Subordinate rulers under the P–alas: Their diverse origins and shifting power relation with the king

Introduction

No matter how we call it, feudal, integrative or otherwise, the political formation of the early medieval North India was characterised by the presence of subordinate rulers generally called sāmantas. ¹ With their own territories, these subordinate rulers were indispensable components of a monarchical state of the period. They served the king in both administrative and military capacities, with different degrees of autonomy.

Subordinate rulers were conspicuous in the early medieval history of Bengal. Since their first appearance, they had been a regular feature of the political structure of the region and their agency and power relations with the kingship had critical bearings on the historical process of agrarian expansion and regional state formation. The same can be said of the time of the Pālas, who established a regional kingdom with firm control over northern and western Bengal and eastern Bihar, and made occasional but extensive expeditions to neighbouring regions and beyond. ² As the Pālas monopolised the issue of copper plate grants, the number of inscriptions left by subordinate rulers which record their deeds is rather small. However, careful reading of those inscriptions and the Pāla grants can still shed light on the subordinate rulers and their relation with the king. Furthermore, the Rāmacarita of Sandhyākaranandin, a Sanskrit kāvya depicting deeds of both Rāma and Rāmapāla in double entendre, and its near-contemporary commentary give us precious information on those aspects in the last phase of the Pāla rule and enable us to discern their dynamism. In the present study, I would like to discuss the characteristics of subordinate rulers under the Pālas and the shifting power relation between them and the kings through the analysis of information provided by those sources. Before that, I briefly outline the early history of subordinate rulers in Bengal.

The earliest references to subordinate rulers in Bengal are found in the two copper plate inscriptions of Vainyagupta, who ruled the eastern sub-region of Samataṭa in the early sixth century. ³ According to them, subordinate rulers fulfilled administrative functions under the kingships of Nāthacandra, a king of the early fifth century whose grant was copied on the later plate, and Vainyagupta, which grew from subordination to semi-independence under the Gupta suzerainty in the fifth and sixth centuries. ⁴ They also showed their presence as functionaries in the sovereign sub-regional kingdoms which emerged in the mid-sixth century in Vaṅga, Rāḍha and Puṇḍravardhana, respectively the southern, western and northern sub-regions. ⁵

As the state enhanced its control over rural society in the following century, the power of the subordinate rulers who constituted a part of its ruling bloc also grew in proportion and their organisation into a hierarchical order proceeded under the kings. Some of them like Lokanātha and Śrīdhāraṇarāta who ruled the periphery of Samataṭa were strong enough to defy the authority of their overlord and issued copper plate inscriptions practically on their own. ⁶ However, they still had to depend on the royal authority for their position and deal with their own subordinates who aspired to improve local control and resource base through negotiations with them. Thus, the political hierarchy in the periphery of Eastern Bengal in this period was replete with tensions among different levels of rulers. ⁷

In contrast, the contemporary kingships based on the developed agrarian tracts like the Khaḍgas ruling Samataṭa and eastern part of Vaṅga had stronger power with which they issued full-fledged royal grants with monopoly. Subordinate rulers were incorporated into the stratified land relation where the king wielded overarching authority, though they still kept their latent agency. ⁸ The Devas, who supplanted the Khaḍgas and ruled Samataṭa in the eighth century, had the established authority monopolising the issue of land grants and wielding the high-sounding titles of paramabhaṭṭāraka, parameśvara and mahārājādhirāja. Under the domination of those dynasties, subordinate rulers tried to extend their interest and resource base by legitimately encroaching upon the king’s authority and territory through the application for the donation to religious institutions in a stipulated manner. ⁹

The Pālas, who started their career as sovereign rulers contemporary with the Devas in northern sub-region of Varendra, wielded the level of authority similar to or even stronger than that of the Devas. They monopolised the issue of copper plate grants and held the same set of royal titles. ¹⁰ Under the Pāla rule, subordinate rulers were firmly incorporated into the administrative apparatus as its components, as shown by the addressees of grants which include rāja, rājanaka, rājaputra, mahāsāmanta and so on. ¹¹ The agency of subordinate rulers is, however, still detectable in the royal grants issued on their applications. The minute information on their genealogy and deeds contained in the Pāla grants and the other contemporary inscriptions gives us clues to their characteristics and relations with the Pāla kings. In the next section, I will discuss the characteristics of subordinate rulers recorded in those inscriptions with emphasis on their origins.

Origins and Characteristics of Subordinate Rulers

The inscriptions related to the activities of sāmantas mostly pertain to Varendra in the early phase of the Pāla rule. The genealogies of sāmantas included in these inscriptions show their diverse origins. The Indian Museum grant of Dharmapāla contains the genealogy of mahāsāmanta Bhadraṇāga with references to the three generations of his forefathers. Balanāga, his great grandfather, is mentioned as sārthavāha, a caravan trader, while Uccaganāga, his father and the son of Nāvranāga, is called mahāsāmanta. ¹² Bhadraṇāga, the son of Uccaganāga and Ahaṇā, is acclaimed for his valour in battle field. ¹³ This is the case of a merchant family having converted themselves to that of sāmantas.

The Bangladesh National Museum inscription of Pāhila records the construction of a Vaiṣṇava maṭha by this subordinate ruler of Devapāla with his genealogy, which mentions the progenitor and the three generations including Pāhila himself and his immediate predecessors. The progenitor Pañcahūti is described as a ruler (adhipa) of Bhaṭṭala maṇḍala, which is locatable to the southeastern corner of Varendra according to the provenance of the inscription. ¹⁴ Karkarāja, born in his family, is described as a chief (nāyaka) of Bhaṭṭala deśa and a friend of king Yaśovarman, who seems to be identical with the famous king of Kānyakubja known for his expedition to eastern India in the first half of the eighth century. ¹⁵ While the position of his son Śivarudra is not clearly mentioned, Pāhila, the son of the latter born to Dottebhaṭṭārikā, is said to have been burdened with the kingdom by Devapāla. ¹⁶ The titles adhipa and nāyaka wielded by Pañcahūti and Karkarāja vaguely indicate their status as local chiefs ruling Bhaṭṭala maṇḍala or deśa. This is the case of a family of local chieftains which had kept their hereditary territory while switching their allegiance from Yaśovarman to the Pālas.

The genealogy of mahāsenāpati Vajradeva in the Jagajjibanpur plate of Mahendrapāla starts with Devaradeva, his grandfather, who is said to have been born in the great family with no clear reference to his position. ¹⁷ His son Nārāyaṇadeva was appointed to a ruler (adhipati) of Dardaraṇḍī maṇḍala by king Dharmapāla. ¹⁸ Vajradeva, his son born to Kalyāṇadevī, is described as a strong warrior and his position as a general of Mahendrapāla is clear from his title mentioned in another part of the plate. ¹⁹ This can be a case in which a local notable family acquired the position of subordinate ruler through designation by the Pāla king and improved their position through military service.

The genealogy of mahāsainyapati Kokkāka in the Mohipur grant of Gopāla II first mentions his family as a sāmanta family (vaṃśa) which was a carrier of the burden of the kingdom of the Samataṭa kings. ²⁰ It had served the kings of Samataṭa at least from the generation of the great grandfather of Kokkāka, for Cāmuṇḍadatta, his grandfather, is said to have been born in this family. ²¹ In view of the chronology of the kings of Samataṭa, their overlord might have been either the Khaḍgas or the Devas, who flourished from the second half of the seventh to the end of the eighth century. Cāmuṇḍadatta had a son named Vapyadatta who married Cundādevī, daughter of Pārāśadatta. ²² Kokkāka was born to them. The relation with the Pāla kings seems to have started with him, according to the description of his military service to Śūrapāla. ²³ A phrase that the king bestowed on him a blazing golden cudgel (ujvalahemadaṇḍa) may refer to his nomination to mahāsainyapati. ²⁴ This is the case of a sāmanta family based in Samataṭa having shifted to Varendra for some reason and become a subordinate of the Pālas.

The Badal pillar inscription of Guravamiśra, which belongs to the end of the ninth century or the beginning of the tenth century, records the deeds of a brāhmaṇa family of which five generations were associated with the six Pāla kings. ²⁵ They served the kings mostly as political advisors and occasionally as priests or generals. Garga, the first member serving the Pāla kings, is said to have made Dharmapāla the lord of all the quarters. ²⁶ By the policy of Darbhapāṇi, his son, Devapāla is claimed to have made the whole earth tribute-offering (karada), that is, subordinate. ²⁷ The king is even said to have come to the gate of his residence for consultation disregarding proper time. ²⁸ He was also offered a prestigious seat by a king identifiable as Mahendrapāla. ²⁹ His son Someśvara is described as a friend of parameśvara denoting probably the same king. ³⁰ Gauḍeśvara, who may also be Mahendrapāla, is said to have held the earth with the lineage of Utkala uprooted, the arrogance of Hūṇas removed, the pride of Draviḍa and Gūrjara kings suppressed, after paying attention to the intelligence of Kedāramiśra, the son of Someśvara. ³¹ It is said that Śūrapāla attended sacrifices held by him. ³² His son Guravamiśra’s association with Gopāla II is implied in the expression ‘causing pleasure to Gopāla’. ³³ Nārāyaṇapāla highly estimated him. ³⁴ Apart from being credited with eloquence and knowledge of āgama, policy, astrology and itihāsapurāṇa, he is also described as skilful in both discussion and war. ³⁵ He also worked as a messenger (dūtaka) in the Bhagalpur grant of Nārāyaṇapāla. ³⁶ This is the case of a brāhmaṇa family having acquired their position as strong subordinate rulers in Varendra through their long-term association with and service to generations of the Pāla kings.

The Bhaturiya stone inscription of Yaśodāsa, which mentions Rājyapāla and belongs to the mid-tenth century, contains a genealogy of his family. It first mentions a place called Aṭṭamūla, a new settlement derived from Bṛhaddhaṭṭa, as an original place (janmabhū) of the Dāsas. ³⁷ Then it mentions Malhadāsa born in this family and his son Śūradāsa known as an abode of heroism. ³⁸ It is said that the latter’s son Saṅghadāsa, whose generosity is poetically described, married Sarasvatī, daughter of Sūryakuṇḍa and Dūrvāyī. ³⁹ Yaśodāsa was born to them and king Rājyapāla made him his councillor (mantrin). ⁴⁰ He was promoted to minister (saciva) and then to tantrādhikārin by the same king. ⁴¹ After the description of submission of various ethnic groups to the king during Yaśodāsa’s tenure, ⁴² it is said that his fame spread all over the world by the excavation of ponds, Vedic sacrifices, construction of temples and maṭhas, knowledge and charitable feedings (sattras) and so on. ⁴³ There is nothing to suggest that they were brāhmaṇa family, while Yaśodāsa was appointed to the positions which required intellectual qualifications and were often occupied by brāhmaṇas. Therefore, the Dāsas may have been a family of non-brāhmaṇa literates located in rural society. Their rise may have accrued from the marriage alliance with Sūryakuṇḍa, their association with the Pāla king and the religious and public activities in the locality. The case of Dakkadāsa, who is the engraver of the Mirzapur plate of Śūrapāla and the Mohipur plate of Gopāla II and prefixed with the title sāmanta in the former, ⁴⁴ can also be categorised as that of a non-brāhmaṇa literate having acquired the status of a subordinate ruler through his service to the king.

The Rāmacarita refers to some more cases of sāmantas under the Pālas. Prominent among them were the Rāṣṭrakūṭas of Aṅga, of whom Mahaṇa, the maternal uncle of Rāmapāla, mahāmāṇḍalika Kāhnaradeva and Suvarṇadeva, his sons, and mahāpratīhāra Śivarājadeva, the son of his brother are mentioned as the most reliable allies of the king. ⁴⁵ Candradeva, the son of Suvarṇadeva, also served Madanapāla as maṇḍalādhipati. ⁴⁶ They seem to have belonged to a collateral branch of the Rāṣṭrakūṭas which settled in Aṅga, a sub-region of Bihar around the present district of Bhagalpur, and served the Pāla kings. This is the case of a scion of the eminent royal family having shifted to the area ruled by another dynasty and become its subordinate rulers, comparable with the cases in contemporary western India of rājpūt clan members serving the neighbouring kings of different clans, especially whom their sisters married. ⁴⁷

The other category of subordinate rulers mentioned in the same text is chiefs of forest regions. The near-contemporary commentary of the Rāmacarita explains āṭavikas, whose help Rāmapāla asked for, as aṭavīyasāmantas. ⁴⁸ Some of the sāmantas who constituted Rāmapāla’s army are also described as such in the same text. ⁴⁹ Kaivarta chiefs, who led the rebel sāmantas against the Pāla kings, may have had a similar social background of tribal chieftains heading peasantised fisher-folks or boatmen settled in a particular area of Varendra. ⁵⁰

The cases discussed above show that subordinate rulers under the Pālas were a composite group with various origins, which included merchants, local chiefs and notables, brāhmaṇa and non-brāhmaṇa literates, members of the other royal lineage and forest chiefs. Whatever were their origins, they got the position of subordinate rulers through their association with the Pāla kings in the form of military or other services and the appointment to administrative positions. Their commonality as subordinate rulers lay in their control over particular territories, which enabled them to make religious endowments from their own assets as recorded in the inscriptions mentioned in the present and next sections. By their association with the king, those holding their own territory as local chiefs like the family of Pāhila were guaranteed of their control over it and the others appointed to administrative positions like the lineage of Vajradeva acquired new territory. In either case, their autonomy was limited under the established state control. In spite of that, they had their own agenda and stakes in rural society which resulted in the tension and negotiation between the king and them, as discernible in their activities recorded in their own inscriptions and the royal grants concerning donations petitioned by them.

Relation with the King: Negotiations and Changing Patterns of Donations

The tension and negotiation between those subordinate rulers and the king can be detected in one of their activities, namely, the construction of religious institutions often followed by the petition for donation of land plots or villages. As mentioned above, the inscription of Pāhila records his construction of a Vaiṣṇava maṭha. ⁵¹ In the Khalimpur grant, mahāsāmantādhipati Nārāyaṇavarman petitioned Dharmapāla to give four villages to the deity Nannanārāyaṇabhaṭṭāraka, who was installed by Nārāyaṇavarman at the temple constructed by him at Śubhasthalī, and to the attendants of the deity (pādamūla) like Lāṭa brāhmaṇa priests protecting him, for the worship and attendance. ⁵² The four donated villages belonged to the two viṣayas in Vyāghrataṭī maṇḍala of Puṇḍravardhana bhukti. ⁵³ This maṇḍala was probably the territory of Nārāyaṇavarman, as Balavarman, who mediated the petition by king Bālaputradeva of Suvarṇadvīpa as a messenger and was called Vyāghrataṭīmaṇḍalādhipati in the Nalanda plate of Devapāla, seems to have been his son. ⁵⁴ Thus this is the case of a subordinate ruler asking for the donation of villages within his own territory to the religious institution established by him, probably in the same territory.

In similar terms, the Jagajjibanpur and Mohipur plates record the donations to Buddhist vihāras constructed by generals. In the former, Vajradeva petitioned Mahendrapāla to give Nandadīrghika udraṅga to the Buddhist vihāra constructed by him in the same locality. ⁵⁵ In the latter, Kokkāka petitioned Gopāla II to give village Kaṅkāvāsaka to the vihāra constructed by him in the same village. ⁵⁶ The construction of vihāras in each locality points to its control by these subordinate rulers and their resources accruing from it, which enabled such an enterprise. The establishment of religious institutions with personal connection and the donation of villages to them may have contributed to the extension of local influence of the subordinate rulers.

Implication of their action is, however, not limited to this. There is an additional clause to the purposes of donation listed in these documents including worship of the Buddha and other divinities, copying of scriptures, provision to the bhikṣusaṃgha and repairs of buildings. ⁵⁷ It says that others wished by the petitioner would also partake in the donation, in the ratio decided by him, for complete usufruct (anavadyabhoga, literally ‘unobjectionable enjoyment’). ⁵⁸ This clause alludes to the involvement of petitioners in the management of donated property and their following influence over it. In view of this, the action of these subordinate rulers can be interpreted as an attempt to secure for the properties under their influence the status of śāsana land, immune to revenue charges and other interferences by the king. The case of the Khalimpur plate can also be understood in the same manner. While accepting claim of the king over land in his territory and exclusive right of its disposal, subordinate rulers could legitimately encroach upon his power and enhance their own through the construction of religious institutions and the petition for grants. We can detect here their negotiation with the king.

The relation and power equation between the king and his subordinate rulers were complicated and diverse. The case of the Indian Museum grant of Dharmapāla, so far the earliest, rather shows the subtlety of their relation. In this record, mahāsāmanta Bhadraṇāga petitioned for donation of several land plots to a vihāra constructed by him in Antarāvanikā and a perfume chamber (gandhakuṭī) and a vihārikā established respectively by him and his wife Saṇhāyikā at Somapura mahāvihāra. ⁵⁹ As far as the first vihāra is concerned, this case is similar to the cases of Jagajjibanpur and Mohipur plates mentioned above. Bhadraṇāga established a vihāra in village Antarāvanikā which included in its premise the two land plots respectively belonging to his vihāra and vihārikā. ⁶⁰ This is the case of a sāmanta enhancing his local control through the establishment of a religious institution within his own territory and legitimately encroaching upon the royal power through the petition for grant.

The constructions of facilities in Somapura mahāvihāra, however, indicate something else. The involvement of Dharmapāla in the establishment of this mahāvihāra is attested by the sealings excavated from the site of Paharpur. ⁶¹ The contribution of some facilities to it indicates the intention of Bhadraṇāga to cultivate a close connection with Dharmapāla, his overlord, through this act. It is endorsed by the inclusion of the king among those the increase of whose merit is said to be the purpose of the donation, a feature not encountered in the other inscriptions. ⁶² The land plots donated to these facilities were most probably located within the territory of Bhadraṇāga, though the portion mentioning them is illegible due to corrosion. ⁶³ The presence of these plots belonging to the eminent religious institution may have enhanced his authority in the locality by showing his connection with it and its founder Dharmapāla.

Thus we can detect in this case the two different manoeuvres executed by a subordinate ruler to enhance his power in the locality. He tried to enhance his authority and resource base by encroaching upon royal authority, while strengthening his tie with the king at the same time. This contradictory behaviour indicates a complicated power relation in which Bhadraṇāga was located. He owed his position and authority to the king, while he needed to extend control and resource base in his own territory. Though dependence on the royal power was the same for his counterparts appearing in the contemporary and slightly later Pāla grants mentioned above, they seem to have had better position in which they could concentrate on one kind of manoeuvre, namely, the establishment of religious institutions within their own territories.

The cases in the later inscriptions allude to the attempt of the kings to reclaim their power in the negotiation with subordinate rulers. In the Bhagalpur plate, king Nārāyaṇapāla donated a village to the temple of Śiva established by him and to the assembly of scholars of the Pāśupata sect. ⁶⁴ What should be noted is an additional clause to the purposes of donation, which prescribes complete usufruct for others wished by the king in the ratio fixed by him. ⁶⁵ It is almost identical with the similar clauses found in the Jagajjibanpur and Mohipur plates and indicates the involvement of the king in management of the institution and its property. By adopting the practice initiated by subordinate rulers, the king seems to have tried to counter their attempt to encroach upon his power. It later developed into the royal patronage of Brahmanical religious institutions with which the kings had personal connections. It became conspicuous in the eleventh century. Mahīpāla I established in Koṭivarṣa a maṭha of Śaiva ascetic order which later provided rājagurus to Nayapāla, his son, ⁶⁶ who seems to have established Śaiva temples in several localities of Bengal and Bihar. ⁶⁷

The effort of the king to block the encroachment by sāmantas is clearer in the case of the Bhaturiya stone inscription. According to it, Yaśodāsa constructed a Śaiva temple complex and installed a liṅga at the main high building (prāsāda) surrounded by eight shrines (suramandira). ⁶⁸ Then king Rājyapāla gave village Madhusrava to the deity after withholding tax (nikara) of 100 purāṇas. ⁶⁹ This is a case of karaśāsana, by which a land plot or a village is given on condition that certain amount of tax would continue to be paid to the king. ⁷⁰ Accordingly, the grant was not accompanied by full immunity. Though it is not clearly mentioned, Yaśodāsa may have petitioned the king for the bestowal of a śāsana land with the issue of a copper plate grant. The fact that this inscription was engraved on the stone slab and not issued as a copper plate grant points to a failure of Yaśodāsa. The king somehow succeeded to limit Yaśodāsa’s control over the donated tract and keep some form of interest in it.

The attempt to reclaim royal power is further reflected in the changed pattern of land and village grants. The grants of śāsana land to religious institutions established by sāmantas disappeared after the reign of Gopāla II in the late ninth century. Instead, the royal grants to individual brāhmaṇas became a norm for the Pāla copper plates from his reign onwards. The donees of royal grants in this period were highly qualified brāhmaṇas with clearer identity, who were in the process of building their own networks with establishment of Brahmanical centres in Varendra and Rāḍha. ⁷¹ Land grants to them prompted their network buildings through their settling in rural society and enabled the king to spread his claim of control through this network. ⁷² Some of the brāhmaṇas were ritual specialists serving the king, like Līlākara and Atthakara who held the titles of śāntivārika belonging to śāntigṛha under Gopāla II. ⁷³ They could be better representatives of the royal authority than sāmantas, including those of brāhmaṇa origin, who had their own territorial power and agenda.

The changing patterns of donations discussed above suggest negotiations between subordinate rulers and the Pāla kings and the latter’s attempt to gain the upper hand, which seems to have been successful to some extent. The effort of the Pāla kings to enhance their power manifested itself in the other phenomenon detectable in the later copper plate grants, which points to their attempt to tighten their control of rural society.

Tighter Control of Rural Society: Land Measurement and Numerical Standard

The change in patterns of donation coincided with the introduction of new practices. It is the land measurement and reference to ‘standard’ (pramāṇa) of particular settlements. The earliest case of land measurement is found in one of the two grants of Gopāla II dated year 4. ⁷⁴ In this inscription, the donated portion (aṅśa) and whole of the village of Bhūtabhada is measured as 1 kulya 1 droṇa 1 āḍhavāpa 2 pādas 3 kākinīs and 9 1/2 kulyas 1 pāda, respectively. ⁷⁵ Kuñjabhaddhikādāma, one of the settlements donated in the Rangpur grant of Mahīpāla I dated year 5, is measured as 41 kulyavāpas 1 āḍhavāpa. ⁷⁶ The Belwa grant of Vigrahapāla III belonging to the second half of the eleventh century records the donation of a half of village Lovanikāma with exclusion of some land tract. ⁷⁷ Both donated and excluded portions are mentioned with the standard of 3 kulyas 7 1/2 droṇas 11 udamānas and 1 kulya 2 droṇas 3 1/4 āḍhavāpas 3 1/2 udamānas, respectively. ⁷⁸ In the Amgachi plate of the same king, the donated village, Viṣamapurāṃśa with Daṇḍatraheśvara belonging to Brāhmaṇīgrāma maṇḍala, is mentioned with standard of 6 kulyas 2 droṇas 2 udamānas 3 kākinīs. ⁷⁹ One of the two land plots which constituted donated village of the Rajibpur grant of Madanapāla dated year 22 is also measured as 35 āḍhāvāpas. ⁸⁰ While the size of donated land plots was generally mentioned in the previous period, the measurement of the whole village, even when only a part of it was donated, was unprecedented.

The reference to standard of settlements first appeared in the two copper plates of Gopāla II dated year 4, in which the village or administrative unit called Suvarṇakārikādaṇḍa, from which some land plots were donated in the two separate occasions, is described as the standard (pramāṇa) of 4,000 without any specification of unit. ⁸¹ The donated portion is mentioned as the standard of 457 with 9 paṇas accompanied by 10 gaṇḍakas in the first plate. ⁸² In the second plate, it is referred to as the standard of 450. ⁸³ As both paṇa and gaṇḍaka are units of currency, ⁸⁴ it is easily inferable that the unspecified standard mentioned here is also a currency unit, of higher denomination. The similar description is used in the Belwa grant of Mahīpāla I dated year 2 for the three donated villages. ⁸⁵ Osinnakaivarttavṛtti, Nandisvāminī and Gaṇeśvarasametagrāmapuṣkiriṇī are described as 210 pramāṇas, 493 pramāṇas 6 paṇas 5 kāṇḍakas ⁸⁶ and 151 pramāṇas, respectively. ⁸⁷ In the Biyala plate of the same king dated year 35, the donated village Palāśavṛnda is mentioned with the standard (pramāṇa) of 1,000 purāṇas. ⁸⁸ Purāṇa is a unit of silver currency, which may have been a notional unit rather than a real coined currency in this period. ⁸⁹ In view of the last case, the unspecified currency unit used in the first three inscriptions must also be purāṇa. The reference to 5,910 parimāṇa (measure) in relation to Rājikāgrāmodraṅga, another settlement donated in the Rangpur grant of Mahīpāla I, can be the indication of standard in the same unit. ⁹⁰

The standard in numerals in these cases seems to be the estimated production of villages and land plots calculated in currency units, in consideration of the similar description in the Bangaon grant of Vigrahapāla III. In this inscription, donated portion, land of half the village of Vasukāvarta is described as a portion of 500 by production. ⁹¹ This conjecture is also supported by the descriptions in the two Rajibpur plates of Gopāla IV and Madanapāla, in which donated tracts are estimated to be 300 in production. ⁹² These cases, together with that of the Manahali grant of Madanapāla which seems to mention land of 320, ⁹³ suggest that the estimation of production continued to be practiced even in the last phase of the Pāla rule in Varendra. It should be noted in addition that such a practice became possible due to the spread of monetary transaction in rural society witnessed in this period. ⁹⁴

The estimation of production in currency units can be a measure taken by the Pālas to enhance their control over rural settlements by computing production by a uniform standard. Its application in tax administration is proved by the case recorded in the Bhaturiya stone inscription which refers to a 100 purāṇas as tax withheld by king Rājyapāla. ⁹⁵

Both practices, the measurement of land of the whole village and the assessment of estimated production in currency units, show the attempt at enhanced control over rural society and settlements of which the administrative machinery would gather information and keep records. It makes a stark contrast with the early Pāla grants, especially the ones petitioned by subordinate rulers, in which these practices are simply absent.

The implementation of such measures was not necessarily thorough in this period, as settlements donated in the Jajilpara grant of Gopāla III and the Bangarh plate of Mahīpāla I have no indications of land measurement or estimated production. ⁹⁶ Its full-fledged development will be witnessed in the later period under Sena rule. ⁹⁷

The royal effort to reclaim power and enhance local control may have given the king a stronger position in relation to subordinate rulers and rural society. It however did not lessen but rather heightened the tensions among these parties. The manifold tensions and conflicts culminated in the Kaivarta rebellion.

The Kaivarta Rebellion and Aftermath: The Balance Overturned

The Kaivarta rebellion was a critical event in the early history of Bengal. It temporarily ousted the Pālas from their ancestral territory of Varendra and decisively weakened their power. The descriptions of the event in the Rāmacarita show its overall character as a revolt of subordinate rulers in Varendra, with element of a popular uprising participated by wider range of social groups. ⁹⁸ The changing relation between the king and his subordinate rulers, in which the former gradually got the upper hand, the enhanced state control over rural society, which accompanied the dominant and even oppressive presence of brāhmaṇas as its agents, and finally the social changes in rural space, which included the stratification of rural residents resulting in a new form of agrarian expansion and the formation of jāti-like organisations based on occupations, ⁹⁹ brought out a conjuncture at which social contradictions and tensions exploded in the form of a large-scale rebellion.

The relation between the king and his subordinate rulers had been the focus of the rebellion. The rebellion broke out as a revolt of sāmantas under the leadership of Divya, the kaivarta chief who was a subordinate of the king. Mahīpāla II, the reigning king, was killed in a battle where his army was totally outnumbered by that of rebel sāmantas. ¹⁰⁰ The Pālas were ousted from Varendra and Rāmapāla, a younger brother of Mahīpāla II who ascended the throne after a short reign of his another brother Śūrapāla II, had to depend on the aid of sāmantas in western Bengal and eastern Bihar for its reconquest. Those sāmantas were vyālas (wicked) and āṭavikas, of which the former is glossed as appropriators of agrahāra and viṣaya (āgrahārika, vaiṣayika), while the latter is explained as sāmantas of forest (aṭavīyasāmanta). ¹⁰¹ These words suggest that those sāmantas consisted of some who had exploited weakened governmental control in appropriating land of donated tracts or lower administrative units, and the others who were forest chiefs under relatively weak control. Thus, the control of the Pālas over their subordinate rulers got weaker even in their remaining territory.

The encroachment upon donated tracts, presumably by sāmantas, was also witnessed in Varendra under the rebel chief Bhīma who succeeded Divya, where the ownership of donated tracts of brāhmaṇas and religious institutions had to be confirmed after its partial reoccupation by the advance raid of mahāpratīhāra Śivarāja, a maternal relative of Rāmapāla. ¹⁰² This common phenomenon attests to the latent tension between the king and sāmantas in which the attempt of the former to strengthen his control frustrated the latter’s efforts to enhance their local power and resource base. When the balance of power was overturned by the sudden decline of the royal power, the sāmantas went for the empowerment of themselves through such encroachment. The fact that donated tracts became their main targets points to the tension between these subordinate rulers and religious agents including brāhmaṇas and the other religious institutions under royal patronage, which the former may have recognised as agents of state control counterpoised to them.

The overturned balance also manifested itself in the autonomous actions of subordinate rulers. As attested by the Ramganj plate, mahāmāṇḍalika Īśvaraghoṣa issued his own grant to donate a village in Gālliṭipyaka viṣaya of Piyolla maṇḍala, which could be located within Varendra, to a brāhmaṇa, in spite of his position as a subordinate ruler indicated by his title. ¹⁰³ Kalyāṇavardhana, whose time corresponds to the last phase of the reign of Vigrahapāla III and the turbulent reigns of his sons, established a mahāvihāra at a place named Kuṭumvavilla according to the genealogy of Vikramavardhana, his great grandson and a subordinate ruler of Gopāla IV, recorded in the Chaprakot stone inscription. ¹⁰⁴ As mentioned above, the issue of copper plate grants was royal monopoly in the early Pāla period. Establishment of mahāvihāra was also limited to the kings, while those constructed by subordinate rulers were strictly differentiated as vihāras. ¹⁰⁵ These cases show that the weakening of the Pāla control enabled those subordinate rulers to exercise their autonomy and defy royal authority.

Due to the changed balance of power, Rāmapāla had to travel around the territories of sāmantas and appease them with gifts of land and movables to enlist their military aid. ¹⁰⁶ They were less dependable than his maternal kin, the Rāṣṭrakūṭas of Aṅga, on whom he relied as much as on himself. ¹⁰⁷ Still he had to depend on these sāmantas and reward them profusely for their military contributions. Vittapāla, his son, is said to have exhausted golden pitchers for war-gifts, presumably to reward sāmantas. ¹⁰⁸ With their help, the Pālas could suppress the rebellion and recover Varendra. Their dependence on subordinate rulers continued in the aftermath of the event. Rāmapāla is said to have been helped by a subordinate ruler called aśvapati in a battle and served by another in conquering Kāmarūpa and the other areas. ¹⁰⁹ The heavy dependence of Madanapāla, the last Pāla king properly fixable in the genealogy, on sāndhivigrahika Bhīmadeva, the messenger of the two Rajibpur grants and the Manahali plate, ¹¹⁰ is evident in the Rajghat stone inscription of the latter. It says that Maṅgadeva was mahāsāndhivigrahika of the Gauḍa king and his son Caṃgadeva gained the position of rāṇaka. ¹¹¹ Bhīmadeva, son of the latter, also became mahāsāndhivigrahaka of the Gauḍa king and is said to have supported the kingdom of Gauḍa-Varendra which moved like a sinking ship in the ocean of enemy troops. ¹¹²

In spite of the efforts of the Pālas in reconciliation and reorganisation which brought out temporary recovery of their authority, including appeasement of people with mild taxation, bestowal of an important position to Hari, an ally of the rebel kaivarta chief Bhīma, and reorganisation of viṣayas, ¹¹³ their control over subordinate rulers was decisively weakened, so that the latter became more autonomous and even independent. In the reign of Kumārapāla, Kāmarūpa was ruled by a subordinate ruler called Tiṃgyadeva. ¹¹⁴ As he showed a rebellious tendency, Vaidyadeva, another subordinate ruler born to a lineage of hereditary ministers who served the king as a minister and won a naval war in southern Vaṅga, ¹¹⁵ was sent against Tiṃgyadeva and appointed to the ruler of Kāmarūpa after defeating him. ¹¹⁶ Then Vaidyadeva himself became an independent king as attested by his own grant and the titles of mahārājādhirāja, parameśvara and paramabhaṭṭāraka. ¹¹⁷ More decisive was the case of the Senas. Vijayasena is identified with Vijayarāja of Nidrāvalī, one of the sāmantas who helped Rāmapāla in his reconquest of Varendra. ¹¹⁸ His acquisition of land by helping Rāmapāla in recovery of Varendra is alluded to in the Deopara inscription. ¹¹⁹ As evident from the presence of this inscription, which records the construction of a temple complex by the king, in present Rajshahi district, ¹²⁰ Vijayasena expanded his territory to the southern part of Varendra at the cost of the Pālas. He claimed the sovereign status for himself with titles of parameśvara, paramabhaṭṭāraka and mahārājādhirāja, as evident from his Barrackpur grant. ¹²¹ The Senas finally finished the Pāla rule in Varendra by ousting them from the sub-region at some point of time around 1165 AD, corresponding to the 22nd year of Madanapāla’s reign, the last date so far known. ¹²²

Conclusion and Further Questions

As their genealogies recorded in the inscriptions show, subordinate rulers under the Pālas had diverse origins ranging from merchants and local magnates to brāhmaṇas and non-brāhmaṇa literates. They also included members of the other royal family and forest chiefs according to the Rāmacarita. Unlike their counterpart in western India, none of subordinate rulers claim any mythical origin or genealogy connected to the famous kṣatriya clans, ¹²³ with exceptions of the Senas who originated from Karṇāta and claimed themselves to be rājaputras or brahmakṣatriyas of Candra vaṃśa, ¹²⁴ and the Rāṣṭrakūṭas of Aṅga who seems to have branched off from a well-known royal family.

One commonality of these subordinate rulers was their service to the king, especially in military capacity. The heavy dependence of the Pāla army on their service, especially in the later period, is clear in the Rāmacarita: The army of the king could be outnumbered and defeated by that of allied rebel sāmantas. ¹²⁵ One question may be raised here: What was the organisation of army under the command of sāmantas, with which they made military service? The inscriptions do not say much on this matter, while eulogising their personal valour and strategic skill. A verse in the Rāmacarita and its commentary indicate that they held substantial number of horses, elephants and foot soldiers, with which they contributed to Rāmapāla’s campaign. ¹²⁶ How they maintained or mobilised their military power is another question and there could be several forms, in view of their diverse origins. Mahaṇa, the Rāṣṭrakūṭa ruler of Aṅga, joined the army of Rāmapāla with his two sons and nephew. ¹²⁷ This was the case of mobilisation of family members, each of whom maintained his own troops: mahāpratīhāra Śivarāja, the nephew, led the advance raid on his own elephant and held excellent cavalry. ¹²⁸ Sāmantas originating from hereditary chieftains, including āṭavikas and kaivartas, may also have had recourse to the similar form of military mobilisation based on family ties.

The other kinds of sāmantas, especially those who originated from literate groups and merchants, may have depended on different sources of supply. One possible source is mercenary warriors and the inclusion of cāṭas and bhaṭas of diverse social groups from Bengal and the other regions in the Pāla army is attested by the address of the grants. ¹²⁹ The army of Madanapāla described as remunerated by cowries (kapardaka) and bread (roṭi) in the Rāmacarita could consist of such warriors. ¹³⁰ The corporation of soldiers (sainikasaṅgha), of which leaders (mukhya) accompanied a general listed as one of addressees in the Irda and Kalanda plates of the Kāmboja king Nayapāla pertaining to Daṇḍabhukti, southern fringe of western Bengal in the tenth century, ¹³¹ could also be a source of their military power.

Another commonality was their territorial control, over which tension was building up between the Pāla kings and subordinate rulers. Under the established royal control, the tension took the form of negotiation between both sides, which manifested itself as establishment of religious institutions and petition for land and village grants by the latter, permitted by the former by issuing copper plate grants. The royal efforts to gain the upper hand in this negotiation, by curtailment of royal grants to such institutions and implantation of brāhmaṇas as agents of royal authority through grants instead, bore fruits to some extent, together with another attempt to tighten control over rural society through land measurement and assessment of production in currency units. The success in those measures, however, heightened tensions and contradictions within society, especially between both parties and brought out a conjuncture at which they exploded as the Kaivarta rebellion. The balance of power was overturned and the Pālas became heavily dependent on their subordinate rulers until the end of their rule.

Thus I discussed the characteristics of subordinate rulers under the Pālas and delineated their shifting relation with the royal power. As is evident from the address of grants issued by the Candras and the Varmans, ¹³² subordinate rulers were also present in contemporary Vaṅga and Samataṭa ruled by these dynasties. How far they had commonality with those under the Pālas discussed above can be answered through studies of their inscriptions, though the agency of subordinate rulers is not easily traceable in them. On the other hand, the case of Bengal studied in the present article can be compared meaningfully with those in the other regions, especially in western India where the deeds of sāmantas and their hierarchical formation are clearer in both inscriptional and textual evidences. ¹³³ The comparative studies within Bengal and beyond, with emphasis on the power relation and its dynamism, would lead us to better understandings of the political formation of the early medieval India and its variable manifestations in different regions.