Jump to ContentJump to Main Navigation
Power, Memory, ArchitectureContested Sites on India's Deccan Plateau, 1300-1600$

Richard M. Eaton and Phillip B. Wagoner

Print publication date: 2013

Print ISBN-13: 9780198092216

Published to Oxford Scholarship Online: April 2014

DOI: 10.1093/acprof:oso/9780198092216.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 23 February 2017

(p.331) Appendix 1 Notes on Method

(p.331) Appendix 1 Notes on Method

Source:
Power, Memory, Architecture
Publisher:
Oxford University Press

Preliminary Considerations and Spatial Research Methods

Three immediate tasks were required before we could pursue our larger goal of elucidating the political roles played by the Deccan’s secondary centres. First, in order to gain a clearer understanding of how secondary urban centres functioned in the sixteenth-century Deccan, we wanted to discern the specific political and economic roles they played, and learn how they were connected with both higher-order cities and lower-level towns and villages. Second, to situate our three main study sites in a larger perspective, we recognized the value of identifying and examining as many other secondary centres as possible. Third, given our interest in frontiers, we needed to develop a typology of interstate boundaries, both to distinguish between more fixed ‘borders’ and less sharply defined frontier zones, and to determine where exactly the different kinds of boundaries lay. We address each of these points in turn.

Secondary centres and their functions: We understand secondary centres as sites that are intermediate, both in size and in number, between the handful of densely populated primary cities and the innumerable small agricultural villages and hamlets that dot the countryside. Secondary centres mediated economically, socially, and politically between the villages within their hinterlands and the capital cities to which they were subordinate. According to an important study by Hiroshi Fukazawa, the `Adil Shahi Sultanate was divided into three different types of administrative divisions: mu`āmala, qal`a, and pargana.1 Fukazawa saw the first two as variants of ‘crown lands’ entrusted to Persian-speaking administrators—called ḥavāldārs—who were directly appointed by the sultan. Parganas, by contrast, were administered by a largely hereditary class of Marathi-and Kannada-speaking officers—desāīs, deshmukhs, kulkarnīs, and so on—who represented a pre-sultanate tradition of revenue administration. Fukazawa based his analysis on some 360 administrative documents that had survived from the `Adil Shahi period, the vast majority of which date from the seventeenth century. Since only twenty-four of those records date to the latter half of the sixteenth century, and none earlier, Fukazawa himself realized that there was insufficient evidence for reconstructing administrative institutions and arrangements in their formative phase.2

Although few paper records from the sixteenth century survive, there does exist another body of contemporary evidence that can shed considerable light on the sultanates’ administrative structure in the earlier periods. This is the rich corpus of historical inscriptions, mostly in Persian, that were carved into stone stelae and fitted into various urban buildings. A systematic review of this epigraphic corpus—greatly aided by the two well-indexed (p.332) topographic lists of inscriptions prepared by Z.A. Desai3—yields a number of suggestive findings. First, sixteenth-century inscriptions refer to only one local administrative unit, the mu`āmala, roughly equivalent to ‘district’.4 With one or two minor exceptions, most mu`āmalas appear to have consisted of a fortified secondary urban centre and the villages within its surrounding territory. Accordingly, the mu`āmala is generally named after the city or fort at its heart. Examples from the sixteenth century include the mu`āmala of Raichur (first mentioned in 1498), and the mu`āmalas of Kondapalli (1524–25), Bijapur (1550), Nusratabad (that is, Sagar, 1561–62), Velha Goa (1570–71), Sholapur (1578–79), and Kalyana (1580).

Second, in sixteenth-century inscriptions the officer in charge of a mu`āmala was never referred to by the term ‘ḥavāldār’ (literally, ‘holder of a position of trust’);5 rather, the term for such an officer is generally nā’ib-i ghaibat, literally, a ‘deputy in [the sultan’s] absence’. This difference in terminology suggests a more patrimonial quality to the administrative arrangements than was the case in the seventeenth century, by which time the bureaucratic element had become more pronounced. The term nā’ib-i ghaibat is associated with mu`āmala in Bijapur, Velha Goa, Sholapur, and Kalyana.6 The fact that one inscription names a certain Dabit Khan as the nā’ib-i ghaibat at Sholapur7 while two others identify him as nā’ib-i ghaibat of the mu`āmala of Sholapur8 suggests that a nā’ib-i ghaibat was the governor of both the city and the district it controlled. Other inscriptions speak of the nā’ib-i ghaibat of the city of Gulbarga9 and of the forts of Yadgir10 and Mudgal.11

Third, many inscriptions mention that their cities were fortified, referring to the ‘citadel’ (ḥiṣār) or ‘fort’ (qal `a), or recording the construction of bastions (burj) and other elements of fortification. These indicate the strategic importance of a mu`āmala as a military centre, something amply confirmed by the archaeological record. One apparent exception that proves the rule is the ‘mu`āmala of Nusratabad town (qaṣba)’, as Sagar was then known. Though fortified with a mud wall in the early Bahmani period, the wall had apparently become breached and ineffective by the sixteenth century. Accordingly, in 1561–62 the governor Mir Muhammad constructed ‘a fort (qal`a) and an auspicous town (qaṣba-i mubārak)’ at the nearby site of Shahapur, evidently to serve the older town’s defensive functions.12

Finally, mu`āmala centres were more than fortified strongholds. Inscriptions at two of them, Kalyana and Shahapur, distinguish between the ḥiṣār or qal`a—that is, the elevated citadel or fort—and the larger town below, designated as qaṣba. In both towns nā’ib-i ghaibats had constructed marketplaces (peth, peint) to promote the commercial livelihood of the town. It therefore seems likely that most secondary cities would have included a sizeable ‘lower town’, or qaṣba, to serve as the primary commercial centre for the surrounding district.

Occupants of the post of nā’ib-i ghaibat evidently held their districts as iqṭā`s, or military service tenures. Such districts were distributed to officers who collected their revenues, using a portion of them to maintain a stipulated number of troops that were available to the sultan on demand. Such officers were also expected to ensure the prosperity of their districts, maintain law and order, and uphold justice. Although the term iqṭā` is not found in sixteenth-century inscriptions, the iqṭā` system appears to have been in use in the Deccan at this time. For we know that fourteenth-century Tughluq officers in the Deccan referred to a holder of an iqṭā` as a mālik al-sharq,13 and also that in the sixteenth-century Deccan the latter term was used interchangeably with nā’ib-i ghaibat.14 It would therefore follow that a mu`āmala, the district that a nā’ib-i ghaibat governed, was understood in the sense of an iqṭā`, that is, a military service tenure. Moreover, sixteenth-century Deccani inscriptions represent the mālik al-sharq as carrying out certain actions that would not be possible unless he held iqṭā` rights. At Bijapur in 1320, and at Raichur in 1498, we find him alienating land revenue or commercial taxes within his district for charitable purposes.15 And at Malliabad in 1513, and at Velha Goa in 1570–71, we find the mālik al-sharq introducing new economic policies that (p.333) are decidedly to the advantage of their districts’ inhabitants.16

We may conclude that secondary centres provided the key economic, social, and political links between the world of the agricultural village and that of the courtly élite in the capital. In effect, secondary urban centres functioned as administrative instruments for converting the productive surplus of the land into political and military power, both for the city’s governor and for the crown, which explains their fortified nature. Not only did they have to provide secure storage for reserves of grain—granaries being one of the most common building types represented in their citadels—but they also had to serve as fortified garrisons for the troops maintained and fed by their governors. These garrisoned troops had two primary functions: to enforce the regular collection of revenue from the surrounding countryside, and to protect the centre from predatory rivals. Despite such precautions, contests for the control of productive secondary centres were very common, especially along the frontiers between neighbouring states. But even centres well within the territory of a given state were open to such contests when that state’s central power was on the wane. This is how Yusuf `Adil Khan, the founder of the `Adil Shahi Sultanate, came to power at the end of the fifteenth century as the Bahmani state was collapsing. As Firishta notes, ‘he wrested many forts from the governors of [the Bahmani Sultan] Mahmud Shah, and subdued all the country from the river Bhima to Bijapur, the inhabitants of which territory submitted to his authority…’17 The implication is clear: the capture of secondary centres enabled control of the surrounding district, and control of the district enabled the submission of its inhabitants.

Identifying and locating the secondary centres: Given the importance of the plateau’s secondary centres, it seemed worthwhile to compile a database of as many of them as possible. Firishta’s comprehensive and detailed Gulshan-i Ibrāhīmī (or Tārīkh-i Firishta) appeared to be the ideal source for the construction of this database. Since neither the Persian text of Firishta nor Briggs’s translation is indexed, we produced our own alphabetical toponym index listing all such places mentioned in the text. Under the name of each site, our annotations listed the page numbers of each reference to it, as well as a summary of the information conveyed and the date of the events described. Firishta’s remarks generally relate to the details of sieges and battles relating to the control of the site in question. In this way, we produced a preliminary list of the names of over 100 fortified secondary cities that figured in Firishta’s text.

A second step in this analysis was to determine the geographic locations of the various sites in order to incorporate them into a spatial database using Geographic Information Systems software (ArcGIS). This we did with the aid of a pre-existing GIS shapefile for populated places in contemporary South Asia.18 As most of the places recorded in Firishta still preserve their sixteenth-century names, they were readily located by searching on these gazetteer shapefiles. In some instances, even though parts of the names had changed—for example, ‘Ramgir’ is today ‘Ramgarh’—the resulting names were close enough to permit ready identification, especially when references to nearby sites offered further corroboration. For a handful of places, the names were completely different (for example, Firishta’s ‘Shahdurg’ is today’s ‘Panhala’), requiring additional research to determine their location. Finally, there was a residuum of sites that still eluded our attempts to locate them with certainty—approximately 15 per cent of all the places mentioned in Firishta. Even so, we managed to identify and locate a total of ninety-six sites, enabling us to create a new GIS shapefile plotting each of the sites and providing associated data in the attribute table (see map in Figure A1.1). The latter included: site category (primary centre, secondary centre, seaport, and so on), the number of times Firishta mentioned the site, the number of times it was besieged, the number of times it changed hands; and the identity of the state that controlled it at even, ten-year intervals (1500, 1510, 1520, … 1600).

Plotting interstate borders and analysing the nature of frontiers: Using our data from Firishta, we next undertook to produce a new set of maps for our (p.334) study area, with three particular aims. The first was to represent political boundaries more systematically than conventional maps, which tend to be impressionistic in rendering political boundaries and seldom indicate their sources. Second, in order to show the shifting nature of political boundaries, our maps are plotted at regular ten-year intervals throughout the course of the sixteenth century.19 And third, our maps aim to distinguish between stable and unstable boundaries.

Having constructed a spatial database and developed an understanding of how fortified secondary centres functioned, we could then prepare our maps. First, using the GIS shapefile constructed from our toponym index, we generated a network of Thiessen polygons around the various fortified sites with a view to projecting visually the approximate area of influence that each site commanded—in other words, the area of the mu`āmala (see map in Figure A1.2). Thiessen polygons offer a generalized hypothetical model for predicting areas of political/military influence around each centre in a non-hierarchical network of sites. Since the sides of each polygon are equidistant from adjacent centres, the borders around any given site effectively define that area to which an army could march from its own centre and arrive before a force marching from an adjacent centre. As the Thiessen model assumes flat, even terrain and the absence of a hierarchy among sites—neither of which holds true for our case—the resulting polygons remain only an approximate model of a mu`āmala’s actual area of control. In theory, we could have modified the polygons to take into account those geographic factors that might affect an army’s speed of movement. But that would have required much more data than was at our disposal.

We made a series of eight separate maps, drawn at ten-year intervals, each showing the approximate territorial extent of each state at that moment (see maps in Figures A1.3A1.10). Recalling Firishta’s remark that control of a fort leads to control of its territory, we assumed that the territorial extent of each state would comprise the combined territories of all the forts under its control. Accordingly, we first displayed each map so that the marker symbols for the secondary centres varied by colour to indicate the particular states that controlled them at that moment. We then drew a subset of the Thiessen polygon lines to distinguish adjacent arrays of differently coloured markers from one another. Once connected, these lines indicated the approximate interstate boundaries separating the various states at that moment. The resulting maps are only approximations, of course, but as models of territorial extent, we believe that they are nonetheless far more accurate and objectively grounded than any offered in previous maps.20 This is because our maps assume that a state’s effective territorial control depends on its control of secondary urban centres. Our maps also attempt to capture the variability of boundaries over time by offering a series of momentary ‘snapshots’ taken at regular intervals.

Finally, we compiled the data from all eight of these maps to produce a composite map revealing three distinct types of territory in the sixteenth-century Deccan (Figure 7.21, p. 274). The light grey areas indicate stable cores that a given state consistently controlled throughout the sixteenth century. The darker grey areas indicate ‘annexed territories’. In other words, having originally belonged to one state, they were at some point annexed by another and remained part of it for the duration of the period. Finally, the variously hatched areas indicate three different grades of ‘unstable frontier zones’. These districts—together with the forts that dominated them—changed hands between states two, three, or even four times over the course of the sixteenth century.

An analysis of this final map revealed several important patterns. First, two qualitatively different kinds of boundaries appeared: the stable, and essentially linear borders that separated contiguous core areas, and the unstable, relatively broad zones that lay between highly contested secondary centres. Second, it revealed that two of our key study sites, Raichur and Kalyana, were contested not in isolation or solely for their own sake, but in the context of broader networks of contested sites in unstable or extremely unstable frontier zones. For example, the fate of Raichur was inextricably linked with that of (p.335)

Appendix 1 Notes on Method

Figure A1.1 Primary and secondary centres mentioned in Firishta.

(p.336)
Appendix 1 Notes on Method

Figure A1.2 Thiessen polygons based on primary and secondary centres.

(p.337)
Appendix 1 Notes on Method

Figure A1.3 Territorial boundaries of Deccan states: 1510.

(p.338)
Appendix 1 Notes on Method

Figure A1.4 Territorial boundaries of Deccan states: 1520.

(p.339)
Appendix 1 Notes on Method

Figure A1.5 Territorial boundaries of Deccan states: 1530.

(p.340)
Appendix 1 Notes on Method

Figure A1.6 Territorial boundaries of Deccan states: 1540.

(p.341)
Appendix 1 Notes on Method

Figure A1.7 Territorial boundaries of Deccan states: 1550.

(p.342)
Appendix 1 Notes on Method

Figure A1.8 Territorial boundaries of Deccan states: 1560.

(p.343)
Appendix 1 Notes on Method

Figure A1.9 Territorial boundaries of Deccan states: 1570.

(p.344)
Appendix 1 Notes on Method

Figure A1.10 Territorial boundaries of Deccan states: 1580.

(p.345) Mudgal, which lay immediately to its west. These were the only forts that changed hands four times during the sixteenth century, and neither fell without the other also falling or surrendering. Third, the map suggested that the nature of the unstable frontier zone around Raichur differed from that around Kalyana. For its part, the Raichur frontier zone lies in the highly fertile tract between the Krishna and Tungabhadra Rivers and straddles a broad political frontier separating the Vijayanagara kingdom from the sultanates of Bijapur and Golconda. In contrast, the Kalyana frontier zone lies at a higher elevation and away from a riverine plain, along the southern flank of a low range of hills running northwest to southeast. It runs between the cores of four states—Bijapur, Ahmadnagar, Golconda, and Bidar—riven by mutual rivalries over territory. Consequently, this frontier zone was not as tightly integrated as the southern one, as the two most contested sites in this zone, Kalyana and Sholapur, were often controlled by different primary centres.

Fieldwork Methods

While helping elucidate sixteenth-century geopolitics, the spatial analysis outlined provided only part of the picture, the other being the archaeological record of the sites themselves. Accordingly, we planned two seasons of fieldwork in order to visit as many sites as possible and to survey and document their urban and architectural remains. We began by selecting forty-eight sites, namely, (1) the three main study sites of Raichur, Kalyana, and Warangal; (2) five primary centres (Bidar, Bijapur, Golconda, Hyderabad, and Vijayanagara); (3) thirty-one secondary centres across the plateau; and (4) nine sites with well-preserved examples of Chalukya temple architecture (see list given later).

We carried out our fieldwork in two seasons of six weeks each in June and July of both 2005 and 2006. Our specific activities at any given site varied depending on the category of the site and the amount and availability of prior documentation. At Warangal, for example, considerable prior documentation enabled us to focus on the buildings and cults patronized by Shitab Khan. On the other hand, neither Kalyana nor Raichur had been thoroughly studied, obliging us to locate and identify as many architectural and archaeological features as possible and to make a detailed photographic record of each one. In addition to the building types that architectural historians have customarily favoured—mosques, temples, tombs, and palaces—we also recorded a wide range of more neglected features, such as moats, fortifications, armories, city gates, tanks, stepwells, roads, pathways, and granaries. For two structures at Kalyana—the main gateway complex and the Raj Mahal palace complex—we commissioned and oversaw the production of measured architectural drawings by a team of Indian architects led by Gunjan Srivastava. We also located and recorded historical inscriptions in Persian, Telugu, and Kannada. Although most of these had already been published in epigraphic publications, being able to record their exact locations proved useful for interpreting them.

Lacking adequate maps for Kalyana and Raichur, we carried out GPS-surveys of both sites, using a Thales Mobile-Mapper and ArcGIS software. Satellite data for the two sites was purchased from Digital Globe and imported into ArcGIS to create a base-map that could be uploaded to the Mobile-Mapper unit. By walking each site and following the circuits of fortifications and other features, we logged and mapped all the primary features to within sub-metre accuracy. This method proved especially efficacious at Raichur, today a bustling district headquarters where much of the town’s innermost stone wall had been dismantled to provide building material for modern structures. By combining the evidence of the satellite imagery, GPS readings, and ground-level photographs, we reconstructed the outlines of the city’s main sixteenth-century features in an ArcGIS mapfile. Since Warangal was already represented by useful site maps published by Michell (1992),21 our map for this site is simply a compilation based on our Digital Globe data and the cultural features shown on his maps.

We carried out more limited exploration and more selective documentation at the other forty-five sites we visited. At the Chalukya temple sites, we (p.346) aimed to refine our understanding of the two regionally distinct variants of the Chalukya style and to gather data relating to sixteenth-century reuse. Fortunately, the five primary centres were well covered in scholarly literature, enabling us to focus our attention on specific buildings and features relevant to our study of the sixteenth-century revival of Chalukya and Kakatiya memory. In contrast, at most of the thirty-one secondary centres that we visited—many of them remote hill forts—we were entering largely uncharted territory. Although occasionally guided by Ghulam Yazdani’s brief but helpful notes in the Annual Reports of the Archaeological Department of Hyderabad, more often than not we were on our own.22 Here we were often torn between two conflicting goals—wanting to document as much as possible, but needing to become familiar with as many sites as possible in the region. In the end, breadth won out over depth, forcing us to be very selective in what we studied at each site. In general, we focussed on (1) gauging the size and overall layout of the fortifications; (2) recording specimens of early modern cannons lying in situ; (3) recording historical inscriptions; and (4) documenting key buildings and features relevant to our primary theme of the interactions between power, memory, and architecture. Despite time constraints, we managed to complete enough GPS surveys at eleven of these secondary fortified centres to map at least their primary circuits of fortifications. This valuable data clarified where our study sites fit within the larger regional site hierarchy. It also enabled us to understand different approaches to designing fortifications in the various regions of the Deccan.

Both before and after each of our field seasons, we carried out text-based research in order to reconstruct the histories of our study sites and to determine the chronology and the motivating factors of the various contestations over them. The primary sources for this analysis included published inscriptions in Persian, Telugu, Kannada, and Sanskrit; roughly contemporary Persian historical works (especially the Gulshan-i Ibrāhīmī [or Tārīkh-i Firishta] of Firishta, the anonymous Tārīkh-i Muḥammad Quṭb Shāh, the Burhān-i ma’āthir of Tabataba, and the Taẕkirat al-mulūk of Rafi` al-Din Shirazi), and the historical prefaces to an assortment of contemporary literary works in Telugu.23

Architectural History and Buildings Archaeology

Our concern in this project lies not only in the initial circumstances in which buildings have been constructed, but also in the changes that those buildings have undergone during their lifetimes. Methodologically, this approach aligns us with the archaeological sub-field known as buildings archaeology. Mainstream architectural history, at least among those specializing in South Asia, has tended to privilege those chains of historical causation that lead up to and culminate in the moment of a building’s creation. The study of a building’s patronage, its intended uses and meanings, the conditions of its production, its impact on the formation of distinct stylistic traditions, or its relationship with past tradition—such themes have dominated the study of South Asia’s architectural history. By contrast, buildings archaeology focusses more on those causal chains that unfold after the moment of initial creation. The buildings archaeologist will typically investigate how the physical fabric of the building has changed over its lifetime, and how these changes can point to new patterns of use, new modes of understanding the building, and even new communities that have used it. This does not mean that, as academic approaches to the study of built landscapes, architectural history and buildings archaeology are mutually incompatible. Nor does it imply that architectural history is somehow irrelevant to the buildings archaeologist. To the contrary, a sound knowledge of architectural history is a pre-requisite for success in any buildings archaeology investigation, for the two disciplinary approaches are closely related and often complement one another in their respective findings.

In practical terms, buildings archaeology relies heavily on the analysis of discontinuities in materials, constructional techniques, and style to reveal a structure’s different ‘stratigraphic units’.24 A buildings (p.347) archaeologist understands that term somewhat differently from how an excavation-based archaeologist understands it. To the former, a ‘stratigraphic unit’ refers to a part of the building that has resulted from a single ‘building action’ and therefore represents a discrete period of construction.25 Stratigraphic units are thus distinguished from different stages or phases of work within an otherwise single, continuous programme of construction.26 One should also distinguish between ‘positive’ stratigraphic units—that is, those that represent additions to a building—and ‘negative’ ones, which represent deletions of certain parts. Analysing the nature of the points of contact between adjacent stratigraphic units (abutments, alignments, misalignments, overlays, insertions, and so on) can disclose the relative sequence of the stratigraphic units, and by extension, of the different periods of construction. Placing the changing character of the building within its proper chronological context allows one to pose higher order questions relating to changes in a building’s use, understanding, and social context.

Unfortunately, buildings archaeology remains little developed in the South Asian context. As a result, the kinds of resources that are often available in other parts of the world are lacking here, for example, reference databases for the dating of samples of mortar (based on chemical composition), wood (based on established dendrochronological sequences), and other materials. Our own application of the methods of buildings archaeology has sometimes been beset with challenges, but the results have nonetheless been encouraging. We believe they have led to many findings that would not otherwise have been attainable.

List of Sites Visited, 2005–06

(all sites are fortified except Gogi, Patancheru, and the Chalukya temples sites)

  1. 1. Main study sites

    • Kalyana

    • Raichur

    • Warangal/Hanamkonda

  2. 2. Primary centres/capital cities

    • Bidar

    • Bijapur

    • Golconda

    • Hyderabad

    • Vijayanagara

  3. 3. Secondary centres

    • Adoni

    • Ausa

    • Balkonda

    • Bankapur

    • Bhongir

    • Bodhan

    • Devarakonda

    • Elgandal

    • Ganjoti (not located)

    • Gogi (`Adil Shahi necropolis)

    • Gulbarga

    • Kandhar

    • Kaulas

    • Kohir (not located)

    • Koyilkonda

    • Medak

    • Mudgal

    • Malkhed

    • Malliabad

    • Manvi

    • Naldurg

    • Parenda

    • Patancheru (suburb of Golconda-Hyderabad)

    • Racakonda

    • Sagar

    • Shahpur (Andhra Pradesh)

    • Shahapur (Karnataka)

    • Sholapur

    • Talikota

    • Udgir

    • Yadgir

  4. 4. Chalukya temple sites

    • Arisibidu

    • Bagali

    • Chaudadampura (p.348)

    • Harpanahalli

    • Hirehadagali

    • Jalsingi

    • Kuruvatti

    • Narayanpur

    • Umapur

Additional sites figuring in the study, visited by one or the other of the authors independently:

  • Ahmadnagar (E 2001)

  • Bankapur (E 2010)

  • Delhi/Tughluqabad (E 2001; W 1999)

  • Devagiri/Daulatabad (E 2001; W 2000)

  • Firuzabad (E 2001; W 2000)

  • Istanbul (E 2011)

  • Pillalamarri (W 1983)

  • Rajahmundry (W 2000)

Notes:

(1) . Hiroshi Fukazawa, ‘The Local Administration of the Adilshahi Sultanate (1489–1686)’, in his The Medieval Deccan: Peasants, Social Systems, and States, Sixteenth to Eighteenth Centuries (Delhi: Oxford University Press, 1991), 1–48.

(2) . Ibid., 6.

(3) . Ziyaud-Din A. Desai, A Topographical List of Arabic, Persian, and Urdu Inscriptions of South India (New Delhi: Indian Council of Historical Research, 1989); and idem., Arabic, Persian, and Urdu Inscriptions of West India: A Topographical List (New Delhi: Sundeep Prakashan, 1999). We used the indices of these two indispensible works to identify all occurrences of the relevant administrative terms such as mu`āmala, qal`a, pargana, ḥavāldār, nā’ib-i ghaibat, mālik al-sharq, and their variant forms. We then consulted the original inscriptions, for the most part published in the series Epigraphia Indo-Moslemica (1907–50) and its successor, Epigraphia Indica: Arabic and Persian Supplement (1951–76). In the notes that follow, we refer to the inscriptions by their serial number in Desai’s two Topographical Lists (prefixing the number with ‘S’ for the South India volume and ‘W’ for the West India volume).

(4) . Parganas are mentioned in only two inscriptions dating from the sixteenth century. One refers to pargana Raibag, in the present-day Dharwar district, Karnataka (W1346, dated 1555), and the other to pargana Potwaram, in the present-day Medak district, Andhra Pradesh (S1283, dating to the sixteenth century). Similarly, the officers associated with the parganas in seventeenth-century sources rarely appear in the sixteenth-century epigraphic corpus. Thus, desāī appears only in two inscriptions. In one, the context suggests that the desāīs were officers in a mu`āmala who were subordinate to the nā’ib-i ghaibat (W2136, dated 1570–1); in the other, the desāīs appear to have been the officers in charge of a pargana (W1346; dated 1555).

(5) . The earliest firm epigraphic occurrence of the title ḥavāldār appears in an inscription from Siruguppa dated 1627 (S1521). In fact, it refers not to an ordinary ḥavāldār, but to a sar-ḥavāldār (‘head-havaldar’) of four distinct administrative units, each of which happens to be a distinct kind, namely, the mu`āmala of Raichur, the qal`a of Adoni, the mukhāṣa of Siruguppa, and the qaṣba of Muhammadnur (that is, Kurnool). An even earlier inscription from Raichur dated 1609 (S1377) refers to the root term ḥavāla (‘custody’), from which ḥavāldār is derived. By the 1640s, the term ḥavāldār had gained widespread usage in place of nā’ib-i ghaibat.

(6) . Bijapur: S250 (1559); Velha Goa: W2136 (1570–71); Sholapur: W1984 and W1987 (1578–79 and 1580–81); Kalyana: S942 (1580).

(7) . W1986 (c.1580).

(8) . W1984 (1578–79) and W1987 (1580–81).

(9) . S581 (1574–75) and S582 (1586), both reading nā’ib-i ghaibat-i shahr-i Aḥsanābād (that is, Gulbarga).

(10) . S1634 (1546), recording the construction of a tank ‘under the supervision (kār-kard) of Amrit Khan, the nā’ib-i ghaibat’; and S1635 (1556), recording the completion of a well ‘under the supervision (kār-kard) of Malik `Ali Beg, entitled `Ali Khan, the nā’ib-i ghaibat’.

(11) . S1240 (1548).

(12) . S1490 (1561–62).

(13) . An inscription from Kalyana records the construction of a mosque during the reign of the sultan of Delhi, Muhammad bin Tughluq, ‘at a time when the iqta` was held by Malik ash-Sharq Qiwam al-Daulat wa’l-Din, Vazir of the province (iqlīm) of Deogir’ (S934, dated 1326).

(14) . The Bahmanis used only the title mālik al-sharq to refer to a governor of a mu`āmalas. The term ‘nā’ib-i ghaibat’ did not appear until 1512, when the Bahmani Sultanate was devolving into independent successor states. The last documented instance of the title mālik (p.349) al-sharq being used independently is the following year, in 1513. After that, it was never again employed by itself, although it continued to be used in conjunction with the newer title nā’ib-i ghaibat throughout the rest of the sixteenth century. For mālik al-sharq as an independent title, see W525 (1271), S230 (1320–21), S934 (1326), W801 (1332–33), S563 (1380), S1076 (1470–71), S1214 (fifteenth century), S1352 (1498–99), and S1213 (1513). For mālik al-sharq used together with nā’ib-i ghaibat, see S940 (1568), W2136 (1570–71), and S942 (1580).

(15) . At Bijapur in 1320–21, the mālik al-mulūk al-sharq Karim al-Daulat wa’l-Din, patron of the Karim al-Din mosque, gave a grant of the revenue from 24 nitāns of land to the builder of his mosque (S230). At Raichur in 1498–99, the mālik al-sharq Malik `Ambar endowed a mosque by granting it the revenue from 9 chāwars of land in the village of Gabur within his mu`āmala, as well as the taxes collected from forty shops in Raichur itself (S1352).

(16) . At Malliabad in 1513, mālik al-sharq Malik `Ambar issued a qaul-nāma setting lower rates of taxation both on agricultural produce and on commercial taxes (S1213). At Velha Goa in 1570–71, mālik al-sharq Malik Zahid Baig, who was the nā’ib-i ghaibat of the mu`āmala, promulgated an order prohibiting the practice of niputrik, whereby the state assumed a deceased person’s personal property if he had no male issue to inherit it (W2136).

(17) . Briggs, trans., History of the Rise, 3: 4–5.

(18) . A ‘shapefile’ is a simple GIS file type that displays geographic features as points, lines, or polygons according to their spatial coordinates, and which links these features to attribute tables recording various kinds of data associated with them.

(19) . Even the best maps have not always dealt adequately with the dimension of time in delineating territorial boundaries. Joseph Schwartzberg’s A Historical Atlas of South Asia (New York: Oxford University Press, 1992) moves in this direction, but even this authoritative source tends to represent territorial boundaries over periods of as long as a century or more. See, for example, Schwartzberg, Atlas, map Plates V.4.d and e, ‘Major states of South India 1390–1485’ and ‘Major States of South India 1485–1605’.

(20) . These maps have two particular limitations. First, because Firishta resided in the court of Bijapur, he refers mostly to places located in the core region of the Deccan, between the Godavari and Krishna–Tungabhadra Rivers. Beyond these rivers, the number of sites he mentions falls off rapidly. Second, there is the problem of edge effects, that is, ‘the tendency for spatial operations to return incorrect values near the edge of a map, often because the neighbourhood used to calculate them has been artificially truncated’. This is most noticeable along the northeastern and southern edges of the map. Because there are no outlying points to limit the outer boundaries of the Thiessen polygons along these edges, the polygons generated there do not accurately represent the sites’ zones of control or mu`āmalas. See James Conolly and Mark Lake, Geographical Information Systems in Archaeology (Cambridge: Cambridge University Press, 2006), 292.

(21) . George Michell, ‘The City as Cosmogram’, with an appendix by Phillip B. Wagoner, South Asian Studies 8, no. 1 (1992): 1–18.

(22) . At two of the secondary sites we visited—Ganjoti and Kohir—we were unable to locate any traces of a late-medieval/early modern settlement, despite searching the area and making inquiries with the local residents.

(23) . These included the Padya-Bāla-Bhāgavatamu of Doneru Konerunatha-kavi, the Vasucaritramu of Bhattu Murti, the Rāmarājīyamu of Andugula Vengakavi, the Yayāti Caritramu of Ponnikanti Telaganarya, and the Tapati Samvaraṇamu of Addanki Gangadhara-kavi, among others.

(24) . This formulation is based on Iain Stuart, ‘The Analysis of Bricks from Archaeological Sites in Australia’, Australasian Historical Archaeology 23 (2005), 81.

(26) . Warwick Rodwell, The Archaeology of the English Church (London: B.T. Batsford, 1981), 68 and 126–7.