boxes Two fragments of this kind of vessel were found: while one reflects the shape and decoration typical of the period (no. 33), the second (no. 34) has characteristics that differentiate it from those found in abundance at other sites in South East Arabia. Most in fact comprise rectangular-shaped boxes with converging walls; in the case of no. 34 at least one of the two compartments seem to be circular. Examples of this type are rare in the Omani Peninsula: a similar vessel, unfortunately out of context, is preserved in the Ras al-Khaima Museum.27 It comprises three compartments, of which one is circular; its decoration consists of horizontal lines and triangular motifs, the space formed by the triangles is filled with a saw-teeth pattern. Although out of context, it can stylistically be dated to the Iron Age. For what concerns decoration, no. 33 shows a dot-in-circle motif, bordered by incised parallel lines. No. 34 instead only bears faint lines belonging to a herring-bone decoration.

Circular lids Circular lids have a slightly convex or flat top surface; the lower surface is also slightly convex or flat. A recessed edge distinguishes the part expected to be inserted into the vessel mouth. Handles may vary considerably, from cylindrical to conical, and have a variable height. The upper circular knob also varies remarkably. The decorations on the upper surface of the lid comprise simple dot-in-circle motifs or saw-teeth motifs that in more complex cases are arranged to form a star-like design. Lids like nos. 36 and 38, where the latter motif is present, are typical of the Iron Age, and find, for example, parallels in lids from Rumeila and Sharm.28 The decoration of the upper surface of the knob also varies: almost all the lids however show a series of radial lines toward the center of the grip, completed by central dot-in-circle or further radial ‘subdivisions’. Lid no. 37 is characterized by a small hole pierced through the middle of its lower face; it may result from reshaping, perhaps to create weights or seals as documented in many cases in the near east and at Husn Salut itself. 29

Rectangular lids Rectangular lids have a slightly convex or flat top surface; the lower surface is slightly flattened. They usually have a rounded edge where insertion into the box was expected. The handles were not found or were heavily damaged. The decorations on the upper surface of the lids comprise simple dot-in-circles or needle-shaped decorations, or saw-teeth motifs and radial lines. No. 43 is a double rectangular lid that is to be associated with a large compartmented box, with a rather simple and coarser decoration than the contemporary examples. During the Iron Age, lids are strangely more numerous than in the previous phases. Most likely, besides the vagaries and circumstances of discoveries, a change in the manufacturing of these objects is a likely explanation. In fact, the lids are thicker and therefore more robust than the previous Bronze Age ones. There is also a much lower number of rectangular lids, single or double, created exclusively for pyxes and rectangular boxes.

27 Probably coming from site Waʿab 4 in Wadi al-Qawr and belonging to the Late Bronze Age or the first phases or the Early Iron Age (Huckle 2003). 28 Boucharlat and Lombard 1985: pl. 61/3-4; Ziolkowsky 2001: fig. 83 and figs 41, 69, 85, 88 for star-shaped decoration realized using different motifs. 29 Degli Esposti 2015c: fig. 17.

284

enrica tagliamonte

Barrel-shaped vessels The small fragment no. 44 would appear to be the only barrel-shaped vessel found at Husn Salut. Other than the shape, the presence of a probable phytomorphic motif made by saw-teeth engravings suggests this hypothesis; these features make it resemble one peculiar vessel found at Rumeila.30 Among the vessels usually reported from other coeval sites it is important to note that, apart from barrel-shaped ones, other shapes are completely absent at Husn Salut, such as biconical vases31 and medium-large beakers with flared walls.32 Despite these differences, which may be the result of randomness in the findings, the site’s assemblage fits almost perfectly the typologies established for South East Arabia as a whole. The distribution of Iron Age vessels, analyzed by C. Zutterman33 based on comparisons between the major sites of the period, underlines their substantial presence in ​​the peninsula of Musandam and in the area north of Qarn Bint Saʿud. Although these results were only partial, due to the limited number of analysed Omani sites, the latter have shown a homogeneous distribution and a consistent similarity of some shapes. In some cases, particular shapes have also been found in distant places such as Ur and Babylon. These data thus confirm the widespread presence of soft-stone vessels over the whole Oman peninsula, mirroring a​​ cultural koinè that is also expressed by the ceramic assemblages. The absence of evidence for a local production at Salut seems an indication that theses chlorite objects were imported; at the same time, the surveys made on the al-Hajjar range; the recent, exceptional discovery of a production site;34 and petrographic studies;35 all confirm a South East Arabian production in the Iron Age. Against this scenario, the location of Salut places it in a preferential point where major north-south and east-west routes converge; with more detailed study and additional fresh data at hand, it is likely that different areas of influence will be identified, shedding light on the mechanisms of intra-regional trade during this important period. Moreover, the discovery in Mesopotamia of some particular shapes of Omani production indicates a continuous use of commercial routes already opened during the third millennium. Unfortunately, the lack of published data did not allow comparisons to be made for the few limestone vessels from Husn Salut (nos 51, 52, 53, and 54), whose discovery in mixed or superficial contexts moreover prevents a definite dating of their production. However, some assumptions about the origin and history may be put forward for what regards the alabaster vessels from the site. One fragment, no. 55, belongs to a vessel which cannot be dated on the basis of its typology alone; the provenance from a dated context (US3) nevertheless provides a chronological indicator and allows us to establish the terminus ante quem for its last use. In this case the fragment, undecorated, has a beaker shape with slightly rounded base. Among the oldest examples of this form are vases from third millennium contexts at Shahr-i Sokhta and Mundigak,36 and alabaster vessels also are part of the famous ‘booty of Magan’ mentioned by Naram Sin during the 23rd century BC.37 One example of a beaker, made in chlorite but with a similar shape, is also known from the third millennium BC Tomb N at Hili, where it was interpreted as an import.38 Simple profile shapes however continued to be produced over the centuries. The material of which these vessels are made is completely absent from the Oman peninsula, the production therefore having to be placed in different areas like Iran-Pakistan or Yemen; moreover, alabaster vessels have been found over an even wider area, as shown by a jar with lid in the Barbar temple in Bahrain.39 The other alabaster fragment (no. 56) found in Area 1 at Husn Salut has a typical beehive shape, with flat bottom



Boucharlat and Lombard 1985: pl. 61/10. Boucharlat and Lombard 1985: pl. 60/5-7; Zutterman 2004: 109, fig. 10. 32 Zutterman 2004: 109, fig. 9; Lombard 1982. 33 Zutterman 2004. 34 Harrower et al. 2016. 35 David 2002b. 36 Ciarla 1981: 50, fig. 3/b. 37 E.g. Potts 1986. 38 David 2002a: fig. 14. 39 Potts 1986. 30 31

6 . stone vessels

285

and slightly inward flaring walls. Vessels with this shape, in chlorite, are dated to the second half of the third millennium from contexts in the BMAC (Murghabo-Bactrian Archaeological Complex), Mesopotamia and Iran.40 At the same time, the production of different kinds of shapes covers a large area spanning from India to the Egyptian territory.41 Given that these chronological and geographical ranges are broad for alabaster vessels, it is not possible to put forward other hypotheses for Husn Salut vessel no. 56 apart from a general Iron Age date strongly supported by the overall context of the site and by the available radiocarbon date for US3, that would place its manufacture before 800 BC (see table 18). Some similar examples, although equipped with four perforated loops, come from the sites of Mleiha, ed-Dur, Dhahran and Karrana.42 These finds are dated to a period comprised between the middle of the 3rd century BC (Mleiha) and the 1st century AD (ed-Dur). The pots are part of funerary goods associated with rich burials: in the Mleiha necropolis they were found together with objects of Greek origin, while in ed-Dur the assemblages included glass jars, bone plaques and metal objects such as gold beads, bronze weapons and vessels. In the Dhahran burials, the finely worked and polished alabaster jar is accompanied by hundreds of semiprecious stone beads and bronze ornaments. The presence of objects coming from far afield and definitely imported, therefore valuable, is the strongest index of commercial networks developed between the Mediterranean and the Near East. Finds from other regions, specifically in the eastern Mediterranean area, clearly indicate that these types of alabaster vessels were used as containers of value-added goods: aromatic oils, perfumes and medical products in the form of powder used in antiquity to protect the eyes or face. However, the vessel itself went on to assume the status of a valuable object: the fragility of the material in fact makes it unsuitable for daily use, and its exoticism surely also played an important role. As mentioned above, stone vessels can be interpreted as cheaper, more easily available skeuomorphs of metal vessels. At the same time, they were reputed as uncommon objects themselves, as shown by the fact that during the Iron Age they had their own skeuomorphs, realized in a typical grey ware, often incised, which usually replicates their shape and decoration.43 Moreover, the presence of repairs or attempts thereof, already attested during the Bronze Age, seems to imply that, although locally available and worked, soft stone was considerably valued. At Husn Salut, fragment no. 75 is completed by a small metal addition, possibly the remnant of a repair. Fragment no. 77 was drilled twice at the base-wall junction: even in this case the two holes bear witness to ancient repair, probably comprising ropes or metal elements44 fixed through them. More interesting is the discovery of the small spouted cup no. 17 from US298. Comparisons for this shape are rather widespread and are found from Maisar to Rumeila.45 This particular shape is inspired by typical Iron Age metal productions, represented at Husn Salut itself by the small cup MB214 (see Chapter 7, no. 22). This vessel is also a clear example of how valuable such an object was for the populations of the time: two small holes, one on the base and one on the wall, are indicative of an initial attempt to repair the vessel. Moreover, it was chosen for being included in US298, a deposit for which a ceremonial nature can be rather safely supposed, together with numerous carinated ceramic cups (see Chapter 4). While it was possible to determine the types of most of the vessels under study, it was not possible to identify specific characteristics – morphological and/or decorative – that would allow recognizing differences between earlier (i.e. Iron Age I + II) and later (i.e. Iron Age III) specimens, with the exception discussed in the next paragraph. In such a situation, contexts of discovery would be useful to establish at least a terminus ante quem for the production of certain vessels. Unfortunately, as said above, most of the Iron Age stone vessels found at Husn Salut do not come from primary contexts but from mixed or superficial contexts that may just confirm their continued use.

40 During caspers 1996: pl. 11. 41 One of the largest known quarries is located in the Negev area. The site shows traces of extraction as early as the vi-v millennium. Yemen was also a major manufacturing region for alabaster vessels. Extraction sites have been identified in three different areas: Wadi Baihan, Wadi Dura and Wadi Markha. According to Hassel (1997) it is conceivable that, given the homogeneity of production and typology between the 3rd century BC and the 1st century AD, some of these vessels could originate from an internal Arab trade. Mleiha’s corpus also poses another problem concerning the production by local artisans that imitated forms of Southarabian tradition (Hassel 1997: 276-277). 42 Hassel 1997: 249-255, figs 4-5. 43 See Chapter 2 for a brief discussion of the few grey ware sherds discovered at Husn Salut. 44 Like the clamps of which a few specimens have been discovered at the site, see Chapter 7. 45 E.g. Yule 2001: pl. 17, Grab M2703; Lombard 1985: figs 307-310.

286

L athe - turned stone vessels : post -E arly I ron A ge occupation

enrica tagliamonte

or early lathe introduction ?

It is commonly held that at the end of the 3rd century BC the lithic industry undergoes a complete transformation with the introduction of the lathe. The production of stone vases using this technology had already been experienced in Mesopotamia and in the Oman peninsula itself during the 3rd millennium, but soon abandoned.46 The use of the lathe completely modified traditional manufacture: from the 3rd century BC closed shapes suddenly disappear. At the moment, it is still difficult to propose a complete typological classification due to insufficient excavation data. Even in this case fragments from Husn Salut match the typical production of South East Arabia, characterized by a limited number of lathe-turned shapes: - small vessels; - pedestalled cups; - large basins. The predominance of open forms is clear, perhaps due to a new functional value attributed to the objects. The decoration is now obtained during lathe turning and consists of studs or grooves. For the small cups, the very thin walls are decorated with small incised lines on the outer surface, as well visible on fragment no. 57, which finds fitting parallels also from Sharm, Dibba and Mleiha.47 Other features, in addition to shape and decoration, make the vessels of this period easily recognizable: contour symmetry, stone polishing, and lathe marks left during processing both on the inner and outer surfaces. Cup no. 58, for example, shows evident rotation traces on the exterior; similarities can be found in specimens from ed-Dur.48 Basins (or large bowls) are now well represented, usually large and with deep carinations just below the rim. The color of the stone, between dark green and black, the internal lathe traces and the highly polished exterior clearly date them after the Iron Age. Probably these vessels did not end with a flat base, but rather with a convex one, just like a few specimens on display in the Sharja Museum and coming from the sites of Dibba and Mleiha, in contexts dated between 250 and 200 BC. As mentioned, the type of stone used is different from previous phases and varies from black to bluish, with rare very dark green artifacts; stone density also differs from previous periods. This might reflect the search for new areas of stone quarrying. The mentioned fragment no. 57 is somehow an exception: while bearing a typical post-Iron Age decoration, it is made in a light-colored stone as the artifacts of the previous period. The work still reflects, as far as it was allowed by technology, shapes and styles that characterize contemporary products, that is, stone vessels resembles objects in different materials such as wood, glass and metal. Despite the use of the lathe having the possibility of speeding up production, archaeological evidence does not record a real increase in their quantity. Sites with comparable materials dated immediately after the commonly accepted end of the (early) Iron Age (c. 300 BC) are quite scarce, this apparently mirroring a contraction of settlement.49 At Husn Salut, the situation after this date still needs investigating; in all likelihood, the site was at least occasionally visited, possibly within a broader purpose that sees the continuation of contacts and commerce along older traditional routes, as might be suggested for example by the recovery of a small number of these late stone vessels also in Saudi Arabia.50 However, some fragments found during the latest campaigns at the site have partially changed the picture, as they appear to be connected with the late phase of the site’s occupation, i.e., the HSIII phase, distinguished by the substantial re-structuring of at least one part of the massive outer wall (see Chapter 2).

46 Zarins 1978: 67, pls 74-75. During the study of the scattered steatite industries on Tarut island, some of the vessels revealed to be smoothed and finished on a lathe. 47 Ziolkowski 2001. 48 Zutterman 2003: 81-87, fig. 2. 49 E.g. Benoist 2013a: 16. 50 Zutterman 2003: 83.

6 . stone vessels

-

catalogue

287

These recently found fragments of stone vessels belong to small size open cup. In most cases the walls are thin, except for base no. 74, which is thicker. In all instances, lathe-marks are clearly visible. One fragment, no. 61, belongs to a small, open vessel, which shows a decoration just below the rim, comprising four parallel horizontal lines. Fragment no. 62, instead, has a rather peculiar open shape, characterized by a flattened rim with a very pronounced hull that resembles an Iron Age III ceramic tulip bowl. In both cases surfaces are polished while tool marks are left on the interior. This presence of lathe-turned stone vessels at Husn Salut clearly raises new questions about the site, and specifically about its end date, further discussed in Chapter 10. Given the commonly accepted reconstruction regarding stone vessel production, these vessels should in fact indicate that the site, or at least part of it, was still occupied after 300 BC c. At the same time, other materials that could be dated to that period, namely pottery, have not been so clearly identified: this could conversely imply a possible earlier introduction of the lathe than commonly thought, or also mean that the long conservatism evident in the pottery assemblage carries on beyond 300 BC, thus generating a scenario in which a few new elements stand out against an unchanged background. Indeed, some first possibly post-Iron Age III sherds have been highlighted during a review of the pottery from later contexts carried out to prepare this work (see Chapters 2 and 10), and this line of research will surely be among the privileged ones during the continuation of the works at Qaryat Salut, the substantial settlement associated with Husn Salut whose investigation started in late 2015.

C atalogue For each entry, the inventory number is followed by the indication of the context of provenance, dimensions, and employed material, and by a short description. Dimensions given in cm: l.= length; w.= width; h.= height; d.= diameter; th.=thickness. N/A = not attested. B ronze A ge 1. Bowl pl. 78 S173, US1 w. 15; h. 3.5; d. 12; light grey soft stone Period: Early Bronze Age Type: strongly carinated open vessel Decoration: N/A. 2. Bowl pl. 78 S168, US198 w. 5.8; h. 4.7; th. 0.7; d. 21; light grey soft stone Period: Early Bronze Age Type: open bowl with outward sloping walls and rounded rim Decoration: just below the rim thee deeply incised horizontal grooves. Other two are visible just on top of the break. The natural colour of the stone is clearly light grey; it fades to brown-reddish due to smoothing of the external surface.

3. Bowl pl. 78 S180, US257 h. 5; th. 0.9; d. 14; light grey soft stone Period: Early Bronze Age Type: open vessel with hemispherical base Decoration: N/A 4. Bowl pl. 78 S187, US257 w. 4.5; h. 2; th. 0.5; d. 6; red soft stone with light grey inclusions Period: Bronze Age (Umm an-Nar) Type: hemispheric bowl Decoration: it is possible to recognize a low and barely visible keel. The deep excised decoration is made by two different lines: the lower one is composed of a continuous excised chevron which runs horizontal; the upper one is very badly preserved and partially visible near the fracture. Probably reproducing the same zig-zag design of the central area.

5. Beehive vessel pl. 78 S94, US40 w. 4.2; h. 3.9; th. 1.5/0.7; dark grey soft stone Period: Late Bronze Age Type: strongly carinated closed vessel Decoration: three different incised rows: the first one is formed by oblique lines crossing a series of horizontal lines; the second one is composed of a row of dotin-circle. The last clearly preserved one is formed by oblique, lines running in opposite directions and separated from the other rows by a horizontal line. 6. Open vessel? pl. 78 S167, US175 th. 0.7; d. 11; grey soft stone Period: Wadi Suq/Late Bronze Age Type: unrecognizable, with rounded base Decoration: on the lower half of the vessel, two parallel horizontal lines enclosing a band containing specular sets

288

enrica tagliamonte

of diagonal parallel lines forming trapezoidal empty fields. 7. Open vessel? pl. 78 S325, US416 w. 4.0; h. 2.6; th. 0.3; dark grey-bluish soft stone Period: Bronze Age Type: open vessel Decoration: the small portion still visible of the decoration shows a band made by four horizontal deep incised lines. Interior and exterior surfaces are highly polished and show evident lathe marks. 8. Lid pl. 78 S32, US3 h. 3.4; d. 6.7; d. base 5.7; d. knob 2; dark grey soft stone Period: Wadi Suq Type: round vessel lid with a flat base and knob-shaped handle, top not preserved Decoration: two incised rows: oblique short lines running along the edge; on the inner part, just near the handle, a stripe of double dot-in-circle.

I ron A ge 9. Bowl pl. 78 S322, M39 w. 6.7; h. 3.5; th. 1.8; light grey soft stone Period: Iron Age Type: open bowl Decoration: the upper half is decorated with a band of three parallel diagonal lines zig-zagging across the upper part of the vessel. This motif is followed by a horizontal line which separates it from the lower half of the vessel incised with a band of vertical lines. 10. Bowl pl. 78 S207, US326 h. 3.3; th. 0.5; d. 8; light grey soft stone Period: Iron Age

Type: open bowl with a flat and rim Decoration: the exterior surface is incised in two stripes: the first one, below the rim, consists of four parallel horizontal lines containing a band of a herringbone decoration; the second one is formed by a saw-teeth motif followed by a series of vertical lines. 11. Bowl pl. 78 S329, US441 w. 5.2; h. 4.3; th. 0.7/1; d. 10; greygreenish soft stone Period: Iron Age Type: open bowl with a deep basin and round slightly everted rim Decoration: the entire external surface is incised with a different band of decoration. Just below the rim a band of herringbone design enclosed between two horizontal parallel lines, followed by a horizontal row of saw-teeth and a single horizontal line. The last band is formed by gadroons. Herringbone pattern on rim. 12. Bowl pl. 78 S380, US665 w. 3/4; h. 4.2; th. 0.4; d. 12; light grey soft stone Period: Iron Age Type: open bowl with a deep basin and slightly rounded rim Decoration: just below the rim, two horizontal parallel lines enclosing a herringbone motif. The inferior part of the vessel is formed by a ‘gadroons’ motif, with a vertical petal-like decoration. Each petal is separated by a vertical line. Clear chisel marks on interior. 13. Bowl pl. 78 S197, US1 w. 2.5; h. 2.8; th. 0.5; d. 7.3; light grey soft stone Period: Iron Age Type: small open bowl with flat base and rounded rim Decoration: on the exterior face, it presents two parallel horizontal lines just

below the rim, followed by a band of a zig-zagging motif composed of two lines. A horizontal line above the base complete the decoration. 14. Bowl not illustrated S358, US594 th. 0.5/0.6; d. 7; light grey soft stone Period: Iron Age Type: small open bowl with flat base Decoration: it is decorated with four different bands, just below the rim a saw-teeth motif delimited by two horizontal parallel lines; a dot-in-circle row and another saw-teeth motif delimited by two horizontal parallel lines. The last band is composed by triangles filled by diagonal lines. 15. Bowl pl. 78 S337, US1 h. 4.5; th. 0.6; interior d. 9; light grey soft stone, with some white band Period: Iron Age Type: open bowl with a deep basin and flat slightly everted rim Decoration: the decoration is only partially visible; the first motif is composed of three parallel horizontal lines followed by a band of two parallel diagonal lines across part of the vessel. Three parallel diagonal lines zig-zagging across the rim. 16. Bowl not illustrated 358, US594 th. 0.5/0.6; d. 7; light grey soft stone Period: Iron Age Type: small open bowl with flat base Decoration: it is decorated by four different bands, just below the rim a sawteeth motif delimited by two horizontal parallel lines; a dot-in-circle row and another saw-teeth motif delimited by two horizontal parallel lines. The last band is composed of triangles filled by diagonal lines. 17. Bowl S306, US298

pl. 78

6 . stone vessels

h. 3.9; th. 0.7; d. rim 8.5; d. base 5; grey-greenish soft stone Period: Iron Age Type: open bowl with a narrow Ushaped spout Decoration: Three horizontal lines below the rim are followed by a band of zig-zag motif composed of several lines. Other two horizontal parallel lines are visible near the base. Below the spout two vertical lines containing a herringbone decoration, which is also present on the exterior surface of the spout. Two holes (d. 0.5), on the base and on the wall, reveal an attempt to repair through the use of metal or perishable material. 18. Bowl not illustrated S357, US576 w. 3.2; h. 5.5; dark-reddish soft stone with large amount of mica inclusion Period: Iron Age Type: spouted bowl Decoration: from the top, the decoration consists of a horizontal stripe of triangle filled by diagonal lines crossing a band made by seven horizontal lines. A similar decoration is visible also near the fracture and consists of a triangle filled by horizontal lines. The two bands enclosing a dot-in-circle row. 19. Bowl pl. 78 S390, M246 h. 4.1/4.7; th. 0.5; d. rim 7.5; d. base 6.4; light grey soft stone Period: Iron Age Type: open bowl with a narrow Ushaped spout Decoration: two horizontal lines below the rim and near the base enclosing a saw-teeth motif. The central area is incised with a very confusing and complex design, two parallel lines zigzagging enclosing a band of saw-teeth. The triangular space formed by this decoration is filled with radiating triangular composition, in some cases with a

-

catalogue

vertical central line. The saw-teeth design near the rim is followed by a horizontal row of dot-in-circle motif. Other two horizontal parallel lines are visible near the base. Below the spout two vertical lines frame a criss-cross decoration, the exterior of the spout is decorated with four deep incised lines in parallel bands. 20. Bowl not illustrated S359, US1 Period: Iron Age w. 2.1; h. 1.7; th. 0.4; dark grey-bluish soft stone Type: spouted bowl Decoration: very thick incised decoration: under the spout three horizontal grooves. On the wall the decoration is composed of a series of three oblique incised lines, alternating with triangular fields, filled by a single dot and circle. 21. Open vessel pl. 78 S13, US9 w. 6.8; h. 13; th. 1; d. 16; light grey soft stone Period: Iron Age Type: open vessel with outward flaring walls; everted rounded rim Decoration: very complex incised decoration formed by a metopal pattern. From the top, the first decoration consists of 13 horizontal lines; the following stripe is made of thin vertical rectangles, filled by criss-cross motifs, alternated with thicker vertical rectangles filled by 5 lines of dot-in-circle motif. The third stripe shows a series of oblique lines, forming triangular shaped fields, which are filled by a single dot-in-circle. Another small stripe is filled by a series of dot-in-circles. The last clearly preserved one is formed by oblique, lines running in opposite directions and intersected by horizontal lines. One small portion of the following stripe shows a specular decoration of the last one.

289

22. Open vessel pl. 78 S69, US1 h. 5; th. 1.5; d. 18; dark grey soft stone Period: Iron Age Type: open vessel with flat rim Decoration: three ‘graffiti’ slightly incised: the first one is an irregular pentagon, the second one is butterfly shaped (formed by two specular triangles divided by a vertical line), lastly a rhombus filled by an oblique cross. 23. Bowl pl. 78 S315, US1 w. 5.5; h. 5.3; th. 0.7; d. 22.5; light grey soft stone Period: Iron Age Type: open vessel with a pointed rim and a wall thickened at mid-height Decoration: two incised horizontal lines run just below the rim, followed further down by a couple of saw-teeth shaped or zig-zagging lines. At the bottom of the fragment, near the fracture, two lines made by a similar decoration are visible. 24. Bowl pl. 78 S334, US470 w. 2.7; h. 2; th. 0.6; grey-greenish soft stone Period: probably Iron Age due the decoration Type: not recognizable Decoration: the decoration visible on the fragment is composed of two dotin-circles followed by three horizontal parallel lines. 25. Conical bowl pl. 79 S318, M39 w. 11.4; h. 7.5; th. 1.9; grey soft stone Period: Iron Age Type: closed shaped vessel with a inward sloping walls Decoration: from the top, a dot-in-circle decoration enclosed by two patterns made by two saw-teeth decorations running between two horizontal lines. This pattern is followed by another formed

290

enrica tagliamonte

by crossing couple of lines, each one enclosing a saw-teeth motif. Rough interior with traces of tool marks on interior.

three incised parallel horizontal lines running along the base and some diagonal lines on the body. Roughly worked interior.

26. Conical bowl pl. 79 S309, US373 Period: Iron Age th. 1.7; th. base 1.2; d. max 19; light grey soft stone Type: closed shaped vessel with a thinned rounded base and inward sloping walls Decoration: band of radiating triangular composition with a vertical line. Three deeply incised parallel horizontal lines above the base. Original grey colour of the stone visible in the fractures; surface yellow-pinkish likely due to burial conditions.

29. Conical bowl pl. 79 S184, US287 w. 5.3; h. 5.2; th. 0.7; grey soft stone Period: Iron Age Type unrecognizable; probably a closed vessel like a beehive vessel or conical vessel strongly carinated at the base Decoration: just below the rim: two groups of horizontal and parallel lines, separated from the each other by a rough saw-teeth motif. The lower part of the fragment is decorated with a triangular composition made by three lines. Those are characterized by a narrow saw-teeth motif. Tool marks exterior and interior.

27. Conical bowl pl. 79 S312, US390 h. 7.2; th. 0.5; d. rim 5.9; d. base 12.3; light grey soft stone Period: Iron Age Type: closed vessel with slightly rounded base and inward sloping sides, strongly carinated Decoration: just below the rim two horizontal lines enclosing a saw-teeth decoration. The same motif recurrs on the following stripes made by a series of radiating triangular composition; the space between the triangles is filled by the saw-teeth motif. A horizontal line, in correspondence of the carination, completes the decoration. Several pointed gouges on the interior, near the base, oriented towards the base. Similar marks at the wall/base junction, oriented sidelong. 28. Conical bowl pl. 79 S385, US696 w. 6; l. 10.4; h. 2.5; th. 1/1.5; light grey soft stone Period: Iron Age Type: conical bowl with rounded base Decoration: on the exterior surface

30. Beaker pl. 79 S330, US1 w. 3.1; h. 3.9; th. 0.3; grey-greenish soft stone Period: Iron Age Type: open vessel with hyperbolic profile, probably a beaker Decoration: Four deep incised horizontal parallel lines. The following motif is only partially visible, it seems to be a triangular composition realized by a pattern of three parallel zig-zagging lines composed of really tight saw-teeth motif. Oblique lines due to working visible on exterior; some striations in various directions visible on interior. 31. Beaker pl. 79 S75, US3 h. 5.3; th. 0.5/0.7; d. 6.5; dark grey soft stone Period: Iron Age Type: cylindrical shaped vessel with flat rim and rounded base, thinner walls towards the rim, slightly concave base Decoration: complex incised decoration: four lanceolate leaves, formed by

a saw-teeth motif, are symmetrically placed on the external surface. Also, some details such as the veins are also reproduced with the same motif. Three deep incised lines are present at the bottom of the vessel. Four small ovoid cavities used for inlays, probably copperbase plaques, are also visible alternating with the leaves. 32. Beaker pl. 79 S388, US706 w. 2.9; l. 6.6; h. 3.9; th. 0.6; grey soft stone Period: Iron Age Type: probably cylindrical shaped Decoration: the external surface bears a decoration consisting of two deep incised horizontal lines which delimit the nondecorated inferior part from the upper one. Above the two horizontal lines a zigzag made by two lines with a saw-teeth pattern. A small remains of a diagonal line, probably enclosing the motif, is visible on the top part of the fragment. Chisel marks on interior. 33. Compartmented box pl. 79 S326, US440 w. 4.8; h. 5.5; l. 5.8; th. 0.3; light grey soft stone Period: Iron Age Type: compartmented rectangular shaped vessel with a rounded base and slightly inward tapering sides Decoration: A band of two dot-in-circle horizontal lines enclosed between four horizontal lines, just below the rim, and followed by other three horizontal ones. Smoothed exterior, chisel marks on interior. One of the compartments is almost complete, the other one is only partially preserved. 34. Compartmented box pl. 79 S86, Surface w. 5.3; l. 6.1; th. wall 0.8; compartment d. 4; light grey soft stone

6 . stone vessels

Period: Iron Age Type: compartmented vessel, with a flat rectangular base with a shrinkage in the central area, which gives a slightly butterfly shape to the vessel. The compartments are circular Decoration: only on one surface is the decoration clearly visible: a series of oblique lines and two horizontal lines at the bottom of the vessel. 35. Compartmented box pl. 79 S366, US601 w. 4.8; l. 5.2; h. 2.4; th. 0.7; dark grey/ black soft stone Period: Iron Age Type: squared or rectangular vessel, probably a compartmented box, with slightly rounded base Decoration: on the inferior part of the walls a series of four horizontal lines roughly incised are visible. Base underside smooth, upperside with tool marks. 36. Lid pl. 79 S190, Surface h. 3.8; th. 1.1; d. 6.9; d. handle 2.7; light grey soft stone Period: Iron Age Type: round vessel lid with knob-shaped handle with a rounded top tapering inward toward the base Decoration: upper lid surface decorated with a dot-in-circle motif in an uneven circular band partially visible due to the fracture of the lid. This motif enclosed a circular band of saw-teeth decoration. 37. Lid pl. 79 S149, Surface h. 4.5; d. 8.0; dark grey soft stone Period: Iron Age Type: round vessel lid with flat base and small knob shaped handle with a rounded top Decoration: rounded top surface decorated with a single dot-in-circle and radiant lines. Along the edge of the upper surface a row of dot-in-circle is incised.

-

291

catalogue

The lower surface of the lid is characterized by a small and deep drill hole. 38. Lid pl. 80 S331, US478 th. 1.5; d. 13; grey-greenish soft stone Period: Iron Age Type: round vessel lid with a flat base Decoration: A circular band of sawteeth motif all along the edge of the rim. The external decoration encloses a star design, only partially visible on the fragment, composed of saw-teeth motif. 39. Lid pl. 80 S191, US315 h. 4.4; th. 1; d. 7.3; d. handle 5.2; light grey soft stone Period: Iron Age Type: round vessel lid with a flat base and squat knob-shaped handle with a flat top Decoration: the upper lid surface is incised with a saw-teeth circular motif along the edge of the lid enclosing a seven-pointed star design composed of lines of saw-teeth motif. Knob top decorated by four intersecting lines with short dashes on the edges. 40. Lid pl. 80 S192, US327 w. 4.8; h. 4.7; th. 1.1; d. 5.9; d. handle 3.1; light grey soft stone Period: Iron Age Type: round vessel lid with a flat base and squat knob-shaped handle with a flat top Decoration: the upper lid surface, badly preserved, has a roughly executed circular saw-teeth motif. Knob top decorated by short dashes along the edges. 41. Lid pl. 80 S361, US5 h. 2.6; th. 1; d. 2.9; dark grey soft stone Period: Iron Age Type: round vessel lid with rounded base and small knob-shaped handle with a rounded top

Decoration: N/A On the underside is a circular hole (d. 2 cm). Traces of fine scratch marks on the surface of the lid. 42. Lid pl. 80 S133, US89 w. 5.8; l. 6.3; h. 2.2; th. 1.1; light grey soft stone Period: Iron Age Type: squared lid with a broken rounded knob Decoration: along the two shortest sides two incised parallel lines filled by a saw-teeth motif. From the 4 corners three parallel lines are incised, forming a cross. The decoration is completed by four incised dot-in-circle motifs. Rough underside. 43. Lid pl. 80 S193, US335 w. 10.6; th. 3; dark grey soft stone Period: Iron Age Type: rectangular vessel lid with division Decoration: two parallel incised lines along the edges, forming a inner rectangle, divided into five panels by four parallel lines, perpendicular to the external ones. On the right side there are three parallel lines, most probably due to a mistake of the engraver. The five rectangles have a decoration formed by saw-teeth motif. 44. Barrel-shaped vessel pl. 80 S346, US493 w. 3.2; h. 3.04; th. 1.1; light grey soft stone Period: Iron Age Type: closed vessel, probably barrel shaped Decoration: Seven deep incised horizontal lines, just below the rim, followed by a non-recognizable design realized by a saw-teeth motif. Possibly a phytomorphic decoration. 45. Wall fragment S341, US473

pl. 80

292

h. 2.5; th. 0.8; grey-greenish soft stone Period: Iron Age Type: probably open bowl Decoration: only few traces of the original decoration survive. Two faint horizontal parallel lines. 46. Wall fragment not illustrated S372, US1 th. 0.5/1; light grey soft stone Period: Iron Age Type: unrecognizable Decoration: incised criss cross decoration. 47. Wall fragment not illustrated S404, US1 w. 8.6; l. 12; th. 1.3/1.6; soft stone Period: Iron Age Type: probably part of a closed vessel (beehive shaped) Decoration: it is possible to recognize one horizontal line and some diagonal lines near the fracture of the fragment. The inner surface is smoothed. 48. Wall fragment pl. 80 S364, US611 w. 4.7; h. 3.7; th. 1.5; light grey soft stone with whitish inclusions Period: Iron Age Type: unrecognizable Decoration: a series of horizontal parallel lines and diagonal crossing lines deeply incised. Deep pointed working grooves on interior. 49. Base fragment pl. 80 S360, US5 w. 6.1; th. 0.8; light grey soft stone Period: Iron Age Type: unrecognizable Decoration: just above the base, a series of diagonal lines followed by a horizontal line which delimits an incised sawteeth motif. 50. Base fragment pl. 80 S350, US570 w. 7.3; h. 2.5; light grey soft stone with

enrica tagliamonte

white quartz inclusions Period: Iron Age Type: unrecognizable Decoration: Band of seven horizontal lines crossed by a series of diagonal zigzagging lines. Traces of tool marks on the upperside.

U ncertain

date

51. Base fragment pl. 80 S335, US470 w. 12; h. 4; d. 12,5; probably grey soft stone Period: uncertain Type: footed bowl. The wall of the vessel is flaring with right angle carination at the foot junction; the foot is composed of a short, in-curving neck over a slightly convex base Decoration: faint traces of incised decoration consisting of some straight lines and a couple of concentric circles. Clear circular tool marks on the foot’s neck and interior. Lathe turning left a polished effect and a brown-reddish colour on the stone. 52. Limestone fragment pl. 80 S319, M138 h. 6.8; th. 2.9; d. 10.5; beige-pinkish limestone Period: uncertain Type: probably a mortar Decoration: N/A Foot of vessel slightly convex at the bottom. The walls are flared with an inclination which doesn’t allow for a precise identification of the shape and function. The thickness and the inclination of the walls suggest it could have been a mortar. The inner surface is very smoothed, but not polished and shows circular hollows probably due to the use of a pestle. 53. Limestone fragment pl. 80 S324, US420 w. 6.1; h. 3; th. 1.7; d. 14.4; white/beige limestone

Period: uncertain, probably post Iron Age Type: uncertain; the small fragment shows a shallow and flat bottom, just 0.9 cm below the rim. Original shape possibly similar to a patera Decoration: N/A Lathe marks on exterior. 54. Limestone fragment pl. 80 S328, US1 h. 4.5; th. 0.8/1; d. base 7; beige limestone with abundant inclusions of quartz Period: uncertain Type: vessel with a deep basin, the base is rounded Decoration: a deep horizontal groove on the link between the base and the wall. 55. Beaker pl. 80 S88, US3 h. 3.6; th. 0.5; d. base 4.1; d. rim 2.9; yellowish alabaster Period: Wadi Suq? Type: cylindrical shaped vessel with pointed rim slightly inward walls, slightly concave base Decoration: N/A 56. Beehive vessel pl. 80 S20, US1 h. 7.3; th. 1.3; d. rim 4.2; d. base 8; whitish alabaster Period: uncertain Type: closed vessel with flat thick base and strongly inward sloping sides. Rounded and thin rim. It belongs to the category of ‘beehive vessel’ Decoration: N/A Working scratches on interior.

L ate I ron A ge 57. Bowl pl. 80 S179, US236 w. 4.3; h. 7.5, th. 0.5; d. 9; light grey soft stone Period: Late Iron Age

6 . stone vessels

Type: small bowl with convex base and quite straight walls Decoration: just below the rim of the vessel a series of parallel and horizontal incisions. The second one is composed by several grooves and it is located on the maximum diameter. Circular lathe marks exterior and interior. 58. Bowl pl. 80 S339, M174 h. 1.6; th. 0.3; d. 6.2; dark grey soft stone Period: Late Iron Age Type: small open bowl with pointed rim Decoration: the exterior surface is entirely worked with horizontal lines. The inner surface is well smoothed, with clear lathe marks. 59. Base pl. 80 S382, US695 w. 2.9; l. 5.8; th. 0.5; dark grey-bluish soft stone Period: Late Iron Age Type: unrecognizable, thin and flat base Decoration: at the beginning of the wall two horizontal parallel lines are visible. The inner surface is really polished with fine scratch at the wall/base junction. A hole in the center of the underside attests an attempt of reworking. 60. Wall fragment not illustrated S370, US647 Period: Late Iron Age w. 2.5; h. 5.7; th. 1; dark grey soft stone Type: unrecognizable, probably a bowl Decoration: two horizontal parallel lines are followed by a series of three horizontal parallel lines just above the bottom of the fragment. Polished interior. 61. Open vessel pl. 80 S384, US696 w. 7; h. 5.6; th. 0.4/0.7; grey-bluish soft stone Period: Late Iron Age/post Iron Age

-

293

catalogue

Type: open vessel with a pointed rim Decoration: four deeply incised horizontal parallel lines, located just below the rim. One more horizontal line is partially visible above the base. Chisel and lathe marks interior. 62. Open vessel pl. 81 S378, US691 Period: Late Iron Age h. 5.8; th. 1; dark grey-bluish soft stone Type: small bowl with a deep basin, strong carination Decoration: N/A Lathe marks exterior and interior. 63. Open vessel pl. 81 S391, US664 Period: Late Iron Age w. 4.9; h. 3.2; th. 0.6/0.7; dark grey-bluish soft stone Type: open vessel with a flat rim Decoration: a raised band (th. 0.5) separates the rim from the body of the wall. A small hole is drilled just below the rim. 64. Basin pl. 81 S161, US163 th. 1.6; d. 23; dark grey soft stone Period: Late Iron Age Type: strongly carinated open vessel with a flat rim Decoration: N/A Circular lathe marks exterior. 65. Basin pl. 81 S343, Surface th. 1.1; d. 9; dark grey soft stone Period: Late Iron Age Type: open vessel with carinated profile Decoration:N/A 66. Basin pl. 81 S363, US1 th. 1.3; d. 16; light grey-greenish soft stone Period: Late Iron Age Type: bowl with a deep basin, the rim is flat. It presents a deep carination on the

body of the sherd Decoration: N/A The surface is smoothed, a small hole (0.4 diam) is drilled near the carination most probably to repair the bowl. 67. Basin pl. 81 S379, US691 w. 3.8; h. 4.7; th. 1.5; d. 18; dark greybluish soft stone Period: Late Iron Age Type: open bowl with deep basin and flat rim Decoration: N/A Deep carination along the body. Lathe marks exterior. 68. Basin not illustrated S393, US717 w. 5.4; h. 5.2; th. 0.9/1.5; dark grey-bluish soft stone Period: Late Iron Age Type: open bowl with flat rim and deep basin; light carination along the body Decoration: N/A Just above the carination the bronze clamp used to fix the broken vessel is visible. 69. Basin pl. 81 S395, US741 w. 5.5; h. 5.3; th. 1.6; d. 41; grey soft stone with abundant of mica inclusions Period: Late Iron Age Type: open bowl with a pointed rim, light carination along the body Decoration: N/A

P ost I ron A ge or uncertain late date

70. Open vessel pl. 81 S396, US743 w. 8.2; h. 5.1; th. 1.2; d. 16.2; dark greybluish soft stone Period: post Iron Age? Type: open bowl with a thin and pointed rim. The wall becomes thicker towards the base Decoration: N/A

294

Slightly polished interior. Probably a fragment of a reworked broken vessel. 71. Lid pl. 81 S308, US1 th. 1.8; d. 13; dark grey soft stone Period: post Iron Age Type: round lid with flat base and flat top Decoration: N/A 72. Lid pl. 81 S196, US1 th. 2.8; d. 6.8; dark grey soft stone Period: post Iron Age Type: unrecognizable. Possibly a lid or a vessel base Decoration: N/A On the upperside a very deep and triangular groove; probably caused by the reuse of the vessel as a stopper. 73. Wall fragment not illustrated S394, US748 w. 11.7; h. 5.3; th. 1.4/2.3; light greybluish soft stone Period: post Iron Age Type: unrecognizable Decoration: N/A Rough working marks interior, thin chisel marks exterior. 74. Base not illustrated S386, US664 w. 4.5; l. 7.9; th. 1.2; dark grey soft stone Period: uncertain Type: unrecognizable Decoration: N/A Lathe marks on upperside. 75. Wall fragment not illustrated S85, Surface w. 3.5; h. 5.2; th. 1; dark grey soft stone (and bronze/copper) Period: uncertain Type: unrecognizable Decoration: N/A A rectangular bronze staple is attached to the exterior surface (th. 1.6), probably used for repairing a fracture.

enrica tagliamonte

76. Conical bowl pl. 81 S313, US1 Period: uncertain w. 6.1; h. 4.5, th. 1.2; dark grey soft stone Type: bowl, probably conical Decoration: two horizontal incised lines mark the bend between the wall and the bottom. 77. Base fragment pl. 81 S153, US161 w. 7.5; h. 3; light grey soft stone Period: uncertain Type: unrecognizable Decoration: N/A Small hole (d. 0.4) drilled into the lower part of the wall, another portion of hole is visible along the fracture, just few mm from the first one. Probably it hosted a bronze staple or a different perishable item used to repair the vessel. 78. Wall fragment not illustrated S403, US1 Period: uncertain w. 3; l. 2.5, th. 1.8; light grey soft stone Type: unrecognizable Decoration: two horizontal parallel lines just below the rim Possibly part of a reworked vessel.

plates

296

enrica tagliamonte

1 9

12

2

13 10

3 15 4 11

5 22

6

23

7

8

24

19

17

21

P late 78 – Soft-stone vessels: Bronze Age specimens (1-8) and Iron Age bowls. Scale 1:3.

6 . stone vessels

-

297

plates

25

33

26

27

28

34

35

29 30

36

31

P late 79 – Soft-stone vessels: Iron Age specimens. Scale 1:3.

32

37

298

enrica tagliamonte

38 39

41

40

42

48

45

44

49

43

50

53

52

55

51

56

54

57

58

59

61

P late 80 – Soft-stone vessels: Iron Age lids and various fragments; limestone specimens (52-54); alabaster vessels (55-56); post Iron Age specimens (57-61). Scale 1:3.

6 . stone vessels

-

299

plates

65

62

63 66

64

67

69

70

76 71 72 77

P late 81 – Soft-stone vessels: late Iron Age/post Iron Age specimens. Scale 1:3.

7. Metal objects marzia sasso

Metallic items were found at Husn Salut in the most different contexts, the majority of which were disturbed and impossible to date. Therefore, a selection was necessary before undertaking their cataloguing. This involved firstly an evaluation of the stratigraphic context, and all materials coming from safe Iron Age layers have been included in the catalogue. Secondly, objects belonging to well known Iron Age typologies have been included also when their context was not reliable.1 The discussion of the different classes of objects follows their order in the catalogue. Apart from the discussed objects, Iron Age levels revealed numerous undecipherable copper-base fragments (scrap metal), mostly parts of clamps and rods. Further iron fragments were on the other hand discovered in late, mixed contexts.

C eremonial

or cultic objects

As discussed elsewhere in this volume (Chapter 9), despite the fact that the multiple aspects characterising the fortified site of Husn Salut somehow complicate its fully clear interpretation, its monumentality, as wells as the nature of a large part of its material culture, strongly indicate some sort of cultic or, however, ceremonial destination.

Copper-base snakes Objects that can be confidently connected to a cultic, ritual significance are copper-base snakes, inasmuch as a series of excavations conducted first at al-Qusais and much more recently at Bithna, Saruq al-Hadid, Masafi and Salut itself have provided convincing evidence about the existence of a snake cult in Iron Age South East Arabia.2 At all of these sites, snake representations are found both on different metal artefacts3 and, much more frequently, on pottery, where they can be applied in relief, incised, or painted. The number of metal snakes discovered at Husn Salut appears finally to be relatively limited when compared with

1 However, a few objects from uncertain contexts, for which an Iron Age date is not so well attested by existing parallels, have been equally described in the catalogue, since lacking was, also, any element that could certainly distinguish them as post Iron Age. This was done also in view of the awaited publication of the thousands of metal items recently discovered at al-Safa near ʿIbri and Saruq al-Hadid in the desert of the Emirate of Dubai. 2 See among others Benoist 2007; Benoist et al. 2015. 3 See for example Nashef 2010: figs 15, 16 (Saruq al-Hadid) and Benoist et al. 2012a: fig. 13 (Masafi 3).

302

marzia sasso

recent findings from Masafi4 or with the abundant, so far unpublished evidence from Mudhmar, near Adam in central Oman (thus relatively close to Salut).5 Their length generally ranges between 16 and 23 cm, only one specimen being remarkably smaller and measuring around 5 cm (no. 9). This is in accordance not only with the figures from Saruq al-Hadid, al-Qusais6 and Bithna, but also with the dimensional range of the specimens discovered at Masafi 3.7 Snake figurines, when sufficiently well preserved, can be divided into two categories based on their production technique. The majority are realized by casting, contrary to what is witnessed at Masafi,8 while the remaining ones were obtained by hammering a copper-alloy narrow strip or small rod. All snakes are depicted crawling, as indicated by an undulating body;9 in two cases, nos 8-9, the head is slightly raised in an alert position. Heads vary in shape and they are mainly rhomboidal, with single examples of triangular, trilobate, and ogive-shaped head. Flat hammered specimens have an even section and are undecorated, as is mostly the case for those from Masafi10 and the ‘Mound of Serpent’ at al-Qusais.11 One slightly different example is no. 8, the partially rounded body of which was obtained by folding a metal strip, a procedure that finds parallels, again, at Masafi.12 Cast snakes have bodies with an oval section, flattened or rounded. Some bear small applied globules on the body and head, while singular occurrences were recorded of scattered pointed incisions and incised small circles respectively, the latter type of decoration being also found on ceramic examples. Applied globules are interpretable, in the case of Husn Salut, as decorative elements sketchily indicating the snake’s scales (pl. 82/3-5).13 In the unique case of a snake with trilobate head (no. 5) it was possible to identify two horns, an element that is also found on pottery examples and that is commonly interpreted as the will to represent the horned viper (Cerastes gasperettii gasperettii), still part of the region’s fauna14 (see Chapter 2). Despite the dangerousness of these snakes, they never appear to be represented in a menacing way, rather their mouth resembles a smile that provides them a somehow anthropomorphous aspect. Moreover, this seems to indicate a positive connotation of these animals among the local people, as possibly witnessed by the representation of a hidden human head just below a snake head in one handle from Masafi.15 Indeed, the snake cult that developed in South East Arabia during the Iron Age16 can be associated with a considerable number of different aspects that are interrelated in many ways: fertility, underground water (thence, the aflaj), healing, magic and technological know-how; specifically metallurgy. At Saruq al-Hadid, the discovery of two large storage jars bearing an applied representation that has also been interpreted as portraying the snake hunt using nets17 raises the question about a possible ‘ritual hunt’ meant to supply snakes for ceremonies to be held in cultic places, where the snakes could have also been further bred.18 Moreover, a possible connection between snake symbolism and funerary rituals – hinting at a possible cult to the

E.g. Benoist et al. 2012a: 156. Gernez 2016; Jean, Pellegrino and Gernez in press. 6 Nashef 2010: 220, 223; see also Weeks et al. 2017: fig. 19. 7 Benoist et al. 2012a: 156. 8 Ibid. 9 Where only the head survives, crawling is suggested by the bending of the head-body joint. 10 Benoist et al. 2012a: 156. 11 Taha 1982-83: fig. 16; Taha 2009: pl. 53. 12 Benoist et al. 2015: 33. 13 The hypothesis of a function for fixing the snake to something else, put forward for some of the Masafi examples (Benoist et al. 2012a: 156), has to be rejected in the case of Husn Salut. 14 The hypotheses has been put forward that the depicted snakes, and specifically those bearing incised dots, could represent the saw-scaled viper of the venomous species Echis carinatus, with triangular head and narrow neck. Similar representations are known from Tarut (cf. Nashef 2010: 220). 15 Benoist et al. 2012a: 158 and fig. 17. 16 Occasional snake representations are however known also during the Early Bronze Age (Umm an-Nar period). 17 See Weeks et al. 2017: fig. 19. 18 Nashef 2010: 220.

4 5

7 . metal objects

303

ancestors – has also been suggested on the basis of findings from the necropolis of Adam north and from two graves at Jabal Mudhmar and Jabal Handali, where snake representations were found in association with long handle bowls and braziers.19 While a discussion of this complex issue in this catalogue would be redundant, given the number of recent papers dedicated to this subject,20 it is anyway worth underlining the very likely connection between the snake cult and metal working, apparently mirrored in the deposition of raw materials, smelting by-products and scrap metal in cultic contexts characterized by the presence of numerous snake representations, as for example at Bithna and al-Qusais.21 Besides, a reverence for metal per se is witnessed by the deposition of non-functional items in collective and ceremonial contexts, such as at Masafi 1.22 The tight metal-snake connection is also evident at Saruq al-Hadid.23 A rather straightforward explanation can be found in the subterranean origin of the ore, that would correlate with the chthonic nature of the snake.

Ladles Within a socio-economic context in which water certainly played an essential role, further enhanced by the previously suggested connection with the snake symbolism, its use during ceremonies has to be considered highly probable. Libations or liquid ablutions, either involving water or other liquids, are generally suggested by the presence of lined pits, sometimes walled with stones;24 by the particular association of spouted jars and small cups commonly attested at sites and buildings for which a ceremonial interpretation has been proposed; and by a few, rare metal ladles that could as well be defined as ‘ritual’ on the basis of their contexts of provenance. At Husn Salut two copper-base ladles (nos 11-12) were discovered in US75, a context that has been interpreted as a foundation deposit given the rich and characteristic material collected (see also Chapter 5): woven baskets containing dates, one of the copper-base snakes (no. 8), a miniature axe (no. 17), an iron handle (no. 147) and abundant, partially burnt, animal bones, possibly to be connected to a ritual banquet. The two ladles differs in dimension and production technique. The cup is almost identical, while the handle length is different, possibly indicating that they were meant to be used with differently shaped – and sized – containers. The longest ladle, that also bears upon its handle two incised crossing lines from where a vertical line departs, was cast, while the other one was folded and hammered. The interpretation of the mentioned engravings remains hypothetical: they could indicate the ownership of the object, as well as the manufacturer, or the use of a specific alloy. Ladles from Iron Age contexts are known from Muweila, one coming from the floor of Building II’s columned hall,25 and from Saruq al-Hadid.26 At a later date, they are attested at Mleiha, in funerary and settlement contexts of the 3rd-2nd century BC and of the 1st-2nd century AD, and in Samad period burials, for example at Sumail.27 Examples with incised handles are known from Mleiha, Sumail and ed-Dur, dated to the 1st-2nd century AD,28 from al-Baruni, where the suggested dating is in the first century BC, and from a Samad grave dated to the 3rd-4th century AD.29 Apart

19 Benoist et al. 2015: footnote 36. 20 E.g. Benoist 2007; Potts 2007; Benoist, Skorupska and Pillault 2007; Nashef 2010; Mouton, Benoist and Cordoba 2011; Benoist et al. 2015. 21 Benoist 2007: 38; Taha 1982-83. 22 Benoist et al. 2012b; Goy et al. 2013: 139. 23 E.g. Nashef 2010. 24 At Bithna and Masafi these were concentrated in areas close to collective buildings (columned buildings); see Benoist et al. 2015: 22. 25 Magee 2003a: 186; see also Magee 1998b. 26 Herrmann, Casana and Qandil 2012: 61. 27 Yule and Weisgerber 1988: 33, 40; Mouton 1992: 55, 109-110 and figs 24/4; 88/4. At Mleiha these tools are rather widespread during the first and second centuries AD (cf. Mouton 1992: 55). 28 Mouton 1992: 110 and fig. 88/4; Yule 2016: 66 and fig. 31/7. 29 Yule and Kazenvadel 1993: figs 8/31; 9/36.

304

marzia sasso

from a significant difference in the incised motif however, they also belong to a typology with the handle ending in an animal’s head,30 as is the case for one specimen reported from Rustag.31 The appearance of ladles in the Near East during the Iron Age has been interpreted as mirroring the adoption of banqueting practices,32 and drinking sets discovered in at least two tombs dated to the 1st and 2nd century AD from edDur may be connected to funerary banquets.33 At Husn Salut another ladle (no. 15) was found in a context that is not strictly identified as ritual, but for which some doubt exists, as it stood close to an area rich in burnt matter. The typology of this ladle (cast) is the same as the other two from the site but with a smaller cup, only around 3 cm in diameter. The reduced dimensions seems to indicate the necessity of pouring small quantities of whichever product; this may be linked to a domestic use as a capacity measure for products to be used in small quantities such as spices or other particularly valued edible goods. Another object with similar dimensions to this above mentioned ladle, no. 14, could be interpreted as a spoon, given the flat junction between handle and cup; however, its state of preservation is rather poor and the ladle interpretation cannot be rejected.

Miniature axes Miniaturist objects at Husn Salut are exclusively represented by four, possibly five axes, most likely to be connected to rituals. One (no. 17) comes in fact from the foundation deposit US75 mentioned above, connected to the re-structuring of the Burnt Building in HSII phase, while the other two (nos 16, 18) come from rather disturbed contexts for which a similar interpretation cannot be completely discarded (see Chapter 5). A non-utilitarian nature is anyhow indicated by their dimensions (ranging between c. 2.5 and 5 cm), their manufacture, and in particular by their blade morphology, generally too thin to allow for actual cutting. Miniature axes from Husn Salut have a trapezoidal blade realized by hammering a single copper-base sheet, with two examples (nos 18, 20) showing a slight concavity of the longer (upper and lower) sides. Where preserved, the tubular shaft is obtained by folding the opposite end of the same metal sheet that forms the blade. A characteristic piece from Husn Salut is no. 19, interpreted as a miniature axe. Indeed, the shape is remindful of the well known fenestrated axes of the Levant, but it would constitute a unicum for South East Arabia. It must however be noted that it does not have a tubular socket and, if it actually represents an axe, fixing it to its shaft would have been similar to the way of tanged blades and arrowheads. Axes interpreted as representative, symbolic artecfacts but not of miniaturist dimensions were discovered at Rumeila, Saruq al-Hadid, and in the ʿIbri/Selme hoard,34 as well as from 1st millennium BC funerary contexts at al-Qusais, Qarn Bint Saʿud and Hili 8.35 All belong to the same typology with a very thin junction between shaft and blade and remarkable flaring of the latter, reckoned as traditionally local.36 An axe with the same morphology is significantly portrayed on an Iron Age seal from Rumeila,37 and on a richly decorated long handle bowl handle from Husn Salut itself (pl. 33/3).

30 These tips in the shape of an animal’s head are frequent from the Achaemenid to the Sasanian period (cf. Mouton 1992: 110). At al-Baruni the handle tip is not preserved. 31 Yule and Weisgerber 1988: 33 and fig. 8/4; Yule and Kazeenvadel 1993: fig. 9/35. 32 Moorey 1980. 33 See for example Haerinck 1994, for tomb AV/G.5156 (1st century AD). Here, given the association with strainers and cauldron with possible filtering spout, the ladles have been interpreted as part of possible wine sets, where the several impurities in wine would have required filtration (cf. Mouton 1992: 109-110). 34 Boucharlat and Lombard 1985; Herrmann, Casana and Qandil 2012: 61; Weeks et al. 2017: fig. 3/SF30085; Yule and Weisgerber 2001: fig. 9. 35 Salles 1981: 44; Lombard 1984: fig. 1/2; Cleuziou 1978-79: fig. 40. 36 Boucharlat and Lombard 1985: 61. 37 Lombard 1998: fig. 1/6.

7 . metal objects

305

However, the axe from Hili is characterised by a relevant tin content, something that would be more consistent with a utilitarian destination rather than a ceremonial one, given the technological advantages provided by the addition of tin to copper. This seems also confirmed by the preliminary results of archaeometric analyses on objects from Husn Salut and Masafi, which showed a very limited use of tin-bronze only for utilitarian items, while ceremonial ones were realized in unalloyed copper.38 The limited dataset currently available however induces caution on this issue. At Mudhmar East, near Adam, an outstanding collection of replica weapons has been recently found inside a cultic building dated to the Iron Age, as well as along the adjacent jabal slope.39 Among these stood five unfinished, cast ‘battle’ axes that have a more compact, trapezoidal blade and a tubular shaft, distinguished by horizontal ribs. Close to the context of provenance of these axes, pottery censers and copper-alloy snakes were discovered, which immediately led to an interpretation consistent with the link between the snake and metallurgy mentioned above.40 At the same time, a symbolic value connected to social prestige and possibly warfare cannot be excluded.

Copper-base vessels Among the possibly ceremonial objects from Husn Salut can be listed a copper-alloy cauldron, no. 21,41 discovered inside the possible oven SF21 (see Chapter 2). It has a hemispheric shape, obtained by hammering a single metal sheet, and two applied loop handles. No closely fitting parallel was found among the contemporaneous South East Arabian materials,42 while some morphologically similar examples are known in Iron Age tombs in Israel.43 Indeed, the provenance of this cauldron and the presence of burning traces could suggest a utilitarian use; on the other hand, it could have also been used for preparations connected to rituals carried out in the uppermost part of the site. A complete spouted cup, no. 22, and the fragment of a rim likely belonging to a similar object (no. 23) were also discovered; the cup resembles examples of pottery and soft stone. The complete cup comes from US568, a deposit that has been interpreted as a secondary ritual fill meant to level the raised platform SF49, also located in the uppermost part of the site (see Chapter 5), although this interpretation must clearly remain a hypothesis. Copper-alloy spouted cups were discovered at Saruq al-Hadid and in the necropolis of Jabal al-Buhais,44 although in the latter case the morphology does not perfectly fit the specimen from Husn Salut.45 The most strict comparisons are instead possible with some of the open bowls from the ʿIbri/Selme hoard.46 A characteristic object is no. 24, interpreted as the umbo-shaped lid of a miniaturist vessel, possibly used for incense or cosmetics and found on US23, the floor of room 2c in Building 1. While the room in itself seems not to have had any peculiar function, its proximity to the pillared room of the building may suggest the storing of items that could be used there. No fitting parallel has been recognized in the literature.

38 Degli Esposti, Renzi and Rehren 2016; Goy et al. 2013. 39 Guernez 2016. 40 Ibid. 41 Based on the the results from most recent stratigraphic study (see Chapter 2) it cannot be safely dated to the Iron Age III period as previously stated (Avanzini and Philliips 2010: 106), but its date rather ranges between the HSII and HSIII phases, that is, Iron Age II or Iron Age III as usually intended. 42 Hemispherical vessels with riveted loop handles are known from funerary contexts in the United Arab Emirates (for example Bidya and Jabal al-Buhais), but are anyhow of different dimensions and shapes than the cauldron from Husn Salut; their chronology is also different, as they date to the second millennium BC (al-Tikriti 1989: pl. 71; Jasim 2012: 27 and fig. 31). 43 Gershuny 1985: 12-14, pl. 8/99, 100. 44 Herrmann, Casana and Qandil 2012: 52; Jasim 2012: figs 24/4; 162/10. 45 This is also the case with some specimens from Hili 8 and Jabal al-Buhais, where the typology is completely different (Cleuziou 1989: pl. 34; JASIM 2012, figs 272-273). 46 Yule and Weisgerber 2001: pl. 22.

306

marzia sasso

Oddities Finally, an object that may be considered as extraordinary, no. 25, was found on the floor of the pillared room. It comprises a hemispheric base from which two antenna-like rods depart, forming a V shape. A tripod interpretation is to be excluded, as there is no trace of a broken third leg; the two rods were therefore pointing upward. Given this aspect, the interpretation proposed is the point of a standard. Also in this case, no parallels are known.

W eapons

and tools

Arrowheads The most common weapons at Husn Salut are arrowheads with 49 specimens, most of which come from late reworked contexts, although there are occurrences from deep, sealed and well dated contexts, as is notably the case for the one from US337 (no. 35) and the one from US454 (no. 49) that are among the contexts interpreted as intentional structural fills and for which a ceremonial nature has been postulated (see Chapter 5). It has to be noted that arrowhead no. 49 was in all likelihood intentionally reshaped and bent in a secondary moment, and this may have been done in order to obtain a sketchy snake representation. In no case was a deliberate grouping of these artefacts identified. All the arrowheads from Husn Salut are tanged, while no example of a trilobate-socket arrowhead (also known as ‘Scythian Points’) has been discovered, the latter usually dated to the Iron Age III or post Iron Age period and often attested together with tanged specimens47 at several sites, such as for example at Rumeila, Dibba (=Daba) and ed-Dur.48 Based on the blade morphology, three types of arrowhead can be distinguished: leaf-shaped (25 specimens), lanceolate (16), and triangular (11).49 For the lanceolate type a subdivision is possible in two variants: one with a clearly distinguished, perpendicular shoulder and an angular blade-shoulder junction, and the other without a well-defined shoulder, which is also more numerous. A similar picture, with minor differences, is attested in the necropolis of Jabal al-Buhais, where arrowheads are numerous.50 Other variables are the tang section, preferentially rectangular or anyhow angular rather than rounded, and the blade transversal section, either with or without the central rib, the two variants almost equally distributed (27 against 25). A more evident rib seems to be reserved for triangular arrowheads, while the lanceolate ones are more commonly without rib. Dimensions range between 3 and 7 cm (again similarly to Jabal al-Buhais), with rare longer specimens that may reach 11.0 cm in the case of no. 65.51 The latter moreover presents particularly concave blade edges, resembling a typology known from Rumeila.52 It is currently impossible to chronologically differentiate the different types of tanged arrowheads, as they often are attested together and alongside trilobate-socket specimens as well in the later context, as mentioned above.53 It

47 While the trilobate-socket Type seems rather clearly meant for military use, more debated is the use of tanged arrowheads in contexts where both Types are present. P. Delrue (Delrue 2007) for example suggested they could be used for hunting. 48 Boucharlat and Lombard 1985: pl. 62; Jasim and Yusif 2014: fig. 43; Delrue 2007: figs 3-4. 49 Four more specimens were too fragmentary to recognize the original shape. 50 Jasim 2012. 51 Although this measure seems remarkable, objects of such size have been classified as arrowheads both by P. Lombard (Lombard 1985: 208 and fig. 105/365) for an example from Ghalila, and by S. Jasim (2012: figs 39/3; 121/2) for other two examples from Jabal al-Buhais, which are also typologically close to nos 62 and 65. 52 Boucharlat and Lombard 1985: pl. 62/6. 53 In fact, although for example the typology elaborated by M. Mouton (Mouton 1990) for the iron arrowhead from Mlehia and ed-Dur only includes specimens lacking the midrib, examples where it is present are discussed by P. Delrue (Delrue 2007), at least for what concerns edDur. Numerous are the indications for the contemporary use of different Types of arrowheads, suffice it to mention the examples of the Sharm

7 . metal objects

307

would appear, therefore, that these main typologies are distinguished by a long continuity of use, as is the case for numerous, relatively simple but highly functional objects that, once they have reached the definitive morphology, remain almost unchanged.54 Parallels for the arrowheads found at Husn Salut are widespread and concern contexts of different nature: funerary (again, the necropolis of Jabal al-Buhais but also the necropolis of Samad as-Shan or the Sharm tomb, the famous ‘Warrior grave’ at Nizwa and some graves at Maisar),55 cultic (Masafi 3, al-Qusais),56 ceremonial or linked to elite gatherings (Masafi 1, Muweila),57 but also residential (Husn Awhala, Rumeila, Hili 2, Maisar, Qarn Bint Saʿud).58 This largely incomplete list is just in itself indicative of the abundance of weapons in Iron Age South East Arabia (mainly arrowheads, daggers and swords); their remarkably high frequency in funerary contexts calls into question the relevance of warfare in the area at that time. One interpretation sees these weapons as an indication of a highly conflicted society, with a focus on military developments; the presence of weapons inside the graves would therefore indicate the prestige of the dead, reflected in the possibility of having available metal objects that would have otherwise been re-used, when not re-melted.59 In an alternative view, this metal in funerary contexts would be conferred a similar ritual significance but in an opposite peaceful context: when the actual need for weapons ceases, they can be offered in burials by virtue of the metal’s worth.60 Archaeometric analyses of nine arrowheads from Husn Salut revealed a clear prevalence of copper specimens, with a single bronze one.61 Six arrowheads had either a folded blade or tang, a fact that is of difficult explanation. With the exception of no. 49, none of these arrowheads comes from possibly ceremonial contexts; they could represent discarded material stored for re-melting.62 Apart from these cases, a number of arrowheads were reshaped in antiquity, usually by longitudinally cutting off half of the blade, which could hint at a secondary use as razors or the like.

Spearheads Four spearheads – one of which fragmentary – were also discovered at Husn Salut. They all have a triangular blade with flat section, with a sharp joint between the shoulder and the tang, consistently with the fact that socketed spearheads appear to be remarkably rare in the Oman peninsula after the Middle Bronze Age, at least until the Sasanian period.63 The specimens from Husn Salut range in length between 16 and 21 cm; one (no. 77) has an angular point, more rounded in the other two complete examples.64 Their weight is rather low, between 13 and 32 g, and blades are extremely thin, ranging between 0.1 and 0.25 cm, parameters that are not consistent with a utilitarian use – sig-

tomb, where the remarkable number of grouped arrowheads suggests the presence of a decomposed quiver (Weeks 2000: 186-187), and of Rumeila (Boucharlat and Lombard 1985: 60 and fig. 62). 54 Copper-alloy arrowheads seem to appear in the Oman peninsula during the Late Bronze Age, becoming widespread during the Iron Age. 55 Jasim 2012; Yule and Weisgerber 1988; Weeks 2000; Yule 2015: pls 2/5-7; 3; Weisgerber 1981: figs 60, 63, 84-85. 56 Benoist et al. 2012a; lombard 1984; Taha 1982-83. 57 Benoist et al. 2012a; Magee et al. 2002; Magee 2003a. 58 Potts et al. 1996: fig. 33; Boucharlat and Lombard 1985; Lombard 1985: 206-208; Tillmann 1981: 224, fig. 60; Lombard 1984: 230 fig. 3. 59 Jasim 2012: 294. 60 Potts 1998: 203-204. 61 Degli Esposti, Renzi and Rehren 2016: fig. 3. 62 An arrowhead with the tip folded to form a loop was found in an Iron Age grave (BHS 84) at Jabal al-Buhais and also shows an incised decoration on the blade (Jasim 2012: 248 and fig. 196/14). Arrowheads with incised marks appears in the late Wadi Suq (Middle Bronze Age) period and become more common during the Iron Age (Magee 1998b). 63 Potts 1998: 185. Socketed arrowheads were collected from Wadi Suq contexts in the necropolis of Jabal al-Buhais – where they disappear at the end of that period (Jasim 2012: 294) –, Jabal Hafit, Batin 1, Qattara and Shimal (Cleuziou 1978-1979). Other examples come from Daya, Samad, Khudra, al-Wasit, Suwaiq, Masira and from residential contexts at Nud Ziba and on the Ghanada island (Weeks 1997). 64 Some scholars label this category of objects as daggers rather than spearheads, probably on the basis of the tanged and not socketed junction with the shaft, while for others what matters are the dimensions, accordingly with the assumption that “the spear is just a dagger at the end of a shaft” (Gordon 1953: 68).

308

marzia sasso

nificantly, they are also thinner than the arrowheads from the site. Moreover, one example was found intentionally bended (no. 76), significantly coming from US471, a layer that has been interpreted as one of the foundation deposits distinguished by some sort of ceremonial nature (see Chapter 5) and connected to the site’s construction. Also the fact that no. 77, the only spearhead analysed so far, was made of unalloyed copper 65 suggests that functional efficiency was not sought during its production.

Hoes When tools are considered, very few are those found at Husn Salut, for which, however, wear marks confirm an actual and prolonged use. This is specifically the case for a few hoes, that reflect the agricultural landscape that in all likelihood surrounded the site during the Iron Age. Among them, no. 82 presents a particular case as it was found broken into two pieces that had likely been stored for re-melting, as shown by the presence of an arrowhead stuck between them. This seems further confirmed by an association in the same context, US565, with ingot fragments, some arrowheads, and copper-base scrap metal.66 The better preserved examples measure around 19 cm67 and have a sub-rectangular shape with rounded edges (although use wear surely accounts for much of the edge smoothing) and a roughly central tubular shaft. Parallels are known from Rumeila, Saruq al-Hadid, Masira, and from some Iron Age funerary context in the Oman Peninsula.68 Tool marks possibly left by hoes of the same kind found at Husn Salut were identified at al-Madam, linked in this case to building activities.69

Axes Somehow among weapons and tools stands the only utilitarian axe found at Husn Salut, no. 83, however found in a context (US337) that has been connected to the construction of the raised platform SF49 and therefore interpreted as somehow ceremonial (see Chapter 5). The associated radiocarbon date secure a dating to the second half of the second millennium (see fig. 41), consistently with the date proposed for a similar object discovered at Tell Abraq70 (see also Chapter 10). No. 83 was cast, as indicated by the casting burr on the shaft’s back, and is 10 cm long, like other typologically similar examples from Rumeila, Khudra, Muweila, and Saruq al-Hadid,71 apart from the mentioned Tell Abraq one. All these specimens have a plain tubular shaft and a slightly trapezoidal blade, except the Muweila example that shows a flaring blade more similar to the ceremonial type mentioned above.72 Despite the context of provenance, that no. 83 was originally meant for actual work is indicated by its visible wear marks, and also suggested by its chemical composition, as it contains 2.6% of tin:73 while this rate is not sufficient to

Degli Esposti, Renzi and Rehren 2016: fig. 3. 66 A ceremonial/votive nature for this context cannot be definitely rejected, although it seems unlikely given the ‘open’ position, atop Terrace I, and the absence of other elements usually distinguishing the so-called foundation deposits, such as pottery and animal bones. 67 For no. 80 this is clearly an hypothetical reconstruction. 68 Boucharlat and Lombard 1985: pl. 63; Herrmann, Casana, and Qandil 2012: 62; Yule and Weisgerber 2001: 85 and pl. 52; Weeks 1997: 25. Further afield in the Arabian Peninsula, a shaft hole hoe was discovered near Dharan (Van Beek and Mandaville 1963). 69 Gallego Lopez 2010: 173. 70 Magee 1996b: fig. 5; see also Phillips 2010: 73. 71 Boucharlat and Lombard 1985: pl. 62/15; Weisgerber 1988: 287 and fig. 159/3; Magee 1996b: fig. 20; Weeks et al. 2017: fig. 3/SF30085. 72 A typological discussion of these items would however be unrealistic, given the extreme rarity of published examples faced with the considerable number of excavated ones, not only from burials (i.e. the Qidfaʿ tomb, on display in the Fujaira museum) but also from large ‘industrial’ sites such as al-Safa and Saruq al-Hadid. 73 Degli Esposti, Renzi and Rehren 2016: fig. 3. 65

7 . metal objects

309

provide the blade with the advantages of a true bronze alloy, it anyhow indicates beyond doubt a deliberate addition.74 A similar tin rate, 3.5%, was detected for the Tell Abraq example.75

Blades Although highly fragmentary, no. 84 has been identified as a blade. Around 1.5 cm wide, it has a straight side and a curved opposite one, consistently with a possible knife blade. ICP-MS analyses indicate a tin content of 1.1%,76 not really technologically useful but undoubtedly intentional (here including the possible use of recycled metal). Objects with a similar shape were discovered at Dibba,77 Mleiha and Ghalila, thus post-Iron Age contexts, and were interpreted by M. Mouton, at least for the two latter sites, as scissor blades.78 Unfortunately, the Salut specimen is two badly preserved to allow this comparison. Metal knifes are known form numerous sites, among which Masafi and, again, Mleiha.79

Weights From a context rich in copper-base scraps comes no. 85, a truncated pyramidal object of uncertain interpretation. Its morphology could suggest it is a small weight, but parallels are unfortunately scarce and made in lead.80 This might also depend on the worth of the metal, which possibly led to the choice of less valuable materials such as, in some cases, pottery, as witnessed at Husn Salut by a large number of coarse ware sherds that were reshaped and pierced through.

Studs Possibly connected to carpentry works is a probable circular head stud with a diameter of 4.5 cm (no. 86). Similar objects, slightly smaller, were discovered at Dibba (Emirate of Sharja), but come from layers dated to the 1st century BC and the 1st centrury AD.81

Needles Three complete needles, a category of objects for which parallels are extremely numerous over the whole Oman peninsula, were discovered at Husn Salut, indicating sewing was likely carried out at the site. Dimensions range between 6.5 and 10 cm, except one exceptional specimen measuring 27 cm, with a diameter of 0.5 cm (no. 89). Its dimensions suggest it could have been used for sewing leather, as proposed for a specimen found at Rumeila.82 Another cast specimen, no. 87, bears two incised parallel lines just below the needle eye.

74 The issue remains however unclear, as these amounts could be obtained by using recycled tin-bearing metal (cf. Degli Esposti, Renzi and Rehren 2016: 85). 75 Potts 1990b: 119. 76 Degli Esposti, Renzi and Rehren 2016: fig. 3. 77 Jasim and Yusif 2014: fig. 43/32. 78 Mouton 1992: 108. 79 Benoist et al. 2015: 31 and figs 5/1, 8/2; Mouton 1992: figs 42 (2nd-1st century BC), 116 (3rd-4th century AD). 80 At Mleiha a few truncated conical lead weights with a hole at the tip are known, unfortunately not datable as they come from superficial or anyhow reworked contexts (Mouton 1992: 112). The original presence of a similar hole in the Husn Salut example cannot be ruled out, given its incompleteness. 81 Jasim and Yusif 2014: fig. 43/37-38, 42. 82 Boucharlat and Lombard 1985: 61.

310

marzia sasso

Chisels and burins Several tools for stone or metal working were discovered at Husn Salut. While a few are reasonably identifiable as chisels, others, being too fragmentary and lacking the working end, can just be supposed to be burins or generic pointed tools when distinguished by a squared section.83 No. 104, with a slightly twisted stem, could be interpreted as a nail lacking its head, although shape and dimensions could also recall the so-called ‘stun bolts’, projectiles used in bird hunting and attested at numerous sites of the Near East during the Late Bronze Age and the Iron Age.84 Chisels from Husn Salut, ranging between 0.5 and 0.9 cm in length and between 0.2 and 0.5 in thickness, usually have a rectangular section, in only one case is it half-circular (no. 92) with a flattened hemispherical, sharpened head. For this latter case parallels can be found from the necropolis of Jabal al-Buhais,85 while those with a flat section find a comparison from Rumeila.86 The diversified working destination of chisels implies the existence of numerous dimensional classes, as well as a marked similarity in shapes, to the extent that very similar objects can be found in contexts that are both geographically and chronologically distant from each other.

Clamps The use of copper-base clamps for the restoration of stone vessels is directly attested at Husn Salut by a wall sherd that still shows such a metal item holding together its two fragments (S85 – see Chapter 6, no. 75). Clamps discovered at the site are composed of two opposite small copper-base plaques (3-4.5 cm long and c. 1 cm wide), connected by two rivets. Similar objects are known from numerous contexts, among which for example the Sharm tomb, Shimal, Asima, and Muweila.87 Clearly, these items could be used for other aims, such as the fabrication of leather items or composite metal objects.

P ersonal

adornments and decorations

Pins and hairpins To this category belong a few simple pins, objects that are widely spread over the Oman peninsula as early as the third millennium BC without significant variations, and a possible hairpin, whose chronology is unfortunately uncertain (no. 115).

Rings Thirteen small copper-base rings (nos 116-125) were obtained by bending a circular section rod (section diameter c. 02. cm) with pointed or rounded ends. Given the overall diameter (between 2.3 and 2.7 cm)88 they may be interpreted as earrings, rather than bracelets. However, as nine of them come from the same context (US23, the original floor of Building 1’s room 2b), it may be suggested that they were part of a chain, although a utilitarian destination can not be excluded.



Parallels can be proposed with objects coming from the necropolis of Jabal al-Buhais (Jasim 2012: figs 162/6; 225/3-4; 296/1; 319/3). See Genz 2007; Chitwood 2015. 85 Jasim 2012: fig. 296/6. 86 Boucharlat and Lombard 1985: pl. 62/13. 87 Weeks 2000: figs 10-11; Donaldson 1984: figs 13/5, 27/15; Vogt 1994: fig. 8/27; Weeks 2000: 189. 88 One only example (MB13) is slightly smaller, with a diameter of 1.7 cm. 83 84

7 . metal objects

311

For these types of objects, clearly rather simple, parallels are numerous and widely scattered chronologically. The case for no. 116 seems to be different, unfortunately incomplete but for which an interpretation as a brooch can be put forward.

Decorative plaques Finally, a number of small, disc-shaped copper-alloy foils could be interpreted as clothes or belt decorations. More interesting however are a few hemispherical plaques, ranging between 1.8 and 2.6 cm in diameter (with one exception measuring 4 cm – no. 131), decorated in three cases with small raised dots (no. 127, no. 128 and no. 129). For these objects fitting parallels are known from Rumeila, where they were interpreted as metal bases for organic material quivers, meant to host the metal arrowhead.89

R aw

material : ingots

The discovery of crucible fragments from contexts related to the outer wall and thus datable to the HSIII phase suggests that small scale copper-working was present at Husn Salut. Scrap metal, copper-base melting droplets and broken fragments possibly stored for re-melting are also attested, although their interpretation is not so univocal and, despite being clearly connected to copper working,90 they do not imply that this was conducted on the site, given also the uncertainty about the function of the small furnace SF21 discovered in the upper part of the site (see Chapter 2). The same applies to a number of circular ingots, almost all fragmentary, discovered from different contexts (nos 133-146). When a sufficiently large part of the ingots is preserved, they can be divided into two types: discoid, with flat section, and circular, with plano-convex section. Archaeometric analysis on six of these ingots and on a big lump of metal that comprised at least four other ingots stuck together revealed that they are not made of bronze but rather of dirty copper, in most cases (five) characterised by a high iron content alongside slag inclusions. This indicates that raw material was circulating as unrefined ingots.91 Where recognizable, ingot diameter ranges between 7 and 12 cm, consistently with that reported for a large number of ingots discovered at Masafi 1, buried inside two storage jars below a floor of the administrative building of the site, which also displays an analogous composition to those from Husn Salut, specifically for what concerns the high iron content.92 Iron content variations in the sampled ingots from Masafi 1 suggested that they corresponded do different degrees of purification, needing in any case further refinement before being suited for production,93 which is thus also the case for the ingots from Husn Salut. That the objects recovered at Husn Salut could not derive from these ingots is also evidenced by the composition of fifteen object samples analysed by ICP-MS, showing very low iron percentages.94 Moreover, the presence of ingot halves and also of a quarter of an ingot, alongside the regular aspect of their fractures, could suggest an intentional fragmentation, possibly linked to specific needs of particular productions (small objects?) as well as to a more articulated system of raw metal circulation. While this last observation would be coherent with some recurrence in ingot weights or in discernible fractions thereof, at Husn Salut no possible scale can



Potts 1998: 199-200. See Degli Esposti, Renzi and Rehren 2016. 91 Degli Esposti, Renzi and Rehren 2016: 85. 92 Cf. Goy et al. 2013. 93 Ibid., see also Benoist et al. 2015. 94 Degli Esposti, Renzi and Rehren 2016: fig. 3. 89 90

312

marzia sasso

be hypothesized. Taking into account the fragmentary nature of the specimens, just a generic range of their weight between 600 and 1280 g can be mentioned. The rough appearance of the ingots’ lower side (the concave part) is somehow similar to those from Masafi, where this aspect led to the hypothesis that they were shaped into a hole dug in the sandy ground.95 For the ingots of Masafi 1 the likelihood of a votive offer has been underlined, on the basis of their context and of the discussed composition, being as they are unsuited for production.96 More complicated appears to be the situation at Husn Salut. Here, both a ritual/votive interpretation and a functional interpretation could be plausible, the latter linked to a small scale production as mentioned above and possibly suggested by the retrieval of copper-base slag and scrap metal from US23, the floor of Building 1’s room 2b (see Chapter 2), together with the mentioned, late, crucible fragments. On the other hand, seven ingots were discovered in US224, a context of Area 5 significantly close to and above one of the so-called foundation deposits that characterises the construction of the site’s outer wall (see Chapter 5). Therefore, this concentration lends itself to different interpretations, mentioned above. Given the ceremonial function that the site surely had, albeit not univocal, it would not be inconsistent to suggest that the ceremonial deposition of metal items, somehow recalling metallurgy tout court as witnessed at sites like Bithna and Masafi,97 was among the ritual practices carried out.

I ron

objects

Out of the three most complete iron artefacts recovered at Husn Salut, two can likely be recognized as knife blades, while the third, extremely fragmented, can be interpreted as a grip. The two blades, nos 148-149, both incomplete, have a straight back and a slightly curved and more sharpened edge, just like the above mentioned copper-alloy possible knife (no. 84); their width ranges between 1.3 and 1.9 cm. The peculiar scarcity of Iron working evidence in South East Arabia until the late second half of the first millennium BC is well known, a situation for which an explanation can be found in the abundant occurrence of copper ores in the al-Hajjar mountains and in the long-established tradition of copper-working. The rare occurrence of iron items and the associate retrieval of such items in specific contexts contribute to connoting them as possible prestige artefacts. From Muweila, around twenty weapons or simple blades were for example discovered in the columned hall of Building II.98 Thus, it is not surprising that the fragments of the possible grip (no. 147) were discovered at Husn Salut in the repeatedly mentioned foundation deposit US75. However, this scarcity may only reflect the situation at an earlier moment of the Iron Age period, as the discovery of the substantial site of Saruq al-Hadid is starting to indicate. There in fact a remarkable amount of ferrous remains was discovered, the majority of which appears to be related to Iron Age II and later contexts, being less abundant in Iron Age I and earlier Iron Age II layers.99 It may be of some significance that the majority of the ferrous objects from Saruq al-Hadid consists in weapons, and double-edged or single edged fragments are also considerably numerous.100 Consistently, when searching for parallels from South East and South Arabia, iron knifes are more easily found, and they all come from later contexts.101 Other ferrous fragments were discovered at Husn Salut in the late, massive layer US3. Although this context contained mixed materials also including Islamic items, it cannot be excluded that these fragments actually are to be

Benoist et al. 2015: 28. Goy et al. 2013: 139. 97 See Benoist et al. 2015: 26-33. 98 Magee 2003a: 189. 99 Weeks et al. 2017: 46 and table 1. 100 Weeks et al. 2017: 46, 48. 101 For example at Mleiha, dated to the 2nd-1st century BC and to the 3rd-4th century AD (Mouton 1992) and ed-Dur (Mouton 1992: fig. 116/69), but also from Building G at Makaynun, in the Hadramawt, dated to the 2nd-4th century AD (Benoist et al. 2014: fig. 8).

95 96

7 . metal objects

313

connected with the late Iron Age phases of the site. Be that as it may, they were not included in this catalogue due to their extremely poor preservation.

G old

ornaments

The only gold object discovered at Husn Salut is a small spherical bead, only 0.6 cm in diameter, coming from US23, the floor of room 2c in Building 1 (no. 150). A few gold beads are known from funerary contexts, such as the tombs at Shimal, the Sharm tomb and the necropolis of Jabal al-Buhais,102 as well as from settlements, such as at Rumeila and Saruq al-Hadid.103 However, these examples are all typologically distinguished from that from Husn Salut, and for the Sharm tomb beads a parallel has been proposed with materials on display in the Dubai Museum and dated to the first half of the second millennium.104 Recent excavations at Saruq al-Hadid have finally brought to light a more considerable amount of gold items, among which a number of beads can be pointed out that rather closely resemble the one from Husn Salut.105 Where the metal came from is currently unclear; it may have come from either outside or inside the Oman peninsula as surveys showed the presence of high gold and silver concentrations in the ore bodies of the Omani mountains.106

Vogt and Franke-Vogt 1987: 54, fig. 17/14-16; Weeks 2000: 180; Barker 2001: fig. 2; Jasim 2012: 291 and figs 7/6, 9; 220/5-7; 311/6-8. Boucharlat and Lombard 1985: 62, pl. 73/10; Herrmann, Casana and Qandil 2012: 62 and fig. 6/c-e. 104 Barker 2001: 211. 105 Weeks et al. 2017: fig. 22/SF20035, SF20384. 106 Weeks 2000: 184. 102 103

314

marzia sasso

CATALOGUE For each entry, the inventory number is followed by the indication of the context of provenance, by the dimensions given in cm (l.= length; w.= width; h.= height; d.= diameter; th.= thickness; wg.= weight), and by a short description.107 When available, main element composition is also reported.

C opper - base

objects

Snakes 1. Snake pl. 82 MB7, US1 w. 1.9; th. 1.3 Composition (wt %): Cu 98.50, Fe 0.01, Sn 0.01 Rhomboidal head. The first curve of the body is partially preserved. Smashed in many points on the head. 2. Snake pl. 82 MB15, US3 l. 16.9; w. 1.3; th. 0.5 Triangular head, pointed tail and impressed dots decoration along the upper surface of the body. This is meant to represent the snake scales, as also visible on pottery, as an alternative to impressed circles and incisions. Mould cast. 3. Snake pl. 82 MB111, US57 l. 20.3; th. 2; wg. 352 Rhomboidal head, marked globular and perforated eyes and applied spot decoration along the upper surface of the body, that tapers at the tail end. Mould cast. 4. Snake pl. 82 MB127, US118 l. 16.3; w 1.8; th. 1.1; wg. 236 Triangular head, rounded tail. The body is decorated with impressed circles while the head shows few small holes.

5. Snake pl. 82 MB130, US1 l. 23; w. 5.5; th. 1.5; wg. 646 Trilobate head with two pronounced vertical elements, possibly horns. The body, made by rather unskilled work, is decorated with applied spots and tapers at the tail end. Mould cast. 6. Possible snake pl. 88 MB133, US132 l. 3.6; w. 2.3; th. 0.4 Fragment of a plaque with a discoidal part from which protrudes an arched segment. Punctuated depression on the upper surface. Possible head fragment of a snake. 7. Possible snake pl. 82 MB140, US165 l. 5.2; w. 3.2; th. 1.5; wg. 128 Fragment of a flat object, slightly arched. Possible body fragment of a snake, badly preserved. 8. Snake pl. 82 MB143, US74 l. 19.8; w. 1; th. 0.3; wg. 27 Raised triangular head, pointed tail. Curved body with four bends, few small circular impressions on the tail likely representing scales (aborted decoration?). Shaped from one flat strip of metal. 9. Snake MB193, US394 l. 4.5; w. 0.6; th. 0.3; wg. 2.4

pl. 82

Raised rhomboidal head bent at the edge, flat body and rounded tail. Shaped from one folded sheet of metal. 10. Possible snake pl. 88 MB206a, US454 l. 5; th. 0.5 Possible body fragment of a snake, badly corroded.

Ladles 11. Ladle pl. 83 MB107, US75 l. 17.5; cup d. 6.1; wg. 65 Hemispherical cup, long handle bent at the end with a rounded profile. A crisscross mark with an extending vertical line is engraved in the upper part of the handle, and may indicate the ownership of the object. Mould cast. 12. Ladle pl. 83 MB112, US75 l. 9.7; cup d. 5.5; wg. 25 Hemispherical cup, handle bent at the end with a rounded profile. Shaped from a single metal sheet. 13. Ladle pl. 83 MB146, US175 cup d. 5; wg. 32.2 Composition (wt %): Cu 83.48; Fe 0.14; Sn 0.15; Pb 0.04 Hemispherical cup, short handle slightly bent at the end with a rounded profile. Badly corroded.

107 For most of the entries, the catalogue necessarily builds upon the excavation database, amended when needed. As such, a few of them replicate those of the selected objects illustrated in Condoluci 2015b.

7 . metal objects

14. Ladle pl. 83 MB194, US394 l. 9.4; w. 2.6; th. 0.2; cup d. 0.5; wg. 2.6 Slightly concave cup, handle sharply bent at the end with a rounded profile. Probably folded. 15. Ladle pl. 83 MB234, US727 l. 8.7; w. 3.2; th. 0.4; cup d. 3.2 Hemispherical cup, handle bent at the end with a rounded profile.

Miniature axes 16. Miniature axe pl. 83 MB76, US28 l. 5.1; max w. 3.6; blade th. 0.25 Trapezoidal blade with tubular opened shaft. Made by hammering a single piece of metal. 17. Miniature axe pl. 83 MB122, US75 l. 2.4; max w. 1.3; blade th. 0.15 Trapezoidal blade with tubular opened shaft, realized by rolling up the tapering edge of the plaque. Made by hammering a single piece of metal. 18. Miniature axe pl. 83 MB167, US355 l. 4.5; max w. 4.9; blade th. 0.2 Flared pentagonal shaped blade with tubular shaft. Slightly damaged on the edge. Hammered. 19. Possible miniature axe pl. 83 MB198, US377 l. 4.6; w. 2.7; th. 0.2 F-shaped object, flat in section with the first arm slightly tapering toward the end. The first arm has a rough triangular shape with a broadly protruding end, originally connected with the second by an arched bridge. Possible fragment of a miniature axe. 20. Miniature axe MB239, M247

pl. 83

-

catalogue

l. 3; max w. 2.6; max th. 0.4; min th. 0.2 Trapezoidal blade with slightly concave long sides. The shaft is missing.

Vessels 21. Cauldron pl. 84 MB85, US36 h. 20; rim d. 31 Hemispherical body with two loops vertical handles overlying the everted rim, flat base. Made by hammering a single piece of metal. 22. Spouted cup pl. 88 MB214, US568 h. 5.0; rim d. 7.4; th. 0.3; wg. 33.1 Spouted cup with. very thin wall and flat base. Hammered. 23. Spouted cup pl. 88 MB228, M227 l. 4; w. 1.3; th. 0.1 Rim fragment of a spouted bowl. The rim is rounded. 24. Lid pl. 84 MB57, US23 d. 1.8; d. 3.1; th. 0.6 Small circular lid with conical knob. Small circular depression in the middle of the lower face.

Oddities 25. Object (standard?) pl. 84 MB19, US3 h. 8.4; base d. 12; antennas d. 1.1/1.8; base th. 0.8 Object composed of a hemispherical base with V-shaped antennas. Possibly the top of a standard.

Arrowheads 26. Arrowhead MB37, US1

pl. 84

315

l. 3.8; blade l. 3.2; max w. 1; th. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, flat blade. The stem is rectangular in section. 27. Arrowhead pl. 84 MB231, M237 l. 4.3; blade l. 3; w. 1; th. 0.1; stem th. 0.3 Type: leaf-shaped Tanged arrowhead, leaf-shaped, flat blade. The tip is broken and slightly folded. The stem is rectangular in section. 28. Arrowhead pl. 84 MB92, US42 l. 4.5; blade l. 3; w. 1.4; th 0.8; stem max w. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is rectangular in section. 29. Arrowhead pl. 84 MB225, US587 l. 5; w. 1.4; th. 0.4; stem th. 0.3 Type: leaf-shaped Composition (wt %): Cu 87.87; Fe 0.42; Sn 0.01 Tanged arrowhead, leaf-shaped, flat blade. The stem is square in section. 30. Arrowhead pl. 84 MB64, US3 l. 5.8; blade l. 4.5; w. 1.6; th. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is rectangular in section. 31. Arrowhead pl. 84 MB86, US45 l. 5.2; w. 1.2; stem w. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is rectangular in section.

316

32. Arrowhead pl. 84 MB128, US1 l. 5.7; stem th. 0.2 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is rectangular in section. 33. Arrowhead pl. 84 MB78, US36 l. 4; blade l. 2.9; w. 1.4; th. 0.1 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is quadrangular in section. 34. Arrowhead pl. 84 MB201, US1 l. 3.6; w. 1.4; th. 0.4; wg. 3.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is almost square in section. 35. Arrowhead pl. 84 MB202, US337 l. 3.7; w. 1.4; th. 0.3; wg. 2.7 Type: leaf-shaped Tanged arrowhead, leaf-shaped, flat blade. The stem is rectangular in section. 36. Arrowhead pl. 84 MB188, M136 l. 4.1; w. 1.4; th. 0.6; wg. 4.4 Type: leaf-shaped Composition (wt %): Cu 82.71; Fe 0.24; Sn 0.33 Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is square in section. 37. Arrowhead pl. 84 MB166, US353 l. 4.8; w. 1.5; th. 0.6; wg. 8.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with elevated trapezoidal midrib. The stem is rectangular in section.

marzia sasso

38. Arrowhead pl. 84 MB176, US333 l. 4.6; w. 1.6; th. 0.1; midrib th. 0.4; stem max th. 0.4; wg. 4.6 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is rectangular in section. 39. Arrowhead pl. 84 MB9, US3 l. 3.7; w. 1.4; th. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The steam is rectangular in section and folded. 40. Arrowhead pl. 84 MB89, US1 l. 5; blade l. 3.4; w. 1.4; th 0.3 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is quadrangular in section. 41. Arrowhead pl. 84 MB178, US1 l. 6.7; w. 1.6; th. 0.3; wg. 10.3 Type: leaf-shaped Composition (wt %): Cu 70.70; Fe 0.08; Sn 0.60; Pb 0.02 Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is square in section. 42. Arrowhead pl. 84 MB183, US371 l. 6.8; w. 2; th. 0.4; wg. 12 Type: leaf-shaped Composition (wt %): Cu 76.10; Fe 0.31; Ni 0.77; As 0.26; Co 0.05; S 0.15 Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib. The stem is triangular in section. 43. Arrowhead pl. 84 MB70, US1 l. 3.2; w. 1.4; th. 0.4 Type: leaf-shaped Tanged arrowhead, leaf-shaped, flat

blade. The stem is rectangular in section. 44. Arrowhead pl. 84 MB88, US26 l. 2.4; w. 1.8 Type: leaf-shaped Composition (wt %): Cu 86.53; Fe 0.07; Sn 0.01; Pb 0.02 Fragment of a leaf-shaped arrowhead, with slightly elevated flat midrib. 45. Arrowhead pl. 84 MB203, US1 l. 6; w. 1.2; th. 0.4; wg. 5.1 Type: leaf-shaped Composition (wt %): Cu 88.24; Fe 0.29; Pb 0.01 Re-worked arrowhead, originally leafshaped, with flat blade. The stem is almost square in section. With the reshaping of the object, one side of the blade was removed, likely from beyond the stem edge. Possibly reused as a razor. 46. Arrowheads pl. 84 (B, C) MB22, US3 A: l. 2.3; w. 1.2 B: l. 3.2; w. 0.8 C: l. 3.3; w. 0.8 Type: A: triangular, B and C: leaf shaped Fragments of three arrowheads, possibly triangular and leaf-shaped, with flat blades. The leaf-shaped ones are re-worked. With the reshaping of the objects, part of the blades were removed. Possibly reused as a razors (C has a sharpened edge). 47. Arrowheads pl. 84 MB108, US79 A: l. 4.7 B: l. 5.1 Type: A: lanceolate (variant B), B: leafshaped Two tanged arrowheads, joined together by corrosion and wrapped in a sheet of bronze. One is lanceolate shaped, with flat blade and the stem is quadrangular in section. The second arrowhead is

7 . metal objects

leaf-shaped and the stem is rectangular in section. 48. Arrowheads pl. 84 MB47, US23 A: l. 3.2; w. 1 B: l. 2.4; w. 1.2 C: l. 2.1; w. 1.1 D: l. 2.9, w. 1 Type: A: triangular, B and C: indeterminable, D: leaf-shaped Fragments of four tanged arrowheads, one triangular shaped, one leaf-shaped and two indeterminable. Three arrowheads have elevated flat midrib, one has a flat blade. The stem section is rectangular in two arrowheads and circular in one. 49. Arrowhead pl. 88 MB206b, US454 l. 4.7; blade l. 3.1; w. 1.1 Type: leaf-shaped Tanged arrowhead, leaf-shaped, with slightly elevated flat midrib on one side and flat blade on the other. The stem is rectangular in section and tapered downwards. Part of a blade edge and the stem are folded. 50. Arrowhead pl. 85 MB136, US132 l. 4.3; w. 1.6; th. 0.2; stem max th. 0.3; wg. 4 Type: lanceolate (variant A) Tanged arrowhead, lanceolate shaped, with flat blade. The tip end is folded. The stem is rectangular in section.

-

catalogue

Tanged arrowhead, lanceolate shaped, flat blade. The stem is rectangular in section. 53. Arrowhead pl. 85 MB173, US333 l. 5.2; w. 1.1; th. 0.3; wg. 4.3 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, flat blade. The stem is rectangular in section. 54. Arrowhead pl. 85 MB49, US23 l. 4.9; blade l. 3.2; w. 0.3 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, flat blade. The stem is oval in section. 55. Arrowhead pl. 85 MB24, US3 blade l. 3.7; max w. 1; max stem th. 0.3 Type: lanceolate (variant A) Tanged arrowhead, lanceolate shaped, flat blade. The stem is rectangular in section. 56. Arrowhead pl. 85 MB223, US576 l. 7; w. 1.7; stem th. 0.4; wg. 9.6 Type: lanceolate (variant A) Composition (wt %): Cu 70.85; Fe 0.12; Sn 9.79 Tanged arrowhead, lanceolate shaped, flat blade. The stem is square in section.

51. Arrowhead pl. 85 MB160, US257 l. 4.4; w. 1.3; th. 0.4; stem th. 0.4; wg. 5.7 Type: lanceolate (variant A) Tanged arrowhead, lanceolate shaped, with flat blade. The stem is rectangular in section.

57. Arrowhead pl. 85 MB25, US3 blade l. 3.4; max stem th. 0.4 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, with slightly elevated flat midrib. The stem is rectangular in section. Slightly arched.

52. Arrowhead MB174, US333 l. 6.8; w. 2; th. 0.3; wg. 6.2 Type: lanceolate (variant A)

58. Arrowhead pl. 85 MB163, US300 l. 5.8; w. 1.8; th. 0.4; stem max th. 0.5; wg. 13

pl. 85

317

Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, flat blade and sharpened edges. The stem is rectangular in section. 59. Arrowhead pl. 85 MB87, US39 l. 5.2; w. 1.4; th. 0.3; stem max w. 0.4 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, with slightly elevated flat midrib. The stem is quadrangular in section. An incised cross is probably visible on one face. 60. Arrowhead pl. 85 MB189, US386 l. 6.6; w. 1.4; th. 0.3; wg. 7 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, flat blade. The stem is square in section. 61. Arrowhead pl. 85 MB124, US88 l. 5.3; w. 1.2; stem th. 0.3 Type: lanceolate (variant A) Tanged arrowhead, lanceolate shaped and intentionally folded. The stem is quadrangular in section. 62. Arrowheads pl. 85 MB191, US387 l. 10.7; w. 2.3; th. 0.4; wg. 18 Type: lanceolate (variant A) Tanged arrowhead, lanceolate shaped, flat blade. The steam is rectangular in section. 63. Arrowheads pl. 85 MB52, US29 A: l. 2.9; w. 1.1; th. 0.2 B: l. 1.7; w. 1.2; Type: A: lanceolate (variant B), B: indeterminable A: one complete tanged arrowhead, lanceolate shaped, flat blade. The stem is rectangular in section and the arrowhead is rolled up. B: one blade fragment of an arrowhead, indeterminable shape, flat in section. The junction with the stem is visible.

318

64. Arrowhead pl. 88 MB80, US36 l. 3.2; w. 1.7; th. 1.1 Type: lanceolate (variant B) Tanged arrowhead, lanceolate shaped, with slightly elevated flat midrib. The stem is rectangular in section. The point and the stem are intentionally folded. 65. Arrowhead pl. 85 MB27, US3 l. 11; w. 2.6; th. 0.6 Type: triangular Tanged arrowhead, triangular shaped, flat blade. The stem is rectangular in section. 66. Arrowhead pl. 85 MB98, US1 l. 4.2; th. 0.5; stem d. 0.4 Type: triangular Tanged arrowhead, triangular shaped, with prominent trapezoidal midrib. The stem is oval in section. 67. Arrowhead pl. 85 MB99, US1 l. 4.8; w. 1.2; th 0.4 Type: triangular Tanged arrowhead, triangular shaped, with prominent midrib, rhomboidal in section. The stem is rhomboidal in section. The blade is bent. 68. Arrowhead pl. 85 MB175, US333 l. 6; w. 1.7; th. 0.1; midrib th. 0.5; stem th. 0.3; wg. 5 Type: triangular Tanged arrowhead, triangular shaped, with prominent midrib, rhomboidal in section. Sharp edges. The stem is rhomboidal in section. 69. Arrowhead pl. 85 MB210, US549 l. 5.4; w. 1.7; th. 0.5 Type: triangular Tanged arrowhead, triangular shaped,

marzia sasso

with prominent midrib, rhomboidal in section. The stem is rhomboidal in section. 70. Arrowhead pl. 85 MB26, US3 l. 4.1; blade l. 2.7; w. 1.3; th. 0.3 Type: triangular Tanged arrowhead, triangular shaped, with slightly elevated flat midrib. The stem is rectangular in section. 71. Arrowhead pl. 85 MB3, US3 l. 3.8; w. 1.2; th. 0.6 Type: triangular Tanged arrowhead, triangular shaped, with slightly elevated flat midrib. Slightly folded tip, sharpened edges. The steam is quadrangular in section. 72. Arrowhead pl. 85 MB126, US103 l. 3.5; th. 0.2; stem th. 0.3 Type: triangular Tanged arrowhead, triangular shaped, with slightly elevated flat midrib. The stem is rectangular in section. 73. Arrowhead pl. 85 MB123, US108 l. 4; stem l. 3.4; stem th. 0.3 Type: indeterminable Fragment of a tanged arrowhead. Only the stem and the lower part of the blade are preserved. The stem is rectangular in section. 74. Arrowhead pl. 85 MB224, US591 l. 11.5; w. 2.4; wg. 26 Type: triangular Composition (wt %): Cu 73.78; Fe 0.13; Sn 0.01 Tanged arrowhead, triangular shaped with concave sides, flat blade. The stem is rectangular in section. Probably reworked for a different purpose.

Spearheads 75. Spearhead pl. 85 MB192, US391 l. 21.1; w. 3; th. 0.25 Tanged spearhead with slightly elevated flat midrib. The steam is quadrangular in section. Probably moulded. 76. Spearhead pl. 86 MB204, US471 l. 16; w. 1.7; th. 0.1; wg. 13 Tanged spearhead with flat blade, rounded tip, arched profile. The steam is rectangular in section. The folded blade seems to suggest a non-functional object. 77. Spearhead pl. 85 MB211, M138 l. 18; w. 2.3; th. 0.25; wg. 32 Composition (wt %): Cu 91.22; Fe 0.05; Sn 0.15; Pb 0.01 Tanged spearhead with flat blade and pointed tip. The steam is squared in section. 78. Spearhead not illustrated MB235, US731 l. 6; w. 2.1; th. 0.3 Fragment of a spearhead with flat blade and rounded tip.

Hoes 79. Hoe pl. 86 MB12, US1 l. 7.4, w. 3.8; th. 0.8; shaft d. 2.2 Ellipsoidal hoe, extremely worn, with a circular shaft in the middle which extends as a cylinder. 80. Hoe pl. 86 MB18, US3 l. 9.5; w. 10.8; th. 0.9 Sub-rectangular hoe, half complete, with sharpened and rounded edges. Moulded.

7 . metal objects

81. Hoe pl. 86 MB161, US256 l. 4.8, w. 5.3; th. 0.8; wg. 97 Probable hoe fragment, with sharpened edge. The preserved side is slightly rounded. 82. Hoe pl. 88 MB215, US565 l. 12.4; w. 10.8; th. 0.5; shaft d. 4.6 Composition (wt %): Cu 91.29; Fe 0.08; Sn 0.31; Pb 0.73 Sub-rectangular hoe, complete, with sharpened and rounded edges. Moulded. Part of a series of objects fused together by corrosion (fragments of arrowheads and scraps) and possibly collected to be re-fused.

Axes 83. Axe pl. 86 MB196, US377 l. 9.6; shaft h. 3.7; max w. 2.3 Composition (wt %): Cu94.10; Fe 0.10; Sn 2.59; Pb 0.03 Elongated axe with slightly triangular shaped blade. Vertical rib on the shaft back side that could be related to the molding process. Moulded.

Blade 84. Blade not illustrated MB199, US377 l. 5; w. 1.3 Composition (wt %): Cu 76.82; Fe 0.14; Sn 1.10; Pb 0.02 Fragment of a possible blade with a pointed tip, slightly bended.

Weights (?) 85. Object not illustrated MB217, US565 l. 1.8; h. 2.4; w. 1.5 Object in the shape of a truncated pyra-

-

mid with rectangular base. Possibly a weight.

Stud 86. Stud pl. 86 MB185, M130 l. 4.5; d. 4.6; th. 0.4; wg. 37 Stud with circular head, slightly convex, and square section stem, tapered toward the end. At the junction with the head, the stem is cylindrical, witnessing its shaping by hammering. More likely a stud than an actual nail.

Needles 87. Needle pl. 86 MB38, US3 l. 10.4; eye max w. 0.5 Complete needle with incised decoration of two horizontal lines below the eye. Moulded. 88. Needle MB145, US36 l. 6.5; th. 0.3 Complete needle, slightly bent.

pl. 86

89. Needle or hairpin pl. 86 MB96, US1 l. 27; th. 0.5 Long pin, with pointed end and head in form of loop. Possible needle or hairpin.

Chisels and burins 90. Chisel pl. 86 MB36, US1 l. 3.1; w. 0.6; th. 0.2 Tool, rectangular in section, with slightly rounded tip. Probably a chisel. 91. Chisel MB55, US1 l. 4.3; w. 0.7; th. 0.3

319

catalogue

pl. 86

Tool, rectangular in section. Probably a chisel. 92. Chisel pl. 86 MB67, US1 l. 3.4; d. 0.5 Tool, circular in section, with a flattened and sharpened end. Probably a chisel. 93. Chisel pl. 86 MB97, US1 l. 6.9; w. 0.9; th. 0.4 Tool, oval in section, with rounded and flattened end. Probably a chisel. 94. Chisel pl. 86 MB102, US114 l. 9.8; max th. 0.5 Tool, rectangular in section, with a flattened end. The object is bent. Probably a chisel. 95. Chisel pl. 86 MB125, US77 l. 4.4; w. 0.5; th. 0.4 Tool, rectangular in section, with a rounded and flattened end. Probably a chisel. 96. Possible chisel pl. 86 MB232, US714 l. 6.3; w. 1.1; th. 0.2 Fragment of a possible chisel with a slightly rounded end.

Rods 97. Rod pl. 86 MB44, US23 h. 3.8; th. 0.2; wg. 2.3 Rod, bent in the form of an ‘S’, probably used as a clasp. Possibly re-worked ring with rounded open ends. 98. Rod pl. 86 MB48, US23 l. 4.5; w. 0.2; th. 0.2 Three joining fragments of a small rod, rectangular in section. Possible fragments of a needle.

320

99. Rod pl. 86 MB59, US23 l. 6.7; d. 0.2 Three joining fragments of a small rod, circular in section, with badly preserved point. Possible fragments of a needle. 100. Rod pl. 86 MB63, US3 l. 2.9; w. 0.5; th. 0.3 Stick fragment, rectangular in section, with pointed end. Possible pin or awl. 101. Rod pl. 86 MB65, US23 Measures: l. 9.2; w. 0.3; th. 0.3 Pointed rod, quadrangular in section. Possible needle or hairpin. 102. Rod pl. 86 MB81, US39 l. 5.6; d. 0.4 Fragment of a pointed rod. Possible fragment of a needle or a hairpin. 103. Rod pl. 86 MB169, US330 l. 5.1; d. 0.6 Complete rod, circular in section. 104. Rod pl. 88 MB197, US390 l. 3.4; head w. 0.6; stem w. 0.4 Composition (wt %): Cu 95.75; Fe 0.21; Sn 0.03; Pb 0.01 Rod with square, slightly twisted, section. Nail or tool missing the tip.

marzia sasso

106. Clamp pl. 86 MB34, US23 l. 2.5; w. 0.9; th. 0.1 Rectangular clamp plaque, with two small holes near the short sides. 107. Clamp pl. 88 MB52, US29 A: l. 3.2; w. 1.1; th. 0.2 B: l. 2.4; d. 0.3 A: rectangular clamp plaque with two small holes, probably for rivets, near the short sides. B: stick fragment, circular in section. Possibly a rivet. 108. Clamp pl. 86 MB159, US264 plaques l. 3.6; w. 1.2; th. 0.2; wg. 3.4 Two rectangular and complete clamp plaques joined together with a rivet. On both plaques are visible small holes for rivets, only one of them is currently preserved. 109. Clamp not illustrated MB200, US377 l. 1; w. 1.5; th. 0.6 Fragment of a clamp plaque with a broken rivet.

Handle 110. Handle not illustrated MB50, US23 l. 2.3; w. 1; th. 0.7 Fragment of a handle. Hammered.

Clamps

Pins and Hairpins

105. Clamp pl. 86 MB6, US1 l. 4.5; w. 1; th. 0.1 Two rectangular clamp plaques joined together with rivets. The complete plaque has three small holes for rivets, only two of them are currently preserved.

111. Pin pl. 87 MB165, US320 l. 11.2; th. 0.5 Complete pin, with pointed end. 112. Pin MB195, US394 l. 5; w. 0.5; th. 0.4 Fragment of a pin.

pl. 87

113. Pin pl. 87 MB219, US568 l. 5.4; th. 0.5 Fragment of a pointed pin, square section. 114. Pin pl. 87 MB226, US588 l. 8.8; d. 0.4 Complete pin, circular in section, slightly rounded end. 115. Hairpin pl. 87 MB5, US3 l. 3.5; d. 0.4 Upper part of a pin, decorated with two deep circular grooves just below the edge. Possibly a hairpin.

Rings and earrings 116. Earring pl. 87 MB21, US8 d. 3.4; th. 0.4 Earring with semicircular shape. Incised decoration of horizontal and wavy lines on the straight segment. 117. Earring pl. 87 MB154, US230 d. 2.8; th. 0.3; wg. 1 Earring with irregular oval shape and rounded open ends. The central segment is straight while the lateral ones are curved. 118. Ring or earring pl. 87 MB13, US3 d. 1.7; th. 0.2 Ring with pointed open ends. Possibly used as an earring. 119. Ring or earring pl. 87 MB42, US23 d. 2.3; th. 0.2 Ring with open ends, possibly used as an earring. 120. Ring or earring MB43, US23

pl. 87

7 . metal objects

-

catalogue

321

d. 2.7; th. 0.2 Ring with rounded open ends, possibly used as an earring.

Hemispherical plaque with a punctuated decoration from inside. Slightly pleated edges.

l. 8.2; max th. 2.4; wg. 282 Fragment of a possible bun-shaped ingot.

121. Ring or earring pl. 87 MB45, US23 d. 2.5; th. 0.2 Ring with rounded open ends, possibly used as an earring.

128. Decorative plaque pl. 87 MB30, US3 h. 1.1; d. 2.6; th. 0.1 Hemispherical plaque with slightly pleated edges.

135. Ingot pl. 88 MB147, US224 l. 11; th. 3 Composition: Cu 81.8; Fe 14.7 Bun-shaped ingot, half-complete.

122. Ring or earring pl. 87 MB46, US23 d. 2.4; th. 0.2 Ring with broken open ends, possibly used as an earring.

129. Decorative plaque pl. 87 MB33, US3 Measures: d 2.3; th. 0.1 Hemispherical plaque with a punctuated decoration from inside. Slightly pleated edges.

136. Ingot pl. 87 MB148, US224 l. 9.5; th. 2.7; wg. 644 Composition: Cu 83.8; Fe 11.7 Discoid shaped ingot, flat section, halfcomplete.

130. Decorative plaque pl. 87 MB41, US23 d. 1.8; th. 0.2 Hemispherical plaque with slightly pleated edges.

137. Ingot pl. 88 MB149, US224 d. 8.2; th. 2.2 Discoid shaped ingot, flat section, halfcomplete.

131. Decorative plaque pl. 87 MB113, US57 h. 1.1; d. 4; th. 0.1 Disc shaped plaque, almost hemispherical, with slightly concave profile.

138. Ingot pl. 87 MB150, US224 d. 10; max th. 2; wg. 386 Composition: Cu 95.9; Fe 0.1 Fragment of a discoid shaped ingot, flat section.

123. Ring or earring pl. 87 MB51, US23 d. 2.5; th. 0.3 Ring with broken open ends, possibly used as an earring. 124. Rings or earrings pl. 87 (A) MB58, US23 A: d. 2; th. 0.3 B: l. 2.6; th. 0.3 C: l. 2.5; th. 0.3 D: d. 2.2; th. 0.2 One fragment and three complete rings with rounded open ends, possibly used as earrings. 125. Ring or earring pl. 87 MB77, US1 d. 2.1; th. 0.2 Ring with rounded open ends, possibly used as an earring.

Decorative Plaques 126. Decorative plaque pl. 87 MB16, US3 h. 0.4; d. 1.8; th. 0.2 Hemispherical plaque with slightly pleating edges. Probably used for decoration. 127. Decorative plaque MB29, US3 d. 2.4; th. 0.2

pl. 87

132. Decorated plaque pl. 88 MB168, US353 l. 7.1; w. 3.7; d. 9; th. 0.3; wg. 32 Circular fragment of a plaque with one plain side and one decorated one. The thickened rim and a further circular rib delimitate a narrow band decorated with spots in relief. The same kind of decoration is visible outside the band. Possible decorative plaque or mirror.

Ingots 133. Ingot pl. 87 MB106, US67 l. 7.8; th. 2.8; wg. 354 Bun-shaped ingot, half-complete. 134. Ingot MB129, US24

pl. 87

139. Ingot pl. 87 MB152, US224 l. 6.2; th. 3.4; wg. 294 Composition: Cu 94.7; Fe 2.5 Quarter of a discoid shaped ingot, flat section. 140. Ingot pl. 87 MB155, US224 l. 7.8; th. 2.4; wg. 412 Composition: Cu 84.2; Fe 11.4 Discoid shaped ingot, flat section, almost half-complete. 141. Ingot pl. 88 MB156, US224 l. 11.8; w. 8.2; th. 8 Composition: Cu 89.6; Fe 7.3 Fragment of a possible bun-shaped ingot.

322

marzia sasso

142. Ingot pl. 87 MB162, US268 l. 8.2; w. 5.6; th. 2.3; wg. 362 Discoid shaped ingot, flat section, halfcomplete. 143. Ingot not illustrated MB172, US236 l. 6.2; w. 4.5; th. 2.2 Fragment of a possible bun-shaped ingot. 144. Ingot MB187, US370 l. 7.5; w. 6.9; th. 2.6 Bun-shaped ingot.

pl. 88

145. Ingot pl. 87 MB205, US272 l. 6; w. 5.3; th. 2.6 Fragment of a bun-shaped ingot. 146. Ingot pl. 88 MB216, US565 l. 4.7; w. 4; th. 2 Composition: Cu 86.2; Fe 9.7 Fragment of a bun-shaped ingot.

I ron

objects

147. Object pl. 88 MI7, US75 A: l. 5; w. 1.6; th. 1.1 B: l. 5.4; w. 1.8; th. 1.1 Two badly corroded fragments of a possible grip. 148. Blade pl. 87 MI5, US45 l. 6.6; w. 1.3; th. 0.7 Blade fragment of a knife with one sharpened edge and a pointed end. 149. Blade pl. 88 MI17, US394 l. 6.1; w. 1.9; th. 0.4; wg. 7.8 Fragment of a thin rectangular plaque, possibly part of a blade and slightly tapering at one end.

G old

object

150. Bead pl. 87 MG1, US23 d. 0.6 Small spherical bead, smashed in many points.

plates

324

P late 82 – Copper-base snake figurines. Scale 1:2.

marzia sasso

7 . metal objects

P late 83 – Copper-base ladles and miniature axes. Scale 1:2.

-

plates

325

326

marzia sasso

P late 84 – Copper-base objects: cauldron; lid; possible standard (no. 25); arrowheads. Scale 1:2.

7 . metal objects

P late 85 – Copper-base arrowheads and dagger blades. Scale 1:2.

-

plates

327

328

marzia sasso

P late 86 – Copper-base objects: dagger blade; hoes; axe; stud; needles; chisels; burins and rods; clamps. Scale 1:2, except no. 89: scale 1-3.

7 . metal objects

-

plates

329

P late 87 – Copper-base objects: pins/hairpins; rings/earrings; decorative plaques; ingots. Iron blade (no. 148). Gold bead (no. 150). Scale 1:3, except nos 115-116: scale 1:1; nos 117-131: scale 1:2; no. 150: scale 2:1.

330

marzia sasso

P late 88 – Copper-base objects, miscellanea. Scale 1:2, except no. 82: scale 1:3.

8. Faunal remains from Iron Age Husn Salut. An archaeozoological study laura strolin

with a contribution by jacqueline

studer

The project of an extensive archaeozoological research on the faunal remains collected during more than ten years of excavations at Husn Salut was launched in 2016.1 Previously in fact, only partial studies had been carried out, aimed at testing some specifically chosen contexts.2 The present study represents the first step in the study of Iron Age animal remains from the site, in the frame of a wider ongoing research investigating complex topics and issues. Malacofauna has not been included in the study, although it is present at the site in a relatively small amount, including marine shells.3

B one

sample

The analysed faunal remains have been recovered throughout specific stratigraphic sequences that were selected by the archaeologists and that are related to three buildings: the Basement, Building 1, and the Burnt Building. All are located in Area 1, the highest sector of the fortress (see Chapters 2 and 3), and can be best understood as places serving a public function, such as collective gatherings and possibly ceremonial activities (see Chapter 9). The availability of a coherent set of radiocarbon dates (Chapter 10) makes of these sequences the best contexts where to try to distinguish possible diachronic trends in animal use and consumption, apart from a general identification of the remains. The osteological sample covers a chronological range spanning the whole Iron Age occupation of the site, i.e. phases HSI (1300-1050 BC), HSII (1050-650/600 BC) and HSIII (650/600-300 BC); however, the large majority of the evidence is attributed to the early occupation of the site. Husn Salut is located on a rocky outcrop that dominates the plain crossed by Wadi Saifam, close to the town of Bisya, approximately 130 km from the east coast. The climate is arid with seasonal rainfalls (see Chapter 1). Recent palaeobotanical studies4 on vegetal samples from some of the same contexts where the present faunal assemblage was collected, indicate that the ancient environment was arid and shrubby, with vegetation typical of sub-desert areas. The plain is composed of gravelly and sandy soils, the surrounding hills have almost barren rocky slopes.

1 The project is a collaboration between the Italian Mission to Oman of the University of Pisa, directed by Prof. A. Avanzini, and the Natural History Museum of Geneva (MHNG), where a remarkable reference collection has been created over the years. Dr. Jacqueline Studer (Curator of the Archaeozoology Department of the MHNG) has directly contributed the section on fish remains to the present study. 2 G. Rasile (Rasile 2011) accounts for the remains found in Area 4, US297 and US298 (NISP = 1150); an unpublished preliminary report by B. Wilkens (Wilkens n.d.) concerns the material of US75 (NISP = 841) in the Burnt Building. These studies are preliminary and therefore cannot be fully exploited in this research. 3 See Degli Esposti 2015b: 79-80. 4 Bellini et. al. 2011.

332

laura strolin

Methods of collection of faunal material must be considered when interpreting results. Given the massiveness of the archaeological deposits in the area under study, remains were hand collected and no flotation or sieving was carried out. This is likely to introduce a bias on their representation, penalising small species, small fragments and fragile bones. Bones showed an overall better state of preservation than teeth. Indeed, the preservation of teeth was poor and additional fragmentation occurred even after collection. Therefore, it was considered worth spending efforts in reconstituting isolated teeth as well as teeth rows with accuracy, in order to provide sufficient evidence for age estimations based on dental attrition rates. Quantification of remains is expressed in terms of number of identified specimens (NISP)5 and weight (g). The total number of remains in the studied sample is 11536 (total weight 26344 g). 46% of these were identified (NISP 5334; 23925.1 g, corresponding to 91% of the total weight). Most of the unidentified fragments are smaller than 4 cm and in many cases belong to non-diagnostic parts of the bone.6

M ethods The identification and a first description of the fragments were mostly performed in situ with the help of animal anatomy atlases,7 as well as by comparison with a small personal reference collection of local fauna. In a number of cases – in particular concerning birds or fishes – the comparison with the reference collection available at MHNG was of fundamental importance. Identification and discrimination between sheep and goat was based on the morphological and metrical criteria presented or assessed by several authors.8 All these criteria were previously tested on the skeletons of sheep and goats from Sudan, Syria and Europe that are available in the MHNG collection. Wherever a clear distinction was not possible, the term ‘caprines’ is used, inclusive of domestic sheep and goats. For camels, the indications given by M.M.S. Smuts and A.J. Bezuidenhout, and by C. Steiger 9 were used (the latter in particular for discrimination between taxa), together with comparisons with specimens of the local reference collection10 and the MHNG collection11. Gazelles have been identified following J. Peters; N. Munro, G. Bar-Oz and A. Hill,12 as well by comparisons with specimens of the MHNG collection. For birds, criteria followed A. Cohen and D. Serjeantson; Z.M. Bocheński and T. Tomek; P.F. Cassoli.13 Equid identification was based on the criteria by V. Eisenmann; C.P. Groves and D.P. Willoughby; S. Payne,14 as well as on those by H.P. Uerpmann.15 For rodents, the features of the occlusal face of teeth have been identified on the basis of Harrison and Bates; Aulagnier et al.16 Fish identification was undertaken with the help of the modern skeletal collection of the MHNG and the estimation of the size of individuals by comparing the archaeological bones with modern specimens of known standard length (SL). Quantification was performed in terms of NISP and of weight per species in each single layer. With the aim of ap-

5 Which counts every fragment or bone identified to the level of the family, the genus or the species. 6 Recent fragmentation is remarkable, with 17.8% of the NISP showing recent fractures. 7 In particular Schmid 1972 and Barone 1976. 8 Boessneck, Müller and Teichert 1964; Payne 1969; 1985; Prummel and Frisch 1986; Fernandez 2001; Balasse Zeder and Lapham 2010; Zeder and Pilaar 2010; Gillis, Chaix and Vigne 2011. 9 Smuts and Bezuidenhout 1987; Steiger1990. 10 Camelus dromedarius n. CdO.1 and CdO.2. 11 N. 1502.2, 1502.3, 1502.4 and 1502.6 for Camelus dromedarius and 1501.1 for Camelus bactrianus. 12 Peters 1986; Munro, Bar-Oz and Hill 2011. 13 Cohen and Serjeantson 1996; Bocheński and Tomek 2009; Cassoli 1972. 14 Eisenmann 1980; 1986; Groves and Willoughby 1981; Payne 1991; recalled in Baxter 1998. 15 Unpublished lecture notes for which we thank the author. 16 Harrison and Bates 1991; Aulagnier et al. 2009.

and

Ambrose 2005;

8 . faunal remains from iron age husn salut

333

proaching at the best the actual value, the minimal number of individuals (MNI) was estimated taking into account age, size and side (right or left) of the different elements, as well as searching for matching fragments (in particular, in the case of mandibles large sets of elements have been examined). Age at death was estimated on the basis of fusion ages in case of long bones.17 For camels, the fusion age was estimated using the specimens of known age of the MHNG collection and the fusion ages of cattle as a comparison.18 An estimate on the basis of the stages of dental eruption and attrition was also performed.19 The work by J. Studer and the recent compilation by O. Putelat20 were used for age attribution to wear stages in caprines and cattle. In the case of gazelles, the references are S.J. Davis and N. Munro, G. Bar-Oz and A. Stutz.21 For camels, M.F.X. Lesbre was followed, and W. Müller for equids.22 Sex was determined following the criteria given by J. Boessneck, H.H. Müller and M. Teichert for caprines; C. Grigson; H.J. Greenfield for cattle.23 The works by L.K. Horwitz, C. Cope and E. Tchernov and by N. Munro, G. BarOz and A. Hill were used for gazelles.24 Measurements were performed following A. von den Driesch; N. Munro, G. Bar-Oz and A. Hill for gazelles; and A. Cohen and D. Serjeantson for birds.25 Additional measurements on camels were taken according to C. Steiger.26 For equids, measurements follow V. Eisenmann; P.F. Turnbull.27 Height at withers was estimated using the factors given by M. Teichert in the case of sheep, and by Z. Schramm in the case of goats.28

F auna

at

H usn S alut

during the I ron

A ge

Species representation is illustrated in fig. 96, tables 2 and 3. Altogether, the analysed contexts show a diversified faunal spectrum, although from a quantitative point of view it is strikingly dominated by domestic taxa. Indeed, the majority of fragments belongs to domestic species, in particular caprines, that represent 92.5% of the total NISP and 80.4% of the total weight. These taxa doubtlessly constituted the main animal resource in the site in all chronological phases. Their role in the subsistence of human communities can be explained by their primary and secondary products, together with their manageability, and is attested on several sites in the Arabian Peninsula. In many cases, the preservation of diagnostic features made the discrimination between sheep and goats possible. Cattle remains are represented in drastically lower proportion (2.1% of NISP identified remains, 8.1% of the total weight) and show marks related to meat consumption. Dromedary represent 0.9% of the remains in terms of NISP, but 7.9% in terms of weight. Dog is also attested in a small percentage (