■ Article ■
Clan System or Samanta System?:
The Polity of the
a.kambhari Cdhamanas
in Early Medieval Rajasthan
●
Masahiko Mita
1. Introduction
Land assignment to clan members of the royal families depicted in the records of early medieval Rajasthan has usually been considered by many historians as a sign of the "clan-monarchies", which were so widely prevalent in later medieval Rajasthan [Ghoshal 1929: 236; Banerjee 1962: 48; K. K. Gopal 1964: 86-103; Yadav 1973: 144; etc.]. This "clan-monarchy" or so-called "clan system", often representing a model of "traditional Rajput polity" is said to be a clan-based state
system where land is assigned to the ruling clan members as their own hereditary possession with the political structure and state-locality re-lations being consequently based on clanship [Tod 1990, vol. 1: 153- 245; Baden-Powel 1972: 196-202; Thorner 1965: 133-43; Banerjee 1962: 127-28; Fox 19711); etc.]. A concrete instance of this in the history of Rajasthan was presented by G. D. Sharma, who analysed a sixteenth-century territorial system of the Marwar kingdom of the Rathors called bhizi-bandh ("bond of brothers") and bhäF-beint (land distribution among
三 田昌彦 Masahiko Mita, Nagoya University, Medieval Indian History.
Articles: "Land Distribution and Kinship of the Nadol Cahamanas: Structure of Rajput Polity in 12th-century Rajasthan", Journal of the Japanese Association for South Asian Studies, 8 (1996), pp. 27-57.
"An Aspect of the Forms of Copper -plate Charters in Early Medieval India" (in Japanese), Rekishigaku Kenkyu (Journal of Historical Studies), 737 (2000), pp. 34-44.
2 Journal of the Japanese Association for South Asian Studies, 15, 2003
"brothers") .2) Therefore, the above historians researching on early me-dieval Rajasthan understand the Rajput polity of the period in the same context as that of the later medieval Rajput states. This clan-based land distribution is often considered to have developed espe-cially under the Cahamana (Cauhan) domination in the early medieval period [Ghoshal 1929: 236; Banerjee 1962: 48; K. K. Gopal 1964: 86-103; Yadava 1973: 144; Chattopadhyaya 1976: 71-72; R. S. Sharma 1980: 143-45; 1990: 88-90].
However, as I discussed in detail elsewhere, the Cahamanas of Naddala (Nadol), a little kingdom subordinate to the Caulukya suzer-ain of Gujarat in the twelfth century, had a unique system, where lands were assigned to royal kinsmen to control the provinces and the frontiers of the kingdom but were not inheritable estates of their own and usually transferred to others on occasions of royal succession, un-like the land distribution of the later medieval "clan system" [Mita
1996; 1999]. It is, therefore, difficult to regard as a clan-monarchy even the kingdom of the Cahamanas which has been taken by many historians as one of the most developed clan-monarchies among the kingdoms of the early medieval Rajput clans.
On the other hand, on the level of suzerain kingdoms, the lord-feudatory relations called the samanta system have so far been consid-ered as one of the most important aspects of the state system of early medieval India. This system was a mechanism of state integration where many major and minor local rulers were linked to a sovereign
king through master-servant relations and more egalitarian alliances, and incorporated into a hierarchy of rulers with various titles like mandalesvara, samanta, ranaka, thakkura, etc. [Mazumdar 1960: 1-76; Yadava 1973; Inden 1981; Sircar 1982: 5-44; Chattopadhyaya 1983; 1994; L. Gopal 1989; Kulke 1995]. However, this system does not seem to match the "clan system" mentioned above. For, in the samanta system, a king usually allowed conquered rulers to continue their rule over their former territory; while in the clan system, conquered land was expected to be distributed among the royal clan members because they regarded any conquest as the result of their collaboration. What kind of political system then did sovereign kings adopt in early medie-val Rajasthan? Can we really grasp the early medieval Rajput state system in the context of the clan system as the above historians do?
How should we place the royal clan members of the suzerain Rajput kingdoms in the samanta system? This paper attempts to answer these questions by enquiring into the political structure and its development of the suzerain power of the Cahamanas who have often been viewed as a clan developing typical clan-monarchies in the early medieval period. The paper will focus especially on the political relations among kings, royal clan members, matrimonial relatives and samantas.3)
The Cahamanas formed several kingdoms in the early medieval pe-riod: the kingdom of Sakambhari (Sambhar), that of Naddilla branched off from the Sakambhari line, that of Javalipura (Jalor) branched off from the Naddula line, etc., among which the object of study here is
4 Journal of the Japanese Association for South Asian Studies, 15, 2003
the Sakambhari line. From the middle of the tenth century onwards, as the Pratihdras of Kanauj, the supreme power of north India, gradu-ally declined, the kingdoms of various Rajput clans like Candellas, Cahamanas, Caulukyas, Paramdras, Gahadavalas, Tomaras and Guhilas sprang out from the Pratihdra suzerain. While forming regional states, they struggled amongst each other for supremacy in north and western parts of India until the Turkish Gharids conquered north India in the end of the twelfth century. One of them was the Cahamanas of Sakambhari, which, later shifting their capital to Ajayameru (Ajmer), grew into the greatest power in north India by the middle of the twelfth century and continued to hold power until they were con-quered by Muhammad Ghari in the late twelfth century.
The sources mainly used here are all contemporary records: i.e., the epigraphical records concerned with the Sakambhari Cahamanas and the Prthviraja-vzja3,,a-mahakavya composed by Jayanaka, a court poet of Prthviraja III, in the end of the twelfth century. The inscriptions checked by epigraphists so far are concentrated on the following two periods: i) the latter half of the tenth century (from VS 1013, the date of Thanwala inscription in the reign of Simhardja, up to VS 1056, the date of Sakrai and Kinsariya inscriptions of Durlabharaja II) when the kingdom just became independent from the Pratihara empire, and ii) the twelfth century (from VS 1161, the date of Revasa inscription of Prthviraja I, to VS 1245, the date of Anvalda inscription of Prthviraja III) when it expanded its territory up to Delhi and beyond and reached its political peak as one of the greatest kingdoms in north India. The Prthviraja-vijaya-mahakavya will mainly be used in the second period because of the date of the work (the end of the twelfth century). This kavya literature praising Prthviraja III and his Sakambhari royal lin-eage gives us a comparatively detailed information about their royal succession and marriage as well as their military achievements in the second period. According to the dates of those available sources, in the following chapters I will discuss the structure of the Sakambhari king-dom in the two tentative stages given above, i.e., the early stage and the suzerain stage.
2. The Sakambhari Cahamnas under the Pratihara Rule: Before the Early Stage
Before discussing the early stage of the Sakambhari kingdom, we need to enquire into the situation of the Cahamanas as feudatories of the Pratihara empire, because the image of the kingdom in the early stage depends on how their situation was before independence.
The Sakambhari Cahamanas are generally believed to have formed a powerful principality under the Pratihara rule of Kanauj. The Harsa
inscription dated VS 1030 states that Gavaka I was honoured in the assembly of kings at the court of Nagavaloka,4) who is identified as Nagabhata II (794?-833) of the Pratiharas [Puri 1986: 56]. Prthviraja-vijaya-mandkavya, though a work of the end of the twelfth century,
also says that Kalavati, the sister of Gavaka II Cahamana, chose the king of Kanyakubja (Kanauj) for her husband among twelve kings and Gavaka II defeated the remaining kings and gave their wealth to her.5) However, none of them are scontemporary sources, and besides, they are sources from the Cahamana side. It is doubtful from the records of the Pratihara side how stable and powerful their principality was under the Pratihara rule. Two copper-plate inscriptions dated VS
8826) and 8987) from Badhal, only about 40 km north-east from the capital Sakambhari, record that the charters of village grant were is-sued by the Pratihara kings, Nagabhata II and Bhoja I respectively. And Daulatpur copper-plate inscription dated VS 900 records that Bhoja I issued a charter of the sanction to restore the lost grant of the village of Siva, modern Sewa, about 70 km north-west from Sakambhari.8 In those copper-plates there is no mention of the names of the Cahamana chiefs. Therefore, even if they formed their own principality, their domain was limited only to Sakambhari and its out-skirts, and their political influence did not extend to even 40 km from their capital, which would be much smaller than what is generally inferred.
Their independent activity from the Pratihara rule started under the reign of Vakpatiraja I in the first half of the tenth century, just when the Pratihdras needed to marshal all their forces against the Rastrakutas of Deccan [D. Sharma 1975: 31]. The Harsa inscription states that Vakpatiraja I of the Cahamnas defeated the army of tantrapala
(fron-6 Journal of the Japanese Association for South Asian Studies, 15, 2003
tier governor) Ksmapala who came to subjugate the Cahamanas by orders of a certain Pratihara king,9 which clearly shows the Cahamanas' rebellion against the Pratiharas. The same inscription says that the next Cahamana king Simhardja defeated many kings and kept them in jail until the emperor of the Raghu family, i.e., the Pratihdras, came to rescue them.10) In the reign of Simharaja and the next king Vigraharaja II, their territory expanded up to about 70-80 km radius from their capital, as will be discussed in the next chapter.
Thus, the Sakambhari Cahamanas were comparatively petty chiefs before independence, but they considerably expanded their territory
and grew as a real kingdom in the process of their independence. That is to say, this process of independence was that of state formation for the Cahamanas. From this stage of the Sakambhari Cahamana history, they began to record various achievements of their own in their stone inscriptions, which will be analysed in the following chapters.
3. The Early Stage of the Kingdom: The Latter Half of the Tenth Century
The Sakambhari Cahamarias of this period had just annexed the southern neighbouring Puskar-Ajmer area,11) and their territory was probably within about 70-80 km radius from their capital Sakambhari, as far as we infer from the spatial distribution of their inscriptions of the latter half of the tenth century (see Map 1). For considering the territorial system of the kingdom in this stage, the Harsa inscription inscribed in VS 1030 in the reign of Vigraharaja II would attract our attention. In this record, the village grants which had so far been made to the temple of the god Harsa, who was a tutelary deity of the Sakambhari Cahamanas, were listed as below:
(i) maharajadhiraja Simharaja granted four villages,
i.e.,Simha-gostha in Tunakupaka-dvadasaka (dvadaaka: a kind of
dis-trict), Traikalakaka and Isanakapa in Pattavaddhaka-visaya (visaya: a kind of district), and Kanhapallika in Sarahkotta-visaya, all of which (visaya) were his svabhoga (own posses-sion).12)
(ii) Vatsardja, a brother of Simharaja, granted with a s'asana
(char-ter) Karddamakhata-grama in Jayapura-visaya which had been
attained as his svabhoga.13)
(iii) Candraraja and Govindaraja, sons of Simharaja, granted with a
kisana with their own signature (svahast-amkita-Sasana)
prob-ably two villages in Pattavaddhaka-visaya and Darbhakaksa-visaya, both of which had been attained as their svabhoga.14)
(iv) Dhandhuka, a duhsadhya (leader of policemen) of Simharaja,
granted with his master's approval Mayurapadra-grama, which
was being enjoyed (svabhujyamana) (by him) in
Khattakapa-visaya.15)
(v) Jayanirdja who was a•\hila-atmaja granted with a s'asana
Kolikupaka-grama which was being enjoyed (svabhujyamana)
(by him) .16
The grantors of (ii) and (iii) were sons and a brother of mahardja-dhiraja Simharaja, and the grantor of (iv) was his officer. Although
8 Journal of the Japanese Association for South Asian Studies, 15, 2003
whether the grantor of (v) was a relative of the king is not necessarily clear,") it is apparent from this record that the nature of the land holding of sons and brothers of the king sharply contrasted with that of non-relatives like officers. First of all, their right to the land (svabhoga or svabhujyamiina) was different. When granting land, Dhandhuka, who was an officer, needed the king's approval, though the king's sons and brothers did not have to seek his consent. Such difference between the territorial right of the royal relatives and that of the officers has often been pointed out by the historians dealing with this inscription [Chattopadhyaya 1976: 71; K. K. Gopal 1964: 80, 91; etc.].
Second, there was a difference between the spatial distributions of their land, which does not seem to have been noticed by former histo-rians. We find this fact from the record that the king and his sons and brother held their lands in a bulk of territory, that is, in an administra-tive unit like visaya (district) and dvadaiaka (12 villages), while the unit of land holding of others was a single village. The records of (i), (ii), and (iii) describe the granted villages as "svabhoge•\visaye•\grama"
(•\grama in the•\visaya which is one's enjoyment) or "svabhogavapta visaye•\grama" (•\grama in the•\visaya which was attained as one's enjoyment), where the terms "svabhoga" and "svabhogavapta" appar-ently modify "visaya", and not "grama". But in the record (iv), the granted village is described as "Khattakupa-visaye svabhujyamana-Mayfirapadra-gramam", where undoubtedly the term "svabhujyamiina" does not modify "Khattakupa-visaye" but "Mayfirapadra-gramam". The record (v) is not as clear as (iv) because it does not mention a relevant visaya, but its description "svabhujyamdria-Kolikupaka-greimam" is the same expression as that of (iv). Thus, as far as this inscription is concerned, persons other than close relatives of the king were assigned land in a village unit. Therefore, if they held several villages, which must have been the case because each person granted a whole village in the above records, their villages must have been scattered over a wide area.
Thus, the Sakambhari kingdom seems to have had two different types of territorial holding even in its beginning stage. The visaya holding by the close relatives of the king would have meant their relatively autonomous rule over the visayas, because they enjoyed the royal prerogative to issue s'asanas of land grant by themselves with
their own signature (svahast-amkita-Sasana), as is apparent from (iii).
On the other hand, scattered land holding in a village unit would have
had a quite different meaning. This scattered holding of land cannot
be attributed to the result of frequent division of visaya property in
the process of inheritance, because it was just after the annexation of
this area, as is discussed in the previous chapter. We must, therefore,
recognize that the king assigned land dispersively on purpose. The
scattered villages can hardly be considered to have been assigned in
order to entrust the assignees with the control of districts or provinces
and, therefore, would have been mere allowances for the assignees'
maintenance, or a kind of reward which was given to persons
render-ing distrender-inguished service. Whether the villages were assigned as
allow-ances or rewards, the assignees probably only had rights to collect
revenues from the assigned villages, and they could make religious
grant from them but not issue charters, probably because, unlike the
king's sons and brothers, they did not have any administrative
respon-sibility for their assigned villages.")
Territorial holding with certain autonomous rights like visaya
hold-ing mentioned above was not confined only to khold-ings' sons and
broth-ers, but seems to have been assigned also to persons of non-Cahamana
clans. Kinsaliya inscription of VS 1056 is a prasasti (eulogy) of the
Dhadicika clan made by Cacca of the clan, where a short prasasti of
the Sakambhari Cahamanas is also inscribed before their prasasti. It
would mean that the Dhadicikas controlled Kinsaliya and its
sur-rounding areas in the west of the Sambhar lake under the political
influence of the Cahamanas of Sakambhari.19) In the southern
periph-ery, king (bhupa) Durgaraja originally ruled over Puskar (Puskar
In-scription of VS 99420)), but according to the Tanwala inIn-scription dated
VS 1013,21) Puskar was under the reign of Simhardja and mahantaka
Durgaraja granted the village of Namda near Puskar with a Sasana,
which would indicate that he still governed Puskar as a subordinate of
Simhardja. Simhardja included the local chief within his domain by
bestowing the title of mahantaka on him. However, it is not known
from these examples whether local chiefs enjoyed the rule over their
land as freely as the close relatives of the king did.
Let us then see the spatial distribution of the land. As the
above-mentioned Harsa inscription clearly indicates, king Simharaja also had
10 Journal of the Japanese Association for South Asian Studies, 15, 2003
his own svabhoga. Four visayas are mentioned as his svabhoga, which was larger than those of any others. The visayas enjoyed by him were located not only near the capital Sakambhari but also in the periphery of the kingdom like Pattavaddhaka-visaya. And as the village of Mayurapadra of Khattakupa-visaya was assigned to duhsadhya Dhanduka by Simharaja (see (iv)), this visaya was also his personal territory. Thus, the major part of the kingdom was under the direct control of the king.
As for the territory of the king's brothers, Jayapura-visaya of Sim-hardja's brother Vatsaraja has not been identified so far. Both Candraraja and Govindaraja were the brothers of the next king Vigraharaja II who was the reigning king when the Harsa inscription was inscribed. One of their territories, Pattavaddhaka-visaya,22) was located in the northern frontier of the kingdom in VS 1030.23), 24)
Thus, in the early stage of the kingdom, the Sakambhari Cahamana kings directly controlled over the vast areas of the kingdom including its heart land, and distributed frontier areas to their sons and brothers and even to the conquered rulers, like Durgaraja of Puskar and the Dhadicikas, entrusting the rule over these areas. The rule of the kings did not extend to the land distributed to their sons and brothers in the aspect of the issue of sasanas. In this sense, this kind of land distribu-tion certainly looks like the so-called clan-system, but unlike the later Rajput kingdoms based on the clan-system (like the Rathors and the Kacchawahas), even the brothers of Vigrahardja II, Candraraja and Govindaraja, who had the visayas of Pattavaddhaka and Darbhakaksa, do not seem to have owned them as their hereditary estates, conse-quently not having formed the principalities of collateral lines. The later records concerned with these areas, like the Revasa inscriptions dated from VS 1161 to VS 1230, do not mention their lines but only the direct rule (rajya) of the Caharndna kings,251 and the inscription of VS 1243 describes the Revasa region as Candela-pratiganaka, i.e., the district of the Candela family.261
Only one exception was the Naddala Cahamanas or the line of Laksmana, who was another brother of king Simharaja. However, ac-cording to their legend in later records like Nainsi's Khyat and Puratanaprabandhasangraha, they formed their principality not by
Naddula region [D. Sharma 1975: 139-40; Chattopadhyaya 1990: 71]. Thus, it would be concluded that, in most cases, the land distribution among the royal members did not lead to the formation of their prin-cipalities in the early stage of the Sakambhari Cahamana kingdom.
4. The Suzerain Stage of the Kingdom: During the Twelfth Century
In this period, especially after the middle of the twelfth century, the territory of the Sakambhari Cahamanas went beyond Rajasthan and extended to Punjab, Hariyana and Delhi. As a result, they bore a large number of samantas within their domain. The Delhi-Siwalik inscrip-tion of Vigraharaja IV dated VS 1220 loudly proclaims that they con-quered the land between Himalaya and Vindhya and made it tribute-paying countries (kara-da), i.e. samantas.27) But unfortunately, the avail-able inscriptions showing the mechanism of the samanta system and the local administration of the kingdom come from the northern, the southern, the western parts, Delhi region and Sakambhari-Ajayameru
region. Here, we enquire into their ruling system in each region re-spectively.
4.1 Northern Part
In the northern part of the kingdom we have only one epigraphical record, the Hansi inscription of VS 1224 (about AD 1167). It is a very valuable record indicating how the Sakambhari Cahamanas put samantas under their control. This inscription is the prassasti (eulogy) made by Laksmana of the Dodas, and admires Kilhana of the Guhilas who was the master of Laksmana's father Valha. This record says Prthvirdja II, king of the Sakambhari kingdom, granted the fort of As i or -Asika. (modern Hansi) to Kilhana in order to protect the kingdom against the approaching Muslim army, and Kilhana defeated the king of Pancapura and captured its capita1.28)
Tod says that the Aika fort was on the route by which most of the Muslim army's invasions were made, and he adds "The route was by Pachapattan, the town of Purity, on the Setlej, to Bhatner and
Fateh-abad, to Asi and Delhi".29) From his words, as D. R. Bhandarkar points out, the importance of fortifying and maintaining the Mika fort to the
12 Journal of the Japanese Association for South Asian Studies, 15, 2003
Sakambhari Cahamanas is quite clear, and the necessity of keeping a strong hold on Pachapattan, which is Paricapura of the Hansi inscrip-tion, is equally clear.30) According to the Bijolid inscription of VS 1226,31) the Cahamanas captured this fort in the reign of the last king Vigraharaja IV, which would suggest that their influence had not yet thoroughly extended over the northern frontier. And it is apparent from this inscription that Pancapura was not under the control of the Cahamanas before the date mentioned. Hence, it would easily be con-cluded that Kilhana's conquest of Pancapura aimed at making the Cahamanas' strong hold on this area as a part of the defensive action
of the Cahamanas against the Muslim invasion from the west. Besides, at this event, the ruler of Pancapura was admitted to govern this city continuously under the Cahamana rule,32) and therefore, even after this conquest Kilhana must have kept watching his movements. Thus,
Kilhana's main duty was to hold the Asika fort and protect it against the Muslims as clearly inscribed in this inscription,33) and as a part of his duty, he watched the movements of the samantas in the northern frontier of the kingdom from the fort, and, if necessary, conquered the neighbouring rulers.
Kilhana had his own army composed of several troops whose leaders were his vassals like Valha of the Dodas in this inscription. He con-quered the ruler of Pancapura with a private army formed by such master-servant connections.
His relation with the surrounding samantas can be known from the following portion of the Hansi inscription:
"Then
, you (Kilhana) wrote a document with your signature that a string of pearl and this city (Pancapura) had been given to me, the ruler of Pancapura, by the lord (Prthvirdja I I)."34)
This is a portion of the letter to Kilhana from the ruler of Pancapura. "A document" (patra) mentioned here is apparently a kind of prasada -patras or praseida-likhitas which were issued when the king assigned
land to his followers or sanctioned their land holding in his favour (prasada).35) Here also, the person who sanctioned the rulership of Pancapura was king Prthviraja II, but the person who wrote this deed and signed it was Kilhana. This means that the person to whom the conquered rulers or the samantas rendered homage was not a conqueror Kilhana but the Sakambhari king. Therefore, Kilhana's domain must have been limited to the Asika fort and its surrounding small area, and did not include those samanta territories. However, according to this inscription, Kilhana would have given a prasada-patra to the ruler of Pancapura without any sanction of the king, otherwise he would have got the king's signature. Thus, though as a king's agent, Kilhana actu-ally formed and maintained the overlord-seimanta connections in this region at his own discretion to a considerable extent. As the Asika fort and its surrounding areas were on the route of Muslims' marching into north India, the protectors of the fort would have been expected
14 Journal of the Japanese Association for South Asian Studies, 15, 2003
to take prompt measures to cope with the movements of these areas. That is probably why he was invested by the king with full power of
diplomacy in the northern frontier.
Such investment of considerable discretion to kings' agents was not unusual in early medieval India. In Saurastra under the rule of the Pratihara overlord Mahendrapala in the end of the ninth century, mahasamanta Balavarman Calukya issued a charter recording land grant out of his territory assigned by the Pratihara over-lord, with the king's agent tantrapala (governor dealing with frontiers) Dhika's signature as well as Balavarman's.") Maha-samanta Avanivarman II of the same principality issued a land grant charter with the same tantrapala's sanc-tion.37) An undated Pratapgarh inscription of the time of Mahedrapala Pratihara in the mid tenth century records that tantrapala mahasamanta dandandyaka Madhava issued a sasana of land grant at Ujjaini at the request of malidsamantaa Indraraja of a Caharna-na family with the sig-natures of Madhava and a certain Vidagdha.38 These land grants should normally have been sanctioned and signed by the Pratihdra king him-self. But in such frontiers as Gujarat and Malwa, the samantas were put under the control of these kings' agents, tantrapalas. In the Caulukya kingdom in the twelfth century, activities of the frontier rulers were more positive and diplomatic like Kilhana of the Sakambhari empire. In VS 1218, Somevara of the Bhinmal Paramdras, who was a power-ful samanta dependent on the Caulukyas of Gujarat, defeated king Jajjaka and captured the forts of Tanukotta (modern Tanot) and Navasara (modern Nosar) in the north-west frontier of Rajasthan and in western Rajasthan respectively, and put him back again in the coun-tries (desas) of these forts, as he became a follower (sevaka) of the Caulukya supreme lord (jagatipati).39) He also seems to have done so at his own discretion.
4.2 Southern Part
As for the south-east region of the kingdom, three Dhaud inscrip-tions say as follows:
(i) VS 1225:4")
Under the reign of the Sakambhari king Prthviraja (II), and under the governance (pratipatti) of Kurnarapala, the adhiraja
of Uparamvala-Antari region, who was subsisted by the king
(tatpadapadmopajiuin), the same Kumarapala constructed a temple
in Dhavagarta. (modern Dhaud) which was enjoyed
(subhujya-mana) by rand Rautvamsu.
(ii)
VS 1228:41)
When one person (his name illegible) was subsisted
(tatpada-padmopajivin) and appointed by king Somevara as the executor
of the administration of Dhavagartd, Cahada sold his house for
god Nityapramoditya.
(iii) VS 1229:42)
Paramamahesvara Kumarapala of Vagadiya-ku/a who was
sub-sisted (tatpa- dapadmopajivin) by king Somevara constructed a
shrine in Dhavagartd.
According to (i) and (iii), Kumarapala was subordinate to two
Cahamana rulers, Prthviraja II and Somevara, and was the adhireija
of Uparamvala-Antari (Bhil-wara-Jahazpur) region. However, an
un-named person of (ii) cannot be Kumarapala, because the father of this
person is inscribed as Kirttisiha in the inscription (ii) while
Kumara-pala's father was thakkura Mangalardja according to (i) and (iii).
Such absence of the name of the adhiraja can be seen even in the
inscriptions of other cities and towns in the Uparamvala-Antari
re-gion. The Bijolia inscription of VS 122643)
in the same region does not
refer to his name. Around Bijolia, the Guhilas were probably
domi-nant because many names of the Guhilas can be seen in the list of the
land grantors inscribed in the inscription, and their titles were rd.
(rauta or ranaka?), maham. and mahattama." But they did not form a
samanta principality. The inscription depicts the praiastis of the
Cthamanas, conquerors, and of the Pragvdtas, makers of the
inscrip-tion, but not that of the Guhilas, which should have been inscribed if
they had formed a feudatory state. The Pragvatas not only inscribed
the Bijolia inscription but also repaired the Great Fort of Mandalakara
(Mandalakara-mahadurga)
near Bijolia. The Pragvdtas were possibly
trusted with the administration of the city of Bijolia by the Sakambhari
king.
The Menal (former Mahandla) inscription dated VS 1225 records
that mandrajni Sahavadevi (or Suhavadevi), queen of Prthviraja II,
16
Journal
of the Japanese
Association
for South
Asian
Studies,
15, 2003
issued a charter of revenue grant, but this also does not mention the
name of adhirdja Kumarapala.45) The inscription from the same place
dated VS 1226, too, mentions only the rule of Prthviraja II at the fort
(gadha) of Mendlakd.46) Anvalda in Uparamvala-Antari area has two
inscriptions dated VS 1234 and VS 1245,47) both of which inscribe
religious grants made by Doda ra. Simghard and Doda rd. Jehada
respectively under the rajya (rule) of the5.-kambhari kings. Neither of
them mentions the name of adhirdja Kumarapala. Anvalda was
prob-ably the dominant land of the Doda Rajputs.
Thus, it is apparent that the rulers and residents of these cities and
villages of Uparamvala-Antari region were directly subordinate to the
Sakambhari rulers and not to the adhirdja of Uparamvala-Antari. How
then should we explain this political situation of the adhirdja? It is
apparent from the above situation that the adhirdja did not form a
principality. This conclusion may be supported by the term "pratipatti",
which is used to express his duty in the inscription (i). This term is
often used for the administration made by town committees like
pancakulas;48) while the term "rdjya" is usually used for the rules of
kings and sdmantas.
The above situations are very similar to those of Kilhana of the
ikika fort. He checked his surrounding sdmantas only as a king's agent
and did not receive their homage, as is discussed earlier. The same
would be the "territory" of adhirdja Kumarapala; i.e., that is not a
principality (rdjya) but the area of jurisdiction within which he must
have checked the sdmantas. That would explain why any sign of
mas-ter-servant relation was not mentioned between adhirdja Kumarapala
and the revenue enjoyer of Dhaud, rand Rautvamsu, in the inscription
(i). Moreover, he belonged to the Vagadiya-kula, probably a branch of
the Cdhamanas," whose homeland was Vagada (Dafigarpur region in
south-east Rajasthan), very far from Uparamvala-Antari. He probably
pledged his loyalty to the Sakambhari king Prthviraja II
(tatpadapadmo-pajivin) and was sent to Uparamvala-Antari as the adhirdja of the
region. Thus, he was similar to Kilhana of Aika even in the aspect of
being an outsider.
4.3 Western Part
the local power of the western part of the kingdom.50) It says, under the rajya of mandalekvara rand Katiyarajam) of Pravdra (Paramdra) vamsa sanctioned through the favour (prasada) of maharaja Prthviraja (III), the same Katiyaraja granted a canal as dohalika (land of religious grant) to god Lokevara of Vijayapura (modern Phalodhi) in the territory of the city of Vikramapura (modern Bikanpur).
This type of siimantas, having such titles as mandalekara and rand, practicing rajya and issuing land grant charters in their own territory sanctioned by overlords, was very often seen under such powerful dynasties of early medieval north India as the Pratihdras of Kanauj,52) the Caulukyas of Gujarat") and the Paramaras of Malwa.54) In this sense, Katiyardja was a typical seimanta and would have exercised al-most fully autonomous power within his territory. His territory was very large, having Vikramapura as the capital and including even the far southern city Vijayapura (about 70 km distance), as the inscription clearly shows. He was probably a considerably powerful samanta in the5.-kambhari kingdom.
Katiyaraja belonged to Kaudiniya-sagotra of Paramdra-vam§a. The inscription mentions that his father Palhana was also a feudatory and had the title of maheislimanta.55) If the"a.-kambhari Cahamanas had the same samanta hierarchy as the Caulukyas had, as was depicted in Aparajitaprccha written by a court architect of the Caulukyas, he achieved a higher rank than his father's.56)
It is impossible to know from this inscription where the territory of his father was located. The inscription says his territory was granted by Prthviraja III through his favour (prasada). But the Phalodhi in-scription dated VS 1145,5') about ninety years earlier than the date of the above inscription, mentions raja Hathadeva of the Pamvaras (Paramaras), which indicates that the Paramaras had ruled over at least Vijayapura (Phalodhi) since this date. The Paramaras had something to do with this area even before VS 1145, for a memorial pillar (govardhana) inscription dated VS 1070 discovered at Pokaran, about 50 km west of Phalodi, records the death of a warrior named Ghimghaka of the Paramdras in a certain battle,58) though he does not appear to have been a ruler of Pokaran.") Thus, it appears that Vijayapura and its surrounding areas were originally the home ground of this branch of the Paramaras. The territorial assignment to Katiyardja by Prthviraja
18 Journal of the Japanese Association for South Asian Studies, 15, 2003
III, therefore, would have meant no more than a confirmation of his ancestral territory. Inferred from the title raja of Hathadeva in the Phalodi inscription of VS 1145, these Paramaras would originally have been independent, and the Sakambhari kings apparently put them un-der their control by vesting them with the title of mahlisamanta (i.e., great frontier ruler) and mandalekara (i.e., ruler of province), which meant that their territories became portions of the entire Cahamana domain. Probably, both Pdlhana and his son Katiyaraja, whose ances-tral territory was sanctioned by the Sakambhari king, gradually raised their ranks in the Cdhamana court through their service to their over-lords and consolidated a powerful seimanta principality in this western frontier of the Cahamana territory.
4.4 Delhi Region
In the north-eastern frontier of the Sakambhari domain, the most important was no doubt Delhi, which had been kept under the control of the Tomaras before the Cahamana rule. The Tomaras had retained strong influence over the Haryana-Delhi region since the period of the Pratihdra empire [D. Sharma 1966: 224-25; 1975: 50-51], and since then, had had continual battles with their southern enemy, the Sakambhari Cahamanas who proposed territorial expansion towards the north-east. According to the records of the Cahamana side, the Caham5nas defeated the Tomaras again and again,601 but they could not subjugate them. In the reign of Vigraharaja IV, they finally suc-ceeded in capturing Asika. (Hansi) and Dhillika (Delhi), and put the Tomaras under their contro1.61)
Unfortunately, we do not have any epigraphical records showing the Tomaras under the Cahamana overlords. However, Kharataragaccha-pattavali, the fourteenth century Jaina chronicle, mentions Yoginipura or Delhi ruled by Madanapala Tomara in VS 1222 [D. Sharma 1975: 68]. This date was no doubt after the capture of Delhi by the Cahamanas, because the date when Vigraharaja IV proclaimed the achievement of the conquest of north India in the Delhi-Siwalik in-scription was VS 1220.62) Madanapdla would have continued to reign over Delhi as the subordinate king probably in the same way as the Paramdra seimantas of Vikramapura in the western frontier. But it is apparent from the Hansi inscription dated VS 1224 discussed above
that the stronghold of Aaika., which had formerly been in the Tomara domain, was not allowed to continue to be in their possession; it was assigned to Kilhana Guhila, a maternal uncle of the reigning king Prthviraja II, on account of defence against the Muslim invasion and check on the movements of samantas. Thus, in exchange for maintain-ing their status, conquered rulers sometimes reduced their territory, following the convenience of the state administration of the Gahamanas.63) 4.5 Sakambhari-Ajayameru Region and Its Surroundings
No epigraphical record which indicates the administrative system of the kingdom has so far been discovered from the centres of this heart-land of the kingdom, i.e., the new capital Ajayameru and the former
Sakambhari.64)valuable inscriptions are found from those surrounding
areas like Narhar, Revasa, Carla and Kekind. Most of them are me-morial stone inscriptions (including sati stones) called govardhanas, devalTs and pallyas in Rajasthan, which record deaths of heroes and other prominent persons sometimes accompanied by their wives per-forming satr, most of whom were of Rajput clans [Chattopadhyaya
1982].
Narhar memorial inscription dated VS 1215 records that Talhana, a grandson of thakura Candra of Naigama family (anvaya), a trader community, died and his memorial was constructed by his uncle thakura Somadeva under the rule (rajya) of Vigraharaja IV.65) It is inferred from their title of thakura, though not beyond speculation, that this trader community was possibly appointed as the administrators of this town by the king like the Pragvatas of Bijolid mentioned before. There would have been no subordinate kings like samantas in this area, be-cause there is no mention of intermediary powers between these thakuras
and the king in this inscription.
Revasa memorial inscriptions dated VS 1243 record the deaths of three warriors of the Candelas, i.e., Nannava, Simharaja and Durlab-haraja in some battle in the reign of Prthviraja III. According to
Nannava's memorial, Khaluvana-grama, modern Revasa [Jain 1972: 407], belonged to Candela-pratiganaka.") A district (pratiganaka) named
after the clan name Candela allows us to consider the area including Revasa as the home ground of a branch of Candela families.
20 Journal of the Japanese Association for South Asian Studies, 15, 2003
of the Mohilas, a branch of the Cdhamanas. Four memorial stones from Carla are published. Two memorials record the deaths of Ahada and his son Ambaraka of the Mohilas on the same date of VS 1241 at the battle of Nagapura (modern Nagaur), which was between the Sakambhari king Prthvirdja III and the Caulukya king Bhima II of Gujarat [D. Sharma 1975: 84-85; Majumdar 1956: 140-41]. These stones were not located at the death place Nagaur, but at Carla. This leads us to the conclusion that the memorials were built in their home.67)
These two Rajput families did not have any titles in their records. If they had titles, they might have recorded them more elaborately in the actual memorial inscriptions rather than in other types of epigraphical records. In fact, memorial stones from Dadana in Nagaur tahsil record the titles of the dead, i.e., mahasamanta Palhana and rdjaputra Valhana.68) Therefore, though they would certainly have been dominant in those homelands, they might have not ruled over them as samantas, who bore such titles as mahfimandalesvara, mahasamanta, mandalika, etc. and whose territories were sanctioned or assigned by kings. However, how they controlled these areas and what their political relation was with the Sakambhari kings are totally unknown from those memorial stones.
Kekind (former Kiskindha) is the closest place from the capital Ajayameru among the towns where the epigraphical records concerned with the Sakambhari kingdom are discovered. Four inscriptions have been checked so far: the inscription VS 1176 records the donation made by rajaputra Mahipala and Rudra Cahamana at Kiskindha; VS 1178, the rule of mahamandalika ranaka Piplardja at Kiskindha; VS 1202, the donations by rdnaka Sahanapala and rani Samvaladevi; and VS 1224, the donations by mahamandaleivara Jasadharapala at Kiskindha.69) It is apparent from those records, especially that of VS 1178, that Kiskindha was the capital of a certain mandala and mahamandalika ranaka Piplaraja was on the seat of the rulership. Jasadharapala would also have been the mahamandalesvara of Kiskindha though the record does not mention his rule. But it is unknown whether other two were the rulers of the mandala, the subordinates to the rulers of the mandala, or mere outsiders. As the relation among the four cannot be known from those inscriptions, it is also unknown whether the rulers inherited the seat as other major samantas in the
frontiers, i.e., Tomaras of Delhi, Paramaras of Vikramapura, etc. There is no inscription from the capital Ajayameru showing the administrative system, but instead, there are many memorial stones from the area of Sakambhari (Sambhar), which was their former capi-tal and their home ground. Six stones dated VS 1221 record the deaths of six persons belonging to the Cahamanas and their wives:70) From them we can infer that many royal clan members had their political base at their former capital area as well as the capital Ajayameru. One memorial stone at the same place dated VS 1244 records the deaths of sahani (general) Vachardja of the Cahuvanas (Cauhans or Cdhamanas) and his mother Gohaldni Visaladevi:71) It is evident from the Delhi-Siwalik inscription, where Vigragharaja IV ordered the royal clan mem-bers (samtanajan= dtmanah) to conquer the remaining unconquered areas,72) that the armies commanded by the royal clan members formed the main forces of the Sakambhari kingdom. Sadani Vachardja, and probably some of the royal clan members of the VS 1221 stones too, must have been such royal commanders.73)
As discussed above, even the area adjacent to the capital Ajayameru was ruled by stimantas, and on the other hand, some places very far from the capital, like Narhar, were under the direct control of the kings. And the Sakambhari-Ajayameru region was the dominant area of the Cahamana royal clan members. From that, we would suppose a complicated territorial system of the heart-land of the kingdom, though the mechanism of their territorial arrangement is entirely unknown
due to lack of information.
5. Maternal Relatives and Collateral Lines
The most interesting point in provincial administration in the suze-rain stage discussed above is that the kings' maternal relatives some-times obtained crucial posts in the state administration: i.e., Kilhana Guhila, who was appointed by king Prthviraja II to the governor of the :Mika fort. Similarly, Bhuvanaikamalla of the Kalacuri family helped Prthviraja III during the regency both as one of the most trustworthy ministers and as one of the most powerful warriors, as Garuda helped Rama.74) According to Prthviraja-vijaya-mahakavya, Bhuvanaikamalla was a younger brother of the father of Karpuradevi, mother of Prthviraja
22 Journal of the Japanese Association for South Asian Studies, 15, 2003
III, and her father named Tejala or Acalaraja was the king (purandara) of Tripuri,75) who is identified as Gayakarna of the Kalacuris by D. C. Sircar [Sircar 1971: 154], and as Nrsimha or Gayakarna's son by D. Sharma [D. Sharma 1975: 77]. At any rate, Bhuvanaikamalla was un-doubtedly a senior member of the Kalacuri court. Apparently, Prthviraja III lured him away from that court through his network of maternal relations.
Kilhana belonged to the family of Guhila and was a maternal uncle of the reigning king Prthviraja II, as is evident from the Hansi inscrip-tion dated VS 1224.76) He was also regarded as an incarnation of Hanumat and king Prthviraja II as that of Rama in the same inscrip-tion,771 which is almost the same expression as in the above case of Bhuvanaikamalla and his master Prthviraja III. But it is unknown from our sources where Kilhana came from. In order to seek the answer, let us check the matrimonial relations of the Sakambhari kings during the suzerain period:
Table
From the above, we can see a distinct pattern in their marriage. The above queens were probably pattarajnis (chief queens) because their sons became kings. All of them were formerly princesses of other kingdoms. Besides, except for Sudhava who was a princess of "the desert country" (Marubhumi),781 they were daughters of independent sovereigns, and not those of petty samantas or governors who did not have their own principalities. Probably because in the kings' marriages, especially in the selection of primary queens (pattareijnis) who were supposed to bear the next kings, candidates would have been required to have socially and politically highly ranked family-background, and also because their marriages were usually one of the most important
tactics of diplomacy towards the neighbouring powerful kingdoms. Therefore, undoubtedly Prthviraja II's mother would have been a prin-cess of a powerful Guhila dynasty, which was probably the Guhilas of Mewar (Nagda-Ahada Guhilas), and her brother Kilhana would have been a senior member of the Mewar court formerly before VS 1224, the date of the Hansi inscription. Possibly, Kilhana's vassal Valha of the Dodas') also might have been from the Mewar region, for Anvalda in the northern frontier of Mewar was one of the territories of the Dodas, which is apparent from the Anvalda inscriptions dated VS 1234 and 1245 recording grants made by the Dodas."
Thus, the kings' maternal relatives recruited from outside were broth-ers or uncles of the reigning kings, that is, membbroth-ers of collateral lines, of other kingdoms. This recruitment has two implications that indicate the characteristics of early medieval Rajput polity. The first implica-tion is relevant to the reason why those kings, Prthviraja II and Prthviraja III, did not select their brothers and other royal clan members but their maternal uncle and granduncle, when they sought trustworthy persons for appointment to such important posts. This may indicate strong influence of the kings' wives and mothers in the royal court politics.81) But this also means that the royal clan members were not always reliable towards the kings and often inappropriate to such cru-cial posts, because they were candidates for the next kings and often became rivals at the time of royal succession. In the early medieval Rajput kingdoms, they instituted crown princes (yuvariijas or kumaras), which were usually the kings' eldest sons and sometimes even other sons selected by kings, in order to avoid unnecessary succession dis-putes. But actually, succession troubles often occurred, sometimes on account of the absence of successors to kings and sometimes due to usurpation by dissatisfied royal clan members; consequently, the royal lines often shifted to other collateral lines. Thus, the Sakambhari Cdhamanas, in the last six or seven generations of the dynasty, experi-enced continual succession disputes. Arnoraja had two wives: Sudhavd, princess of the Maru country, and Karicanadevi , daughter of Caulukya king Jayasimha (see the Table above). The former's eldest son, Jagad-deva, killed his father Arnoraja and usurped the throne, and in turn his younger brother Vigraharaja IV revolted against Jagaddeva and ascended to the throne. Then, the former's son Aparagangeya
suc-24 Journal of the Japanese Association for South Asian Studies, 15, 2003
ceeded him, but Jagaddeva's son Prthviraja II took up arms against Aparagangeya probably as a son of the eldest son of Amor*, defeated him and acquired the throne.82) But as he probably died without suc-cessors, the ministers (mahamatya) brought his uncle Someavara, the only surviving son of Arnordja, from the Caulukya court of Gujarat and offered the throne to him.83) That led to the shift of royal lineage from the Sudhavd-line to the Katicanadevi-line.84) Someavara was suc-ceeded by his son Prthviraja III, but as soon as he ascended the throne at only twelve years old, Nagarjuna, a grandson of Arnoraja in the Sudhavd line, revolted against him and captured Gudapura. By annihilating his army completely, Prthviraja III consolidated his reign, and started his digvijaya or "conquest of the world".85) Thus, even in the zenith of the Sakambhari kingdom, their royal court politics was often unstable especially at the moment of royal succession because of the royal members being candidates for king. In such a situation, a matrimonial network of relations would have offered an alternative way of seeking reliable persons of talent.
The second implication is the political situation of the supply side, i.e., the royal families on the maternal sides. It seems that in early medieval Rajasthan the king's brothers were not usually allowed to form collateral junior kingdoms or samanta principalities within the territory of the kingdom. They appear to have avoided distributing the hereditary territory among them as far as they could. Thus, as men-tioned repeatedly, in the Nadalla Cdhamanas, a contemporary king-dom of the collateral lineage of the Sakambhari kingdom, land and revenues were certainly distributed among the royal kinsmen for their maintenance, but they were not inheritable and very often shifted to possession of others; consequently, king's brothers could not form their own principalities in most cases, whether dependent or independent [Mita 1999].86) That is why kingship was not open to king's brothers who missed the throne, unless they usurped the throne or founded new kingdoms by colonizing outside the territory of the kingdom. Otherwise, when they attempted to seize power within the royal court as advisers of the king, if they were very capable and ambitious, the king might have avoided them or tightened his guard against them as rivals who had rights to the throne. Those able royal members were rather required in the other royal courts into which their sisters and
nieces married, as mentioned earlier. This balance in supply and
de-mand would suggest that many more of the kings' maternal relatives
than those known from our records might have played vital roles as
the kings' advisers in the royal court. Matrimonial network in early
medieval Rajasthan functioned not only as alliance and restraint
be-tween the dynasties related by marriage as is often pointed out so far,
but also as one of the most important channels to bring competent and
trustworthy persons of the non-Cdharndna clans as a counterbalance to
the royal clan members who were potential political enemies to the
kings.
6.
Conclusion:
Political
Structure
of the Sakambhari
State
through
the Stages
As discussed above, the Sakambhari kingdom can be considered to
have stepped on the two stages of the state formation since their
inde-pendence from the Pratihdra kingdom. Before indeinde-pendence,
the
Sakambhari
Cdharndnas was a petty chiefdom ruling over only
Sdkambhari and its outskirts, though they gained a certain grade of
status in the Pratihdra court. They probably formed a kingdom in the
process of their efforts at gaining independence from the Pratihdras.
Hence, the first stage, in the latter half of the tenth century, would
have been a very early stage of their state formation. In this stage,
some frontiers of their territory were divided among the close relatives
of the reigning king in order to control the frontiers. They enjoyed
considerable autonomous power within their territories, but their rule
seems to have been unstable and not to have continued even for
sev-eral generations. Some other areas were left to other clans who were
originally the rulers or the dominant groups there like Durgaraja of
Puskar and the Dhadicikas in the west of Sambhar lake. This system
in the early stage is strikingly similar to that of the Naddala Cdhamdnas,
a siimanta kingdom subordinate to the Caulukyas, in which the king's
sons ruled the distributed territories at their own discretion but their
territories were not hereditary estates, and therefore, they did not form
their own principalities as big samantas did [Mita 1996; 1999].
On the other hand, in the twelfth century, great numbers of local
rulers of non-Cahamdna clans were incorporated into the territory of
26 Journal of the Japanese Association for South Asian Studies, 15, 2003
the Sakambhari Cdhamanas as they advanced conquest. Thus, the sec-ond stage of their state formation was the suzerain stage composed of many feudatories i.e., samantas. In this stage too, the royal clan mem-bers still continued to be some of the most important staff of the kingdom as is evident from the Delhi-Siwalik inscription.87)
However, unlike the first stage, the existence of frontier chiefs or governors belonging to the Cahamana clan is unknown from the sources on the suzerain stage. The royal clan members advanced conquest with their own troops, however unlike the clan system, the conquered areas were not distributed among them but usually left to the con-quered rulers. As the Harsa inscription dated VS 1030 says that Vigrahardja II subdued the earth by forcing the enemies in all direc-tions to bend to his will,88) and as the Delhi-Siwalik inscription dated VS 1220 expresses Vigrahardja IV's conquest of north India by saying that he made the land between Himalaya and Vindhya tributary coun-tries," their ultimate ideal was to subdue all the rulers of the world, i.e., digvijaya, and not to distribute the whole world among their clan members. Consequently, their state system became a samanta system in the suzerain stage and cannot be understood in the context of the so-called clan-monarchy.
The samanta system was considerably unstable and had difficulties in maintaining integration. For example, the Cahamanas had many battles with the Tomaras, where the formers defeated and killed the chiefs of the latter repeatedly before Vigrahardja IV succeeded in put-ting their forts of Delhi and Asika under his contro1.90) It would indi-cate that state power in the samanta system essentially could not con-trol the samantas perfectly. Thus, the crucial problem the dynasty faced must have been controlling them, that is, state integration. As B. D. Chattopadhyaya pointed out as the mechanism of the samanta system, state integration would have been realised through the ranking system of samanta hierarchy, where seimantas vested with various titles like mandale.s'vara (lord of province), mandalika (ruler of province), etc. rendered services according to their titles as units of the state and raised their title-rank in the central court [Chattoapdhyaya 1983: 219]. It is worth noting that those samanta titles cannot be seen in the records on the first stage of the Sakambhari kingdom, though some kind of conquered rulers certainly existed in their domain. In this
sense, the second stage was the real samanta system in the historical process of the Sakambhari Cdhamana state.
Another and more direct apparatus to control the samantas was the governors who checked them from the strongholds of the frontiers. The governors were often given a free hand in their diplomatic activi-ties towards the frontier rulers in order to take prompt measures against their movements. So, this check system would have worked well as far as the appointees were able and loyal to the kings. The kings in this stage were not inclined to leave such a keystone of the state integration to their close kinsmen who were their potential rivals, probably for fear that they might be connected to rebellious samantas of the fron-tiers. This would have been the most critical situation for the kingdom under the siimanta system. Accordingly, the kings introduced many trustworthy and loyal persons of various clans other than the royal clan members, including members of the collateral branches of the Cdhamanas like Kumarapala of the Vagadiya clan. One of their most reliable channels of recruitment was their maternal connections, as in the case of Kilhana Guhila of Asika fort. It was not unusual that the king's maternal relatives were appointed even to important posts other than frontier governors, as is evident from the example of a trustwor-thy minister and warrior of Prthvirdja III, i.e., Bhuvanaikamalla of the Kalacuri family. Probably, in the second stage, the ratio of the non-Cdhamanas among the staff of the kingdom became incomparably higher and they played much more important roles in the state administration than in the early stage, though the royal kinsmen still took a major part in the military apparatus of the empire.
Even in the suzerain stage, they appear to have something like a clan ideology. In the Delhi-Siwalik inscription, Vigragharaja IV addressed the order of the world conquest only to his clan members (samteinajiin=
atmanah).91)
This probably indicates that the conquest was consideredto be the collaboration of the Cdhamana clan members. It is not be-yond speculation, but the royal clan members including the kings them-selves might possibly have viewed their kingdom as their common property. This kind of ideology, i.e., joint rule of the royal family, appears to have existed in the Naddula kingdom, too [Mita 1999: 103]. The difference from the Naddula kingdom appears to have been their territorial scale, which led the kingdom to a full-fledged samanta
sys-28 Journal of the Japanese Association for South Asian Studies, 15, 2003
tern where the importance of non-Cahamana clans and the royal ma-ternal relations considerably increased.
But however dominant such clan ideology was among the royal clan members, potential rivalry among them always existed as in the clan system of the later periods. So-called "clan disorders" in the later clan system often meant the clan members' secession from the central power. As they had their own hereditary territories, they could secede from the kingdom of their chief and establish new kingdoms in the territo-ries. However, in the case of the Sakambhari Caharnanas, there is no record indicating that one of the royal clan members increased his power and his "principality" became independent of the central power.
"Clan disorders" in this kingdom were all the usurpation of the throne by them, for they usually could not form their own principalities be-cause there was no assurance of heredity of their revenue districts and villages. In this system, only samantas, especially powerful siimantas, could form principalities and break away from the central authority. That is why the kings would have worried most about rebellious royal clan members connecting with powerful samantas of the frontiers. Thus, although the system of the Sakambhari Cahamanas was certainly clan-based, it was considerably different from the so-called clan system of later medieval Rajputs. The significance of its historical change would be clarified by enquiring into the historical position of the samanta system, that is, its difference from the state systems of the Delhi-Sultanates and the Mughal empire and the mechanism of change into the latter, which should be one of the most crucial topics of research in pre-modern Indian history.92)
Abbreviations
ABORI: Annals of the Bhandarkar Oriental Research Institute. ARIE: Annual Report on Indian Epigraphy.
ARRM: Annual Report on the Working of the Rajputana Museum, Ajmer. CII: Corpus Inscriptionum Indicarum, Archaeological Survey of India, New Delhi. EI: Epigraphia Indica.
IA: Indian Antiquary.
JASB: Journal of the Asiatic Society of Bengal.
JBBRAS: Journal of Bombay Branch of the Royal Asiatic Society. JPASB: Journal and Proceedings of the Asiatic Society of Bengal.
Kr.: Krtyakalpataru (K.V.R. Aiyangar (ed.), 1953, Krtyakalpataru of Bhatta Laksmidhara. Baroda Oriental Institute, Baroda).
LP: Lekhapaddhati (C. D. Dalal and G. K. Shringondekar (ed.), 1925, Lekhapaddhati. Baroda Central Library, Baroda).
PRAS, WC: Progress Report of the Archaeological Survey, Western Circle.
PV: Prthvirizja-vijaya-mandkavya (G. H. Ojha (ed.), VS 1997, PrthvIraja-vijaya-mahiikiivyam. Ajmer).
SP: Sodh-Patrika.
Notes
1) R. G. Fox re-interprets this system in the context of the segmentary state theory [Fox 1977]. But his "re-interpretation" is not beyond a usual image of clan system depicted by other scholars.
2) In this system, conquered land was distributed to the sons and brothers of the king, and their hereditary holding of land was sanctioned. Each landholder ap-pointed his officers and could freely exercise his political power within his terri-tory. Their territory was considered to be their share of the collaborative activities in which the ruling clan members engaged. Therefore, the basis of their landhold-ing sanctioned by the klandhold-ing rested on the fact that they belonged to the rullandhold-ing clan rather than merely enjoying the favour of the king. Although territorial holders accepted the suzerainty of the head of the ruling clan (i.e., the king), they treated him simply as primus-inter-pares, and even the choice of the king had to be confirmed by them [G. D. Sharma 1975; 1977: 1-24].
3) So far there have been very few intensive researches about the state system of the Cdhamana dynasties, especially not at all on the Sakambhari kingdom. Even in the monumental study by Dasharatha Sharma, Early Chauhan Dynasties, he de-scribed the "state system" in the chapter of "Chauhan Institutions" in a trite style of the dynastic historiography like the topics of kingship, royal court, administra-tive division, official designations, revenues, etc. [D. Sharma 1975]. A.B. L. Awasthi and R. V. Somani also adopt this kind of itemization of topics, though they collected much more valuable information about the samanta system of the Sakambhari kingdom from their epigraphical records and the literary sources like Prthviraja-vijaya-mahakavya [Awasthi 1968; Somani 1981]. However, such item-ization of topics might be suitable for the institutionalized state like modern ones rather than the samanta formation of the early medieval kingdoms. Besides, in Sharma's study, the data of various Cahamana dynasties are mixed together and analysed as "Chauhan Institutions". As for this methodology, other dynastic re-searches on the Cahamanas like R. G. Singh and Anita Sudan too, are totally the same [Singh 1964; Sudan 1989]. However, the Cdhamanas of Sakambhari, Naciclala, Javalipura, etc. formed independent kingdoms, which is apparent even from the fact that the origin myths of their lineages formed in each kingdom were consid-erably different from one another [Asopa 1976; Chattopadhyaya 1976: 66-68; N. S. Kapur 2002: 238-42]. Even if they were actually interrelated one another in their various aspects of activities and sometimes came into alliance each other, that never means that we are allowed to mix the data from these kingdoms to describe their political system. Such methodology would lead us to some kind of fictional images on the Cdhamanas. Therefore, this paper will be the first attempt
