RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL 151 IN THE TRENCHES: RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL By Alison L. Gascoigne*, David Thomas** and Fiona Kidd*** *University of Southampton **La Trobe University, Melbourne ***Metropolitan Museum of Art, New York Abstract The Bala Hissar was the royal, military and administrative heart of Kabul for a significant period before it was occupied by British forces during the first two Anglo-Afghan wars in the nineteenth century. Despite its archaeological and historical significance, part of the site continues to function as a military base, an expan- sion of which began in 2007 when nine large holes were bulldozed into the site before protests halted the work. This paper details the findings of an archaeological impact assessment undertaken in July 2007, and incorporates an analysis of satellite images documenting further construction in 2009. The results provide the first explicit archaeological (in particular ceramic) evidence suggesting deep continuity of occupation at the site. The contested ownership and uncertain future of the Bala Hissar in Kabul exemplify the pressures placed on archaeological sites around the world, in the face of uncontrolled development and competing agendas. Keywords Bala Hissar; Kabul; Afghanistan; historic fort; ceramics I. INTRODUCTION Bala Hissar was considerably built up by him and his descendants.2 No explicit archaeological evidence of The Bala Hissar, or High Fort, is the historic heart of activity pre-dating the Mughals in the Bala Hissar has, Kabul, situated at the town’s south-eastern edge on however, been identified to date. Traces of fortifica- a natural rock outcrop rising some 50 m above the tions in the south-west of the site have been attributed surrounding plain (Fig. 1). The site comprises upper (without supporting evidence) to the Hephthalite and lower fortified enclosures and is registered in its period (AD 450–565).3 Pre-Kushan pottery has been entirety as a national historic monument. The Upper identified in Kabul city near the fort, as have Greek Bala Hissar is located on the south side of the fortress, and Achaemenid coins, while the surrounding valley while the lower enclosure is more extensive, cover- contains remnants of the city’s Buddhist, Hindu and ing in the region of 42 ha and containing most of the pre-modern Islamic past, and a historic cemetery, the historically attested buildings as well as, until the late Shuhada-i Salahin (Pious Martyrs).4 The uninves- nineteenth century, a substantial population. The sur- tigated archaeological remains on the Bala Hissar, viving stone and packed-earth fortifications on the site therefore, have the potential to yield a wide variety of are primarily mid-eighteenth to late nineteenth century data about the changing occupation of the site over at in date, but a few fragments of Mughal-era architec- least one and a half millennia. Despite the significance ture survive.1 of the site, its history has only recently been studied in Consensus has it that the Bala Hissar was occu- detail (readers are referred to the excellent synthesis of pied long before the intensive activity initiated by Mughal rulers. The site was clearly in use prior to the 2 Woodburn 2009: 3. 3 Caspani 1946; Ball 1982, I: 136–37; Dupree L. 1980: 302; sixteenth century: the first Mughal emperor Babur had Woodburn 2009: 3; Schinasi 2008: 42; Dupree N.H. 1972: to besiege the fort to gain entrance in 1504, and the 71–72. 4 Dupree N.H. 1977: 82–83; Ball 1982, I: 136–37; Omrani 1 Lee 2009. and Leeming 2005: 571. Iran LI 2013, 151–96 © 2013 The British Institute of Persian Studies 152 ALISON L. GASCOIGNE eT AL. Fig. 1. Locations of the Bala Hissar’s major features, indicated on the 1879 Ordnance Survey map, “Plan and Survey of the Bala Hissar or Fort of Cabul shewing the present state and nature of its defences and pointing out repairs and improvements recommended for its better security”. Photozincographed at the Ordnance Survey Office, Southampton, 1879; from a lithographic copy taken in the Surveyor General’s Office, Calcutta, December 1878; copied from a plan signed by Lieutenant J.L.D. Sturt, Engineer in HM Shah Shooja’s Service, 13 December 1839. (Courtesy of the National Archives, Kew, reference MPHH 1/675. Some of the indicated locations [e.g. 14, 16] are approximate.) the post-mediaeval historical sources by Woodburn) Afghan Ministry of Information and Culture (MoIC) while—as noted above—archaeological investiga- and other cultural organisations successfully lobbied tions have been negligible.5 for the work to be halted. In July 2007, at the invitation Despite legislative protection, a major expansion of the Deputy Minister for Culture, Mr Omar Sultan, of the current military facilities at the Bala Hissar the authors visited the Lower Bala Hissar to undertake began in May 2007, funded at least initially by the US an assessment of the damage caused by the construc- government. This work, which resulted in the digging tion work, in collaboration with colleagues from the of nine large trenches, was undertaken without archae- National Afghan Institute of Archaeology (NAIA).7 ological supervision, in contravention of a 2004 law protecting Afghan cultural heritage.6 A broad spectrum 7 Since our visit, further archaeological work, in the form of of local and international protests followed, and the limited excavations, has been undertaken at the Bala His- sar by the Délégation archéologique française en Afghani- stan in March 2008 (DAFA 2008). This work focused both 5 Woodburn 2009; Omrani and Leeming 2005: 571. on the edge of the higher ground of the Lower Bala His- 6 AKTC 2007b: 2. A translation of the law is available at sar north of trench 2, and in an area outside the walls in www.cemml.colostate.edu/cultural/09476/pdf/afghan- the lower town further north still. This work will no doubt antiquities-law-2004.pdf (accessed 11 Sept. 2012). provide more valuable archaeological information on this RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL 153 Analysis of satellite images available through Google ish forces were billeted in the walled garden (chahar EarthTM has since revealed that further construction bagh) to the south of the palace, which along with work was undertaken in mid-2009, resulting in the its Diwan-i ʿAmm pavilion is visible in photographs creation of a large walled compound in the centre of dating to 1880 (Fig. 1: garden, no. 3; palace, no. 6; the Lower Bala Hissar, some activity in its north-west pavilion, no. 12).9 The royal residence fell into dis- corner and development on the Upper Bala Hissar, repair during this period, and was later moved to the which was inaccessible to the authors and has never outskirts of Kabul by Abdur Rahman Khan (amir of been investigated archaeologically. Afghanistan, 1880–1901). A military academy was In this paper, we attempt to correlate the areas built on the site of the royal palace and court in the impacted by these recent construction programmes late 1930s and faint outlines of its walls can be dis- with the historically documented structures of the Bala cerned on Figure 3. More recently, an uprising against Hissar, and to present and analyse the archaeological Communist rule broke out at the site in 1979, prior to (primarily ceramic) evidence for both recent and earlier the Soviet invasion later that year, and was put down periods of activity on the site that was salvaged from following heavy bombing.10 the trenches during our 2007 visit. Although our field- Most of the non-royal structures in the Bala His- work was necessarily brief due to difficulties of access, sar, including the Armenian quarter, the buildings the results are of particular importance in light of the occupied in 1879 by the ill-fated British Resident Sir current lack of available archaeological data relating to Louis Cavagnari and the domestic and commercial the Bala Hissar in Kabul, and to post-Timurid remains quarters that grew up around the palace and gardens, more widely. The broader issue of the continued use of were obliterated, either accidentally through arsenal a historic fortress as a military base has ramifications for explosions in 1879, wilfully by the occupying Brit- other archaeological sites in Afghanistan, such as those ish forces the following year to clear lines of fire in at Qunduz, Bagram and Herat, as well as further afield. preparation for an attack, or during the violence of The paper also highlights the role of satellite imagery in later, well-documented conflicts.11 In particular, the monitoring impacts upon contested historical sites that fierce fighting in and around the site in the early 1990s are difficult to access for reasons of national security.8 destroyed almost all the extant standing architecture in the Lower Bala Hissar, and left the site pitted with shell holes and scattered with unexploded ordnance II. THE 2007 AND 2009 CONSTRUCTION WORKS: and mines.12 Satellite images show the site of Cavag- HISTORICAL DEVELOPMENT OF THE nari’s Residency, towards the south-eastern side of the AFFECTED AREAS lower fort, as marked with trenches, presumably from recent conflicts.13 With the exception of the remaining The trenches dug by military contractors in 2007 were sections of the massive perimeter walls, therefore, the located in the Lower Bala Hissar, in the centre of the primary surviving historic value of the site resides in plateau, which stretches to the north of the upper fort its subterranean, archaeological remains, which have (Figs. 2–3). This part of the fort was the location of been compromised by the recent construction works, numerous structures of significant historical value. and in its symbolic value as the scene of Afghan resist- The lower enclosure was the setting in which many ance against successive invaders. uneasy, fractious relationships were played out, both among Afghan elites and within the geopolitical arena 9 The evidence for many of the Bala Hissar’s historic struc- of the First and Second Anglo-Afghan Wars. Sir Wil- tures has been collated and analysed by Bill Woodburn and liam Macnaghten, who led the mission that installed will not be revisited in detail here; Woodburn 2009: 5, fig. Shah Shuja on the Afghan throne in 1839 in place of 45. Dost Muhammad, briefly had his residence here. Brit- 10 Woodburn 2009: 41; Vogelsang 2002: 307. 11 Woodburn 2009: 32, 34–37, figs. 44–46; for the develop- ment and situation of the Armenian quarter, see Lee 2002. intriguing site. We are grateful to David Jurie of DAFA for 12 Woodburn 2009: 41–42. sharing information about these excavations with us prior 13 Woodburn 2009: fig. 40 indicates the construction of tanks to its publication. (for water and/or fuel?) in the area of the Residency, which 8 Myers 2010, on the US military base at Guantánamo Bay, must also have impacted significantly upon below-ground Cuba, is another such example. deposits. 154 ALISON L. GASCOIGNE eT AL. Fig. 2. Plan of the construction trenches and spoil heaps recorded by the authors in July 2007, overlain onto part of the 1879 Ordnance Survey map. Fig. 3. Plan of the construction trenches overlain onto a satellite image from 19 July 2007 (copyright 2013 DigitalGlobe Incorporated, Longmont CO, USA). RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL 155 Fig. 4. Sequence of satellite images from 17 June 2004 to 20 November 2009, showing the expansion of military construction in the Lower Bala Hissar (Google EarthTM map data copyright Google and image providers DigitalGlobe [upper images] and GeoEye [lower images]. Bulldozing and backfilling of the trenches took place between the taking of the first and second images). A site visit in July 2007 revealed that the sequence of satellite images available through Google c. 5,300 m2 of construction trenches had been bull- EarthTM subsequently demonstrated that a second pro- dozed to an average depth of 1 m.14 Analysis of a gramme of works commenced at some point between 9 July and 30 August 2009, with the construction 14 Due to the sensitive nature of the site, we were unable to survey the trenches with Total Station or Global Position- and correlating our sketch plans with satellite images of the ing System, relying instead on hand tapes for measurements site by means of readily identifiable landmarks. 156 ALISON L. GASCOIGNE eT AL. of a large compound, apparently housing a missile the walls have associated white plaster surfaces and launcher, covering 12,300 m2 in the centre of the site, are of considerable size, being 1.4 m wide and 0.7 m and an additional 9,950 m2 of gravel being laid to the high. As one would expect, the trenches yielded a east and west of the compound walls (Fig. 4: C–D). significant number of artefacts, including fragments This later construction phase overlies, but appears to of white (and a few pale pink) bathroom-style tiles,18 be otherwise unrelated to, the trenches inspected in yellow-painted wall plaster, and the spout of a metal 2007; it thus represents another major impact upon the teapot.19 Severed electrical cables and a cement pipe archaeological and aesthetic value of the site that has were revealed in section below the surface of trench gone unreported and unmonitored. 4. Although not particularly illuminating in them- We will now consider in turn the most significant selves, the structural remains and associated artefacts historical structures—the Military Academy, palace, salvaged from the bulldozed trenches illustrate the mosque and residential areas—that comparison of potential to reconstruct a detailed plan of the Military historic maps with the satellite images indicates must Academy, and to examine aspects of the use of space have been affected by the recent construction works. within the building. We will also present the architectural remains and material culture documented during our archaeologi- cal impact assessment, in light of historical data. II.2. The royal palace and gardens As noted above, the Military Academy was built over II.1. The Military Academy what had been the dominant structure of the Lower Bala Hissar in the nineteenth century, the royal pal- The civil war of 1929 caused considerable destruc- ace.20 Although the palace may have had its origins tion to Kabul and necessitated rebuilding from 1930 in the Mughal period—the fourth Mughal emperor onwards. Work commenced on a new, modern, Jahangir demolished existing buildings in 1607 to con- Afghan Military Academy in 1933, situated in the struct a palace and audience hall—this royal complex, Lower Bala Hissar on the site of the earlier palace with its associated buildings and gardens, took a more and palace court, and this complex remained in use well-documented form when Timur Shah Durrani (r. from 1938 to 1969.15 Photographs of the Academy as 1773–93) moved the capital from Kandahar to Kabul it neared completion show it to have been a long, two- in 1775. The palace was further renovated by his son storey establishment, with an exercise ground and a Shah Shuja.21 In 1832, British deserter, traveller and large park with young shrubs, and an aerial view of agent Charles Masson wrote that the palace “is most Kabul taken in 1965 shows the Academy, with the area substantially constructed, and the interior is distrib- around it by this time being quite built up.16 The Acad- uted into a variety of handsome and capacious areas, emy structure was not completely destroyed by recent surrounded by suites of apartments on a commodious conflicts as it showed clearly in satellite imagery taken in 2004 (Fig. 4: A), and more faintly, in 2007 (Fig. work at the site of Babylon in Iraq (Curtis 2005). 3), despite being levelled, presumably at the time of 18 The tiles had textile impressions on the back, and one the 2007 construction work. The remnant walls and preserved part of a maker’s mark “…ALIERA…” on its reverse. foundations were cut in many places by trenches 3, 19 A copper-alloy 5 paisa coin dating from AH 1313 (AD 4, 5a, 5b, 6 and 7, and were clearly visible in section 1895–96) was also found, and clearly pre-dates the Acad- (Fig. 5).17 Built of stone and baked brick, several of emy. This coin was produced after the 1890 currency reforms of Abdur Rahman, which made the Kabuli rupee 15 Schinasi 2008: 183–84; Dupree N.H. 1972. the only currency in Afghanistan, and introduced English 16 The photographs in question are for the most part in private minting machines into the country; previous coinage was collections and have not been seen by the authors; they are irregular and hand-struck (Noelle 1997: 398–400). Metal reported by Schinasi 2008: 183–84, pl. 35. finds from the Bala Hissar were conserved by Jane Hamill, 17 By the time of our visit, the southern end of trenches 3 and and all artefacts were deposited with the Afghan National 4 had been levelled and covered with fine gravel. It was Museum in Kabul at the end of the study. not possible to ascertain in the time available whether the 20 Masson 1842, II: 254–58, esp. the sketch on p. 257; Wood- gravel was sterile or whether it contained cultural material burn 2009: 4, 16–22; Schinasi 2008: 43. from elsewhere, as happened during military construction 21 Woodburn 2009: 4, 16–23. RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL 157 Fig. 5. View north along trench 4 towards the north spoil heaps; the walls visible in section are those of the Military Academy. and magnificent scale. These are embellished with taken over by Shah Shuja for the use of his family.25 ornamental carvings, and highly coloured paintings of By the 1870s, the palace had been replaced by a new flowers, fruit, and other devices”.22 In 1836, British structure to the east, Sher Ali Khan’s (r. 1863–66, traveller Godfrey Vigne noted a recent redesign of the 1868–79) palace, and the old building was given over gardens, with the addition of the pavilions at either to less high-status activities: during the second British end.23 By 1839, however, British Lieutenants Nash occupation in 1879–80, it housed Gurkhas.26 and Havelock noted how dilapidated the palace had The Lower Bala Hissar was so run down by the become, and that the collapse of the audience chamber late nineteenth century that Amir Abdur Rahman Khan roof had nearly killed Shah Shuja.24 Barracks built decided to build a new palace elsewhere. Indeed, in that year with British troops in mind were instead according to British diplomat Sir Mortimer Durand, Abdur Rahman deliberately left the shell of Sher Ali’s palace in the Lower Bala Hissar standing in order to contrast it with the splendour of his new residence, but 22 Masson 1842, II: 257–58. For more information on Charles Masson, see Whitterage 1986. Echoes of the appearance of the fate of the palace pre-dating that of Sher Ali is not these paintings may perhaps be found in the beautiful fres- known.27 Abdur Rahman’s physician from 1885–89, Dr coes of flowers and birds that are still preserved in a bath- house in the Upper Bala Hissar, Herat. 25 Woodburn 2009: 19; Kaye 1874, II: 141. 23 Vigne 1843: 164–65. 26 Woodburn 2009: 26, fig. 20. 24 Woodburn 2009: 19; Kaye 1874, II: 141. 27 Cited in Woodburn 2009: 41. 158 ALISON L. GASCOIGNE eT AL. John Gray, noted that the Lower Bala Hissar “is now the spoil-heaps on the northern edge of the plateau. It almost all in ruins or demolished. The gateway stands, seems unlikely that, in this central area at least, much and a part of the old [again presumably Sher Ali’s] pal- of the palace could have survived. ace. This is used as a prison for women, political prison- Something of the style of the fort walls and palace ers, Hazaras, and others. The wall and the moat exist, architecture, although not the palace itself, can be seen and inside, some rough barracks have been built for a in Irish photographer John Burke’s 1880 panoramic few troops”.28 By 1912, structures in the Lower Bala image of the Bala Hissar.34 Any extant structural Hissar had been completely dismantled.29 remains of this northern range, however, are now Faint outlines on a 2004 satellite image available buried by the extensive heaps of soil and rubble that through Google EarthTM may indicate the northern were dug from the recent trenches and dumped over range of the older palace (Fig. 4: A). It was built on the edge of the plateau. The archaeological survival top of the defensive walls of the Lower Bala Hissar, or otherwise of the possible palace walls that can be looking northwards over the countryside. Immediately traced on the satellite images was thus unverifiable; to the south of this structure lay the palace court, for- perhaps the spoil heaps will provide the best defence merly the site of the Dafta Khana, or record office, for whatever remains of this important structure in the a “very gay” building until its demolition by Dost future. Muhammad;30 and the Durbar Khana or audience hall. South of the court lay the palace gardens.31 One area that may retain some archaeological potential, as indi- II.3. The mosque of the amirs cated by the satellite imagery (Fig. 4: A–D), is the site of the raised pavilion located on the south side of the Structures servicing the palace were located around it garden, now situated at the base of the slope of the in the Lower Bala Hissar, but with the declining status Upper Bala Hissar between a modern swimming pool of the site as the court moved elsewhere, the gardens, and an unidentified structure; this pavilion was used as diwan khanas (traditional buildings opening onto a officers’ quarters in 1839.32 central courtyard) and shrines were allowed to decay. Trenches 3, 4, 5a, 5b, 6 and 7, as noted above, cut Many were pulled down by local inhabitants who recy- through substantial recent walls associated with the cled the building materials, particularly the wooden Military Academy. No walls were visible in section roof beams.35 The Royal Mosque, which may have that could be connected with the earlier palace. The originated in the seventeenth century (a mosque in the stone walls of the Military Academy, however, were Lower Bala Hissar was said to have been built by the sunk deeply into the ground, and the construction of sixth Mughal emperor Aurangzeb, r. 1658–1707),36 their foundations (as well as the sheer number of the was protected from demolition, but had become a ruin walls themselves), must have turned over the ground by the time of Masson’s stay in Kabul in the 1830s.37 long before the current works. Substantial or obvious As we have seen elsewhere at the site as a whole, the preservation of Mughal-era buildings should perhaps capricious nature of royal patronage contributed to its not be expected given the predominant use of perish- decline, and it suffered from wilful neglect during the able or reusable materials such as mud or earth and reign of Dost Muhammad, who may have resented (recycled and recyclable) wood; Woodburn has sug- its associations with the Durranis whom he had over- gested that it may have been as easy to demolish and thrown.38 Little architectural information is available build anew as to repair, and the ground level in the fort on this historic mosque, although it appears (probably may have risen as a result.33 Furthermore, the northern in idealised form) in British army surgeon and artist ends of the recent trenches were shallower, with com- James Atkinson’s panoramic sketch of Kabul dating pacted ramps to provide access for the bulldozers to from 1839.39 Later plans and drawings show no sign 28 Gray J.A. 1890–91: 29. 29 Adamec 1985: 324. 34 Khan 2002: 123–25. 30 Masson 1842, II: 256. 35 Woodburn 2009: 31, 41. 31 See Woodburn 2009: 20–23, esp. fig. 28 for an axonometric 36 Woodburn 2009: 4. reconstruction of the palace. 37 Masson 1842, II: 258. 32 Woodburn 2009: 20–22, fig. 27. 38 Woodburn 2009: 22. 33 Woodburn 2009: 7. 39 Reprinted in Woodburn 2009: 24, fig. 29. RESCUE ARCHAEOLOGY AT THE BALA HISSAR, KABUL 159 of the mosque; instead two blocks of military quarters Trench 2 was considerably deeper and less neatly are found in the area, slightly further to the south.40 finished than trench 1, but like trench 1, it had lit- Trench 1, fortunately, appears to have been cut into tle coherent archaeological material other than ashy deposits to the north-east of the Royal Mosque, miss- lenses visible in the east section, although the remnants ing the location of any extant archaeological remains of a stone wall were noted in the east end of the south associated with that structure. Some evidence can be section. The northern, unfinished part of the trench identified, however, for the nature of the mosque’s contained a jumble of baked bricks, bone, ceramics immediate surroundings. The exposed sections in and other debris of uncertain date. To the west, trench the east and south, which are of considerable depth 8 was left unfinished and at the time of inspection in (1.4 m), show little by way of architectural remains 2007 resembled a ploughed field. Large quantities of and the trench appears to be dug into silty fill or ceramics, bone and stone were visible in the disturbed wash. Ashy layers are visible in the south-west, with soil, presumably indicative of the situation of the other comparable lenses in the west and north sections, but trenches prior to their being topped with gravel. A these are not visibly related to a major structure. More small, mortared wall, standing seven courses (0.9 m) significantly, a fragment of a human cranium in the high, was noted in the south-east corner of the trench. south-west section may provide evidence for burials Other features included a possible stone-lined drain. near the mosque, and further pieces of human bone The architecture exposed in these sections appears to were observed on the northern spoil heap. Masson be relatively recent. recorded a strong local belief that the Bala Hissar Additionally, an area in the north-west corner of was once a cemetery, while British political agent the Lower Bala Hissar, just inside the now disappeared Sir Alexander Burnes (killed in Kabul in 1841) noted Derwaza Naqqara Khana, or city gate, has also been “burying-grounds” in or near the Lower Bala Hissar.41 impacted by the recent build-up of military infrastruc- Our observations support the presence of a graveyard ture (Fig. 4: A–D). A public road cutting across the edge in the vicinity of the Royal Mosque. of the site was widened, and inside the military area buildings were erected and a parking area created. What historic structures, if any, may have been impacted by II.4. The artillery yard, bazaar area and town this development is unclear. British plans from the First and Second Anglo-Afghan Wars only note the presence No historically attested structures are known to have of the road running from the palace area to the city gate; been located in the vicinity of trenches 2 and 8, which the military engineers did not map the dense housing might have cut through the edge of the artillery yard that covered this area.44 We were not able to inspect this and into the bazaar area. The Lower Bala Hissar, part of the site during our 2007 visit. outside the palace complex, gradually became more plebeian over time as people dependent upon the royal court for their livelihoods built houses and a bazaar in II.5. The Upper Bala Hissar the vicinity over the following decades. The area must have been fairly densely built up: Burnes, who resided The expansion of construction associated with the in the Lower Bala Hissar for some time, recorded a ongoing use of the site as a military base has not been population of some 5,000 people in the early 1830s.42 restricted to the Lower Bala Hissar. For the upper fort, Lieutenant Henry Durand’s 1839 report on the fortifi- analysis of imagery available through Google EarthTM cations for the British army recommended the clear- also indicates the development of infrastructure over ance of all private dwellings within the Bala Hissar, areas of historical significance during the last decade but this was not carried out until 1879.43 (Fig. 4: A–D). Inevitably, these developments will have impacted upon the historical remains in the area. 40 Woodburn 2009: figs. 44–45. In the late eighteenth and early nineteenth centuries, 41 Masson 1842, II: 253; Burnes 1843: 266; as noted above, the Upper Bala Hissar contained a prison complex, Schinasi records large Sunni cemeteries to the south of the where sons of Timur Shah were imprisoned follow- Bala Hissar, including many places associated with local saints (2008: 56–57). ing a succession dispute after his death in 1793, as 42 Burnes 1834: 56; Schinasi 2008: 43; Woodburn 2009: 5, 22. 43 Durand’s report is quoted in MacGregor 1871: 441. 44 Woodburn 2009: figs. 5, 29, 30, 44. 160 ALISON L. GASCOIGNE eT AL. was deposed ruler Mahmoud Shah in 1803 (who later At first glance, the distribution of sherds indicates escaped).45 A building known as the Kulah-i Firangi or that the most prolific exposed archaeological contexts “European’s hat” was noted by Burnes c. 1832. It was should lie at the western end of the area into which demolished by 1839, apparently because it overlooked the trenches were cut, with trenches 2 and 8 yield- the palace area of the Lower Bala Hissar, and seems to ing the most fragments. Trenches 2 and 8, along have been located at the eastern end of the upper fort.46 with trench 1, also yielded pre-Islamic sherds. These During his 1836 visit, Vigne noted that the Upper western trenches, and the central ones (trenches 4 and Bala Hissar as well as the Lower, was in considerable 5), also contained some early modern ceramic mate- need of repair.47 British army engineer Lieutenant rial, reflecting the known use of the site in Mughal John Leigh Sturt’s 1839 plan indicates ruined walls; times. Trenches 3 to 7 were less productive, with the likewise, in the 1830s Masson described the upper most common ceramic type being white-glazed por- fort as empty and ruinous, but noted the presence of celain tile pieces of twentieth-century date. It should marble sitting platforms on the north side.48 There is be noted, however, that trenches 1, 3 and 4 had been not enough evidence to examine the situation of these cleaned and their bottoms had been compacted and/ particular structures in any detail, but the Upper Bala or overlain with gravel. Consequently, few sherds Hissar is arguably a location in which some evidence were recovered. Ceramics from trenches 5 and 7 were of pre-Mughal activity might be expected, and thus scarce for similar reasons. The distribution of pottery the damage to this area and the lack of archaeological therefore probably reflects the state of the individual investigation are significant. As with the city gate area, trenches at the time that work was stopped as much as we were unable to visit this part of the site. their potential archaeological significance. III. THE CERAMIC EVIDENCE III.1. Recording and analysis of the ceramic assemblage So far we have considered structures known from his- toric sources, in particular those documented by par- Pottery was collected by means of a 100% diagnostics ticipants in the Anglo-Afghan wars of the nineteenth (here defined as anything indicative of a specific form century, although as noted above, the origins of the or identifiable as a particular ware) pick-up policy in Bala Hissar are thought to pre-date this period con- the trenches and around the spoil heaps. This resulted siderably. Investigation of the 2007 trenches identified in a total assemblage of 178 sherds, presented in full the first explicit archaeological indications of early and in the Appendix below. A series of fabric groups was pre-Islamic activity on the site, in the form of ceram- defined, all sherds were catalogued and a sub-sample ics. In addition to recording the locations and sections of 96 sherds was drawn. Where possible sherds were of the trenches, surface pottery was collected on two assigned to broad phases, in order to investigate the his- occasions: during our first, brief visit, a few pieces tory of activity at the Bala Hissar over the long term were picked up from trench 2 and the large northern (Fig. 7). These phases were: “modern” (eighteenth to spoil heaps (NSH), with an additional collection being twenty-first century); “early modern” (fifteenth to seven- made across the whole area by members of NAIA. teenth century); “mediaeval” (eighth to fourteenth cen- During our subsequent visit, sherds were collected sys- tury); and “pre-Islamic” (seventh century and earlier). tematically from each trench (Fig. 6). Although small At 178 sherds (all quantification is by sherd count), and unstratified, this collection provides the means to the size of the assemblage is too small to yield more consider the area in terms of the ceramic assemblages than a preliminary insight into the historic ceramic in use at the site over time and, to the authors’ knowl- traditions of the Kabul area, but in light of the lack edge, represents the first archaeological publication of of available data for recent ceramic production in post-Timurid ceramics in existence. Afghanistan, it is hoped that it may nonetheless rep- resent a useful contribution to future work. In the absence of many published parallels, especially for 45 Vogelsang 2002: 237–40; Schinasi 2008: 42. 46 Woodburn 2009: 5. the most recent phases of activity, the judgements 47 Vigne 1843: 164. presented here must be regarded as subject to ongo- 48 Woodburn 2009: 16.