311

CHAFIER X

THE AUTHORS

REVIEW OF THE TIL^DITIONa L ACCOUNTS

In this section an attempt Is made to reinterpret the accoiints in of colonisation given the Pali Chronicles namely the Djpavamsa I • and the i^havarosa in the light of archaeological data. This becomes necessary because all the traditionsd historical views are invariably influenced by these Chronicles which talk about the coloniaation ot Sri Lanka by the Indo-^ryan language speak­ ing Sinhala settlers in the 6th century B.C. Before analysing these legends it is relevant to present some observations re­ garding the value and nature of these Chronicles,

Law (1947 * 43) says of these sources that "the Chronicles of Ceylon were all written or compiled by the Elders whose pri­ mary interest naturally lay in the history of Buddhism and

Buddhist foundations. Although the religious motive finding its expression in the edification of all things connected with

Buddhism predominates over everything else, there is really an interplay throughout of two distinct motives, devotional and patriotic. The first of them may be construed as puritanic and sectarian and the second as national and racial. The

Chroniclers who were mad with the ieda of the Indo-Aryan rule 31

did not foresee the difficulties to be met by the modern historian.He concludes that "the main theme of the Dlpa- vamsa is the conquest of Lanka both politically and culcurally.

The Mahavamsa amplifies and supplements the Dioavamsa. . This apparently shows as to how these works are really meant to serve an ideological function which the author of the Mahavamsa is frank enough to admit at the end of each chapter, the pur­ pose of his compilation being *'for the serene Joy and emotion of the pious". Hence extracting the historical facts from ^

* the ideological bxhuberence becomes our main task.

The two main legends are those of relating to Vijaya and

Pandukabhaya respectively* The classical model of this version of colonisation of Sri Lanka by the Indo-Aryan language speaking Sinhalese often assumes a complete or partial genocide by the invading North Indians on the existing population embodied in the Vijaya-Kuveni legend and the demographic replacement by the invaders. This model also envisages that at that time Sri Lanka was popul

Referring to the earlier setx:lements the Dlpavamsa (C h.IX t

31-36) and one tradition in the Mahavamsa (Ch. VII t 39-^5) attribute the founding of the villages to Vijaya's cofupa- nions in the regions of north and north-western Sri Lanka

(F ig . 50). The other tradition of the Mahavamsa (Ch. IX i 9-12) 313 ir f '

attributes this to the brothers of Bhaddakaccana (the

wife of Panduvasudeva who was the successor of Vijaya),

However it should be noted here that it is only in this tra­

dition that south-eastern Sri Lanka is mentioned. On the

whole the main theme of the Chronicles is the story of Anu-

rad!iapura region.

Scholars like Geiger (bechert 1960 i 4 9 ), Basham (1952 »

163-171), Paranavitane and Nicholas (Ray 1959) while accept­

ing that these stories are overlain with myths, legends and

folk&ore are yet of the opinion that these contain a kernel of truth - Sri Lanka was colonised by the Indo-Aryan language

speaking Sinhala settlers from North India during the two or three centuries before the 3rd century B.C. According to them this is confirmed by the linguistic afliliations of the

Sinhalese language as well as the earliest Brahmi inscriptions

in Prakrit dating back to the 3rd or 2nd century B.C. They

also infer that these people had arrived by sea from north­ west as well as north eastern India in more than one wave.

However one could see a confusion even among the

scholars who accept the version given in the Pall Chronicles

as an evidence for the early Aryan seti;lements. Thus Ellawella

(1969 » 105) concludes that "it can be inlerred that there

were several streams of immigration from different parts of

India, spread over an extensive period of time ... But the

full authenticity of this account is doubtful, for it is 314

confused by a number of traditions unsklllfully blended. It

Is therefore, difficult to ascertain the types and the charac­ teristics of these settlements from the Chronicles", (Nicholas

(19!?9 t 8) says that "it will probably never be resolved satisfactorily why the i^uth Indians did not resist the North

Indian colonisation of 5ri Lanka or if they did, how that resistence was overcoffle.

On the other hand Senaratne (1969) has dismissed this story by saying that there does not seem to be any evidence, either historical or archaeological, which could substantiate this story, Mendis (1965 t 263-279) however concludes that the

Vi^aya legend, according to the evidence available, cannot be taken as a historical account,and also is not the story of the first Aryan settlement as it actually took place.

In his analysis he has admirably shown as to how besides the inspiration from the Jataka stories, the peopling of

Tamil Nadu by the Dravidian speakers, (speaking a different language from that of the Sinhc.lese) with whom the early

Sinhalese had a closer contact, would have also contributed to the belief that the Sinhalese had come from North India only.

The legendary nature of this story of the first i^ryan settlements becomes obvious when one examines the legend more closely. Mendis who had exai&ined the origin and the growth of this legend is convinced that it is not even earlier 31b

than the 1st century B.C. According to the Dipavamsa (Ch. IX i

5) the provocation for the Vi;Jaya story was the name Sihala and thus it says; "This was the island of Lanka called

Sihala after the lion. Listen to this Chronicle of the origin of the island which I narrate" The Fiahavamsa (Ch. VII i

42) elaborates this and adds that the king i>ihi»bahu, since he had slain the lion was called Sihala and by reason of the ties between him and them, all those (^followers of \Ti;)aya were also called Sihala".

The word Sihala is hardly mentioned in the Chronicles,

We find a solitary reference to it in the Dipavamsa (Ch. IX :

5) in connection with the story of Vi;)aya and here too Sihala is referred to as the name of the island. In the Mahavamsa. this name occurs twice: once in Vijaya*s story (M.V, VII : 42) and next in the account of the king Vattagamini Abhaya (89-

77 B .C .) , where he is described as 'Mahakalasihala* *the black lion* only (M.V. XXXIII t 4 3 ). Thus in respect of these three instances, the first two occur within the Vijaya legend and the third is only a comparison of the king with a lion. But in the Pali Chronicles the most popular names to the island are Tambapanni (Tamrapami) and Lanka. Actually

Sihala or Simhala became popular only in the later Pali and

Sinhalese writings. £ven any of those earliest Brahnii inscriptions of Sri Lanka do not mention tl:iis name (Appendix I). 31G

The earliest reference in Indian literature to Sri Lanka as Sihala (ijiuihala) is found in Cilappadikaram (Swaminctha

Iyer 1920 i i>9» 80,91)» a Tamil epic and in the Diwavcidhana. the works of about the 3rd century A,D, Similarly it appears in the earliest Indian epigraphical sources also around this time (Mendis 19t»3)* Thus all the available evidence seems to confirm that Sihala (Simhala) in this form was not originally the name of the Island or the people and around 3rd century '•I A.Dj only the island derived it from some other old name

Ilam about which we shall discuss later, 0

The above analysis also shows that there is no evidence to indicate that the first Aryan settlers gave the name Sihala

(Simhala) to the island. Had they done so they would have been known by this ntjise from the earliest times. Hence the story which emerged five centuries after the first settlement is evidently unhistorical,. It is not likely to provide re­ liable evidence as to how, when, and why the original 'Aryan' settlers came to the island. This legend arose to account for the name Simhfc.la which was applied to the island then and was produced by one or more Buddhist Elders based on the knowledge of India and Sri Lanka they possessed at that time and limited by what they thought and believed with regard to human natiare and human activity. The landing of Vi§gya on the day of the Purinibbana of Buddha and the artificial nature of the legend, however, whows that the Vi;)aya myth should be seen not only as a fictional myth of origin but also 317

as a myth of legitimisatlon vls-a-vls Buddhism, the religion

of the writers of this history, and parallels in many respects

the Agastya and Par:^surama legends of bouth India and the

Kaundinya legend of South-ii^st Asia. This apparently implies

a cultural infiltration which had been wrongly attributed to

a large scale migration of people. A thorough assimilation

of cultural traditions by the local population need not imply

and did not require a fullscale Indo-Aryan migration.

The above conclusion becomet more and nrare eaqjlicit when

one analyses the second cycle of legends known as the Pandu-

kabhaya legends of the Kahavamsa (Chs, VIII,IX & X). The two

independent legends are connected by a personality named

Panduvesudeva who is introduced in the Chronicles as the brother’s

son of Vijaya and the l&i>ter*s successor to the ttirone in £>ri

Lanka. The Dipavamsa (Ch* IX) describes the story in fifteen

verses. Almost all the aspects of this version are given in

greater detail in whe Mahavamsa, Nevertheless Mendis (1956 :

81-84) views this legend as an attempt of the Chronicler to

connect the successors of Vijaya with the ir'andavas, the heroes

of the i-'lahabharata on the one hand and on the other with the

Sakyas, the clan of Buddha. Lastly commenting on these legends of alleged Aryan migration, Susantha Goonatilleke (1980) says

that it is clearly mythological whilst the factual period of

the Mahavamsa begins with the era after the introduction of

Buddhism. 31b

It Is now opportiine to turn to archaeological sources

and see how far they confirm the legendary migrations of the

Chronicles, benaratne (1965 * 8-9) was the first scholar to make a survey of the places mentioned in the Chronicles. He

sui^eyed a triangle bound by Puttalam, Karaitivu, and Tobbova

and concluded that "in many instances the location of the site was very d ifficu lt. At others there were no surface indi­

cations which invited even a trial dig. In yet others such

remains as were visible related to much later periods. It is

extremely doubtful, therefore whether any sites of this type

could be made the starting point, for a study of this period •••

Certain regions of the south east sector were similarly

examined. This was somewhat cursory but the initial indica­ tions were no different”. And finally he says that in parti-

cular, there are no finUs which could be traced to either the

West or the nast of North India (Senart-tne 1969 * 31)*

Conversely our study of the data from the arcaaeological,

anthropological, sociological, epigraphical and the linguistic

sources to be discussed below, however^ shows that the legends

of the proto-historic period may be given a fresh interpreta­

tion besides what has been discussed so far. First, inspite of

all the embellistiiiients as mentioned earlier tnese legends

grew to account fox the early settlements in the Island,

Surprisingly enough the settlement areas mentioned in the

legends more or less coincide with the zone of the megalithic

settle ents as well (Figs, 1 & 5 0 ), Secondly, a closer 313

scrutiny, alaova tnat tnese legends embodied laany craditions interwoven together. Besides the traditions reflecting the religious and mythological aspects discussed earlier* there also lies yet another identifiable tradition in .hese legends which is cez^ainly factual. This is che tradition of Pandu, the Pali form of Pandya, Appellation 'Pandu* has been taken for granted as a reference to Pandus or Pandavas the heroes of the liahabharata by naloney (1968) and others(Kandis 1956). On the contrary, the continual usage of this term in the Pali

Chronicle^.' of Sri Lanka to denote the Pandyas of

(Nicnolas 1939)» demands a fresh approach to this problem.

In this context the observations of Nicholas (1959 t 7-8) merit serious consideration. He says that "In the legendary be, innings of the Sinhalese civilization related in the early chapters of Mahavcimsa. the kingdom of i: andu (wtiich n^ime is consistently used for Pandya both in the i''iahavuinsa and Cula- vamsa) is said to have been in existence in the 6th century

B.C. and the leader and chiefs of the first Indo-Hryan arri­ vals in Ceylon are said to have sent envoys to the Pandyan king and obtained wives, craftsmen and material assistance from that kingdom." The pez-sistence of this 'core* in the legendary history of the Sinhalese, inspite of the Chroni­ cler's awareness to their differences with the Tamil speakers of the Psmdyan coast makes one to t^ive a most serious thought to this aspect. 32u

As mentioned earlier, in the Chronicles the 'i'andus*

are equated vrith the Fandyas of Ta il Nadu. I'he derivation

of the term *i-andya* itself shows that this explanation holds

good linguistically as Pandyar is derived from Fandu, means

old, ascending on uhe respectable suffix ^ (Subrahmanian,

N. 1966a t 547). Jciven the name Palayar, for a branch of the

PandyciS means the same. Apart from the Vijaya legend, it also

occurs in the names of the successors of Vijaya as well. For

instance woile the Diuavamsa (Ch. X t 2) mentions the successor

of Vijaya as *ianduvasa'; the /lahavanisa (Ch. VIII) calls him

as *Panduvasudeva. The other names occurin^ in the Pandu-

kabhaya legend of the i’iafiavamsa such as Pandusakka, Panduka-

bhaya and 'Pandula', are again apparently of the same impoi--

tance. The landings place of Paxiduvasudeva and Baduakaccana,

again according to Geiger (I 96O ( 63t 64) lay alon^, the coast

' of hannar area. Thus the geographical proximity of this'region

to the Pandyan coast is significant.

i>imilarly the term "Tamba" occurs as a prefix in the old

name for the island of Sri Lanica. A river bå the same

name also has been located on the Pandyan coast. In the early

Pali texts the term used to denote the Pandyan territory,

however, remains to be Tambarattha (Venki. tasami, S . 1950 *

63) . The occuirence of the term i’amba as an intial word of

places in south eastern ^ri lAnka has also been noted by

Parker (1885)* In this context, it is rather opportune to

take cognisance of Parkers (1865 : 18) observations. While 3k;i

referring to Tlssamaharaina he says that "I siay mention, also, that from four to five miles distant from the Tlssa dagabas there is a lank called the Tammanna wewa, through which there runs a stream termed the Taiiunanna-aru • (a Dravidian name in the south east of Ceylon ! ) • This stream Joins the

Kirinde - i,ange four miles from the Tlssa ruins. The name

Tummanna is so coiomonly applied to tanks and rivers in Ceylon that this fact cannot be considered to throw any litht on the ancient name of the city* 1‘ambapanni being originally a ::;>outh

Indian name, it is quite jpossible that this appellation was bestowed on the district long previous to Wi^jaya's landing".

The above observation takes us back to the another aspect of the early colonisation, a s said earlier, one of the vei-sions of the rianavamsa mentions only two settlements viz:

Digtiavapi and hohana in south eastern Sri Lanka. Besides this no further details are available about the region during this p-riod, for the main purpose of the Chroniclers seems to l;iave been to deal in detail only with the history of the Anu- radhapura region. However, the reler nee to the Kshatriyas of Kataragama, and Candagama in Lhe account of the Ftahavomsa

(Ch. XIX t 54) in connection with the planting of the Bodhi tree in Anuradhapura is significant, x'he br. hmi inscriptions of the Kshatriyas show that fish was their dynastic symbol.

A reference to f'ta.ihi-mara.ihaha (^aranavitane 1970b. No. 421) which is in lact an equivalent ol Fanayan title iKXnavan shows 322

that the early settlers of this region too hailed from the

Pandyan coast. This is also true in the case of the extreme north of iJri Lanka, wnere Punch-marked coins with fish symbols have been discovered (Pieris 1919 * 5 0 ), Thus the Pandyan element in the account can be interpreted as an arrival of an earlier wave of people rei-ching Sri Lanka from the oppo­ site Pandyan coast.

In the Mahavamsa (Uh, VII t 49-51) Vijaya is said to have sent messengers to the Pandyan king who had his capital in southern Madurai ( Dakkhinam Madhuraa ouram). This has been interpreted by the scnolars who saw a kumal of truth in these legends as a reference to a city of Matnura in northern India.

This seems to be an untenable suggestion for the reasons sugg­ ested below, tirsc oi all no such southern Mathura has been identified in north India. The reference in the Mahavamsa is actually to the Pandyan i-adurai and not to Mathura , v.hich was not a seat of political authority during this time.

Secondly in the Xaroil sources thexe is a reference to soutnem

Madurai in tne Pandyan coast where the first Tamil academy was convened. The existence of southern i'^tadurai in the 7th cexitury B.C. is confixmed by the recent study of komila Thapar

(1975) who nas identified udal of the cuneiform texts with the early Pandyan capital which had also another name Kudal. This, together with their second capital Kavatapuram, lay in the

Tamrapami delta of the Pandyan region (Maloney 19t>8), hadurai. 325

the present city is only the third capital of the the Pandyas.

Lastly the C-1A date for Korkai going back to 8 century B.C. is s i ^ ^ f i c a n t , lor Korkai remained to be a sea port capital of the Pandyas who had their capital in southern Madurai, All this showsthat the Pandyas had an earlier beginning in the far south of Tamil Nadu, unlike any other dynasty of this region,

Megasthenes refez'S to Pan'^ia. the daughter of Heracles as a queen of the Pandya teriitory. This iias some sim ilarity to the legends of >illiyaracani, a Tamil queen of northern

Hri Lanka who anciently ruled from Kudiraimalai on the north western coast (iiornell 1922 t 4-16), i‘loreover there does exist a common flood tradition oX deluges between these tvo regions. This is mentioned in the Tamil as well as in the

Pali sources (Kaiiapatnippillai 1948 i 216-228), About this

Haloney (1968 X 293) observed that **These legends of the sea encroaching uuon Cape 6amorin» old Madurai, Kapatapuram,

Kaveripottinam, i^ahabalipuram, Kalyani and Kudiraimalai

(the last two places are in Sri Lanka) and others not mentioned hexe are no doubt based on facts to some extent. A number of such events have occurred in the present century". The OK>dem parallels, however, show that these legends do reflect the folk memories of cin earlier tradition common to both the

CO: sts of Palk straits. The geological study of this region

(Dtraniyagala, P,E,P. 1956b t 127-142) has further confirmed the evidences of the literary tiaditions. 324

All this shows that there was a co. inon base of cultural stratum between the two regions. Perhaps the evidence for the early emergence of the Pandyan Kingdom and the development of language, literature and script in the same region, indicates that the inlluence radiated from that region only rather than from any other parts. The control of the Pandyas even on the trade of this region has been commented upon by scholars, Nicholas (1959) is of the opinion that since the Pandyas were the closest to ^ri Lanka in all probability they established trading contacts with

Sri Lanka from their early years of existence. On the other hand Perera (1952 i 196) speculates that one of the ancient names of ^ri Lanka (Tambapanni) appears to have been the result of close commercial connections between the two count­ ries. This status quo continued even during the early historical times (Silva, K.M.de. 1977 * 37).

The above survey indicates i^w the legends regarding the the proto-historic period, which represent a cluster of traditions put together, can be given a new interpretation in the light of the archaeological and other evidences. Of these, the tradition rei^resenting the Pandu element is signi­ ficant, Perhaps theie were other elements as well, about which we shall deal later. However^ the strut>fele of Pandu- kabhaya against his uncles could be interpreted as a struggle between the elements representing the Pandu tr<^dition and that of the proto-blu element for the control of Anuradhapura 32o

region. This aspect is being discussed in the linguistic section. Xhe conspicuous absence of any reference to the

Pandu element with the of Mutasiva, shows that this group was slowly eclipsed by the other elements representing the proto-Lilu group. A parallel development took place in south-east Sri Laixka as well, where the Kshatriyas of Katara- gama representing the Pandu element held sway. This is embodied in a story of Mahanaga, the brother of Devanamijiya Tissa.

Mahanaga is said to have fled from nnuradhapura and established a hold here. 6oon after during the time of his successor, the relations between the two groups became hostile and it paved the way for the eclipse of the Kshatriya j^^wer in the region. As far as the extreme north is concei-ned, the proto-

EXxx elements seem to have had little control over this region, for it is mentioned only during the time of Devanpiya Tissa.

During the time of his successors for nearly three centuries there is hardly any reference to this region in the Chi'onicles

(Pieris 1917 s 1 1 ). This indiciutes that there were other elements such as Pandu which were in contr*ol of this region.

Al^ihKO^-OLQGIJAL LyiuixtiCE

The study of the Physical remains is a more direct way of analyzing the population, a s discussed elsewrire Kennedy's

( 1975) study of the human skeletal remains of the megalith builders of ^outn India also included the human material unearthed from Poa, arippu, (Fiti. 51). ^-he ^omparippu evidence showed that it was a part of the iSouth Indian population of w r r V, ' ■------I ■ ^ 32b

megalith ouilders. Similar view has ueen expressed by Cha-

nmugara and Jayawardane (193^) in their study of the human

skeleton discovered at Tiz*ukketiswarain. They observed that

these remains could v^ry well be classified as those of a

modem South Indian type (PI. VC). Howeve^ Kennedy (1975)

places ti:iis burial to the late iron age of iiiri Lanka.

Later Kirk (1976 t 91-99) studied the blood groups of

Sinhalese* Tamils, and ocher Indian population recounted

in the Vijaya legend, while acknowledging the limited souice

material for his study he observed that "a sufficient number

oi' studies on the distribution of genetic maktrs among popu­

lations in Sri Lanka and India during the last two decades

enable a worthwhile genetic distance analysis to be made....

The main conclusion which emerges from these individual compa>

risons, however, is that north west India and the more extreme

north seem to be excluded as a major influence on the Sinhalese,

whilst the north-east and the south are genetically much more

closely related to them ...." Moreover he adds that the gene­

tic distances indicate that the Sinhalese are closest to the

Tamils, followed by the Keralities for two of the indices, whilst the other shows the Sinhetlese to be closest to Lhe

Bengalis with the Tamils slightly further away.

He concludes that "the colonisation of Sri Lanka itself

probably resulted from population pressures in the agricultural areas of neighbouring India, bjqjansion from South India would have been relatively easy and durin^s the historical period records indicate that a lively intercourse took place between populations in i>outh India and £>ri Lanka. The genetic distance study agrees with this in giving low values for the distance between the Sinh^lese and Tamils and Keralities. The lowest distance value for the Sinhalese however, for F 0 is with the

Bengalis, but the differences in genetic distance between the

Sinhalese and Tamils on the one hand and the Sinhalese and the Bengalis on the other are not great. Indeed Bengalis are closer to the South Indian populations included in these compa- risions than they are to the other Indian populations in the north and north west. This aspect of Bengali relationships will be explored in more details elsewnere..

Wijesundara (1977) examined the Sinhalese-Tamil relation­ ship in depth in 1977. He remarked that "Amon^ the determinable single gene effects in humans, are the abnormal haemoglobins and blood groups, Tambi Pillai, Seneveratne and Abeyratne in

1972 foimd an equal prevalence of ab D and E in Sinhalese and

Tamils, ABU blood group distribution also shows no statistically significant difference."

The above survey clearly indicates a close physical affinity between the Sinhalese and the Tamils and also shows how they are closely related to the South Indian populations, the descen- dents of the neolithic-megalith builders of South India. This is amply corroborated by the other common elements in their cv|jLture. 32S

SOCIOLOGICAL

The theoretical superlmposltlon of the fourfold caste system never changed the basic Sinhalese social structure which was essentially i^ravidian. Ananda Coomaraswamy (1936 t

21) observed tliat the caste system of Sri Lanka differs from the wellknown fourfold system of the Hindus and is similar to the Dravidian system in South India in which the culti­ vator is ranked as the highest, Bryce Kyan (1953 * 17-20), a sociologist, while endorsing this view, remarked that

"recognition of the South Indian structure makes less sur­ prising the fact that among Sinhalese the highest caste is also the largest". Ihe Sinhalese society is broadly divisible into agriculturalists (kulina) and non-agriculturalists China) .

(Bechert 1960 t 32). This social stratificational evidence hints at the Velir (Dravidian agiiculturalists) orientated

Sinhalese society about whom we will discuss later. However, the term Velir disappears by the beginning of the Christian era and is replaced by Prakritised Goiaama which denotes the s . me

The existence of a common basic kinship terms in both the

Sinhalese and Taaiil societies is again an indication of the basic oneness of these two cultures. Most of the anthropo- gists who have worked in South India and Sri Lanka are agreed that the kinship terminology of this region has some important general characteristics in common. ^Kanre (1968 t 39&*'^12) who 323

studied the kinship system of India, came to the conclusion

that the Dravidian kinstiip terms are patterened differently from the nortnern sanskritic system, Now, let us see the important characteristics of the Dravidian systerji. In the

Dravidian system there are primary terms for three generations of imiiiediate relations and the terras for one generation are not exchangeatole for those of another generation. All the other terms are derived from these primary terms i . e . father's brother receives the same appellation as the father, and his wife is called mother. Their children are called brothers and sistexr* This set of relations are completely identified with own parents and siblings respectively.

The same could be said of the Sinhalese itinship system,

Wijesekera (19^9 * 62) observes : "The Sinhalese kinship terms are designed to suit a classificatory system in which persons who stand in the same relationship to another are addres;Led by the same term by him. For example the mothexs

sisters will all be called by the same term by her children.

But for purposes of identification and convenience of refere­ nce in their absence various descriptive epithets are added to the names. They are the names of the villtige or adjectives like small, medium, big, light, fair or dark. Sometimes also a petname comes to stay,** A few examples would suffice to illustrate the point. In Tamil father is known as *Aopa* and in as Achchan. In Sinhalese Apuchcha, Appachchi. 330

are the terms used to denote father. In Sinhalese, father's

elder brother is called as Hahaopa or Loiuaupa and the younger

brother r^udappa. flalappa or Bappa. In a similar way in Tamil

Perivappa and Kunciaopa are the terms used to denote the

fathers elder and the younger brothers respectively. Like­

wise the same terminology is used to denote the mother and

her sisters as well. In both the languages the mother's

brother and father-in-law as well as father's sister's husband

.are called as Mama (uncle). This last mentioned example brings

out an another salient feature of the Dravidian kinship termi­

nology, for in the northern system nobody who is a relation

by blood of an ego could be a relation by marriage also. The

example of uncle (Mama)becoming the father-in-law in the Dravidian

system, of cross cousion marriage is one of the basic concepts

of a matl:ilineal system and underlines the basic oneness of

the Sinhala and Tamil social structure, as noted elsewhere

(Ch,VII) the megalithic evidence points to an important fact

that the economy and technology of Sri Lanka in the early

phase was the same as in South India. The Kinship system, it

should be noted, has an ultimate association with the pro­

duction system and so is conditioned by it .

i!.vidence of the Folk and Syncretic Relations Practices

The conquest of Buddhism which took place in 3rd century B.C.

could not totally eliminate the Austro-Asiatic cum Dravidian

primitive animistic religion. The claims of the Buddhist monks

for the prevalence of pure form of Hinayana buddhism in Sri

Lanka can never be justified, for since the time of its 33i

introduction it is only a syncretic religion (Bechert 1978).

Becau.^e of their faith in the Vijayan legend and the Aryan migration scholars like Sarachchandia (1966) attributes its first introduction to the Aiyan ii^igiators and consider the

Dravidian parallels as a later wave. But a careful analysis and correlation ol these elements points out the auoriginal character and iriomo^’tnlty a;aong pre>BuddcJ.stic .Sri Lanka and pre-Brahiaanic bouth India.

Various terracotta objects of mother goddess, lingas. animals such as Iriorses, elephants, and other votive offerings in the vicinity of the uanks as at Tobbova, Maruda Kaduwa, i’algasv.-;va near Arippu, Urutirapuran near Kilinochi, Pansal- vava near Vavaniya, Ilukuvava near Sigiriya, Uttikulam near

Mihintale and Pandr<11a wewa (Deraniyagala, P.E.P. 1960,1961) and few places in AmparlLi district (Deraniyagala, a 1972 : 164-

165) have ^een unearthed at various times. These clearly show that the fertility cults had an earlier origin and were associated with pastoral cum agricultural society reflected in the proto-historic periods of both Sri Lanka and South India,

The Pre-Brahmanic 2 eligion of iiouth India is a conglo- maration of ^astro-Asiatic and iJravidian elements which cannot be distinguished now. This is also the case regard­ ing Sri Lanka, where Austro-Asiatic rituals were mixed with inegalithic iJravidian elements before the arrival of Buddhism. 33^

However, the ir^ail Chronicles spec^k about two cradltions,

Naga and Yakka which prevailed in iari Lanka before the mythical arrival o£ the Aryans, It seems that the Naga cult was readily syncretised with Buddhism probably even before its arri al in <^ri Lanka as no lOiiger the Buddhist works speak about it as a separate cult.

In the ^ri Lcmkan Brahmi inscriptions theire are many references to Buddhist monks and kings with I4aga names (I'ara- navitane 1970b). Even the first uuddhist emissaries came through the north of ^ri Lanka which was then known as Naga- dipa and the Maiiavamsa never misses any opportunity to quote the Naga enttiusiasm in receiving Buddiiism, Contrarily not even a single Yakka name is to be found in the list of Buddhist kings and monks. Further, the Buddhist works always treat

* the word Yakka indifferently. Curiously enough at present the term 'Yakko* in Sinhalese is used to scold a person. All these hint that the N^:.gas mentioned as a non-human aboriginals in the hahavamsa were more sophisticated than the Yakkas and they were the people who enthuisiascically accepted Buddhism. The occurrence of Naga suffixes to the names of early kings of iiri Lanka, indicates the political power they enjoyed. The

Naga tradition is found in the Sangam litei'ature which men­ tions certain Naga kings, poets and lands (Pillai, K.K. 1963 s

20-22). We may say with some caution that the term Naga in

Sri Lanka stands for the megalithic Dravidian element and the Yakka for the Austro-Asiatic descendants.

As Is evident from the Pali Chronicles, anlml&m was che original religion of the Island, for which identical trends can be traced in the Sangam liteiature (bubrahmanlan 19b6to t 350) and among the survivals of the i>outh Indian tribals. The ori­ ginal spirit worship forms of Sri Lanka wtre common to that of bouth India in the model of Poota worship mentioned in the

Sangam literature (iSubrahmanlan 1966b s 316-62). These Poo las were the megalithic goda (Fig, 49) and they still survive amon^, the lower class co-iununities of western j^hats, both in

Sri Lanka and South India such religious practices were assi-

::iilated in a similar manner into Buddhism and Bnahminlsm respec­ tively, These rcligio-magical rites wiiich are comnion lor

Sri Lanka and South India form the necleus of the folk art.

Unfortunately the scholars who have worked on Sri Lankan folk art consider it as an outside influence imported either from

South India or North India, Actually this fox-m of folk art and witchcraft are aboriginal in character and are the sur­ vivals of a common religious stratum. This is illustrated by

Nadarasa (1968 : 421)*"There is in the Sinh-lese areas a kind of ritual dance known as devil dancing or liali Tovll. Tovll being a corruption of the Tamil word i'holil which has been flourishing in the villages for the purpose of curing dis­ eases and warding of evil effects such as Asvas and Katavas as in Slntiala (the navuru and kannuru of the Tamils) coming down from early times. This dance has its artistic merit and

1 sucial significance. This reminds one of the yeriattu

and xheivameri Hdu;u Kuttu found in CilaoDadiKaram and Sangam works. These are religious dances to propitate gods and Yakkas.

This dance is invariably connected with magic and witchcraft which during the passage of time has influenced other types of

dances of the Sinhalese and even contributing to the develop-

laent of ^inhala tradition of dances ... This type of dancing > like the binhala and Ta.7iil nadagam goes on throughout the

night and ends with mimetic vicarious nuances by the villagf^^ priest (who is quite different from the liuddhist clergy) who

exorcises devils who have aiflicted the patient with various

disceases. Xn this type of dance sometimes dialogues take

place between the priesxs and the demons who appear with hedious

* masks depicting various Yakkas, the Bhutas found in Cilaptaadi- kaj am."

Another example is given by Sarachchandra (19t>6 t 92-93) who describes ^okari as follows: "Sokari bears the stamp of being one of the earliest folk plays that we have. In some basic features, iJokari is si.ilar to the Takil Kuttu. but tnis

siifiilarity is noticeable in Kolam as well and points only to the close connection that seeros to have existed, even from

early times, between the Sinhalese folk culture and the folk

culture of the Tamils. One is tempted to suggest that Sokari was originally a ritual drama enacted for purposes of promoting fertility and general prosperity for crops and cattle..,*' However, wfciile tracinfc the origin of this word to Hinoustani

Chokadi or Chokari. Kt further adds that he would be happier

if one could trace it to some ramil word rather than the more

' remote Hindustani, Here our view is that it would be appro­

priate to derive the word from Tamil Cortcali like the other

parallel Kathakali of the foik art«

Buddhism replaced ii^any of these folk traditions and some

of them remained in an assimilated form. Evidences point out

that the earliest Buddhist shrines were constructed on the

very sites of the animistic spirit dwellings. To quote

barachchandra again s "An inscription of the 10th century states

that the Issurumuni Viriara was situated by the side of the Tissa

t&nk, the water of which formed the dwelling place of the spirit

(rakus) who were converted by saint r^aninda and was made to be

of service to the religion as well as to the world, hahasena

in the 4th century A . D . is said to have constructed a Ceitya

on the spot where stood the shrine of Yakea K.alavela and the

Thuparuma (The earliest Stupa oi ^ri Lanka) itself was situated

in the precincts of a s.jrine dedicated to ^iahe;}a" (barachchandra

1966 I 3)* Similar observation has also been made by Hahula

(1936). There are I0 places said to have been sanctified by

Lord Buddha during his legendary visits - they are Katara^ama,

Tissamaharama, (south eastern Sri Lanka) l^iahiyangana, and

Mudiyangana (both in ii^dula d istric t), i^umanakuta (Adams Peak)

Divaguha (Ratnapura district), Kalyani (Colombo district). 33G

Nagadipa (Jaffna district) and 7 places In /inuradhapura district, ifaey are Bodtiitree, Maricbvatti cetiya, Maha tupa, i'huparama, Abhayagiri, Jetavana, and Selacetiya (Bechert 1960 :

207). However, it should be noted atiiat all the above places have been associated with the elements of early cultures from the late stone age to laegalithic.

Apart from these folk survivals, when we come to the syncretic Buddhism one could observe many proto-Qravidian gods are placed next to Buddha in che Sinhalese pantneon.

They arrived here in an archaic fonn and later syncretised with

Buudhism, however/they found a dominant place in Tamil Nadu when syncretised witxi £rahmanical gods but in £?ri Lanka because of the dominance of Buddhism they remained as subsidiary gods retaining their arciiaic character.

We shall now discuss the origin and syncretism of tne four main gu' idian gods of the Buddiiist pantheon such as Uppulavan/

Uppuiavanna, god of Kataragama, i^atha and Pattini. Besides them even so many Brahmi-nical gods were syncretised later such as Indra, Ganesha, iiaraswati etc, uuite obviously all the four guardian gods had come from i'amil Nadu, It should be noted here that the pre>Brahix.anical practice in I'ainil Nadu was to assign a god to each geographical division (Bechez*t 1973 :

199-206), Thus we find Oeyon« Mayon, Ventan, Varunan and

Korraval each ruling Kurunci, hullai, Haioidan, Neydal and

P alai. I'hese gods were already in the process of Syncretism 33V

with the Brahmanical gods when the was written,

A l cording to the fiahavamsa(Ch. VII), Indra (Sakka) after being ordered by Buddha, appointed Uppulavanna (Blue coloured) as the guardian god of Sri Lanka, The fact that Uppulavan receivfcd Vijaya and his companions and assui’ed them protection indicates the presence of this god since early times, Uppulavan is none but the god Mayon, mentioned in the as a blue coloured god who lules the forest areas and the favourite of Ay^s (^velebil 1975 s 30-39) It is interesting that these Ayar or Ays are agricultural-cum-pastoral communi­ ties whose presence in Sri Lanka and Tamil Nadu is attested by Brahmi inscriptions and Sangam literature respectively.

About this time oi slightly later Mayon was syncretised with the Brahinanical Visnu and was accepted in Sri Lanka, In the

Tolkappivan we find that Mayon was given the first place in the local pantheon. Similarly in Sri Lankan tradition Uppulavan occupies the place only next to Buddha (Bechert 1973),

The n.xt god of Kataragama occupies the top of the hill at Katarctgama. He is described as a red coloured god in the

Sinhalese traditions and he ranks second to Uppulavan (Bechert

1973)* Identically the Dravidian Ceyon is considered as the god of hilly land in Sangam literature and many of his abodes are found on hill tops. His name is derived from his red 333

colour and M Is ranked second to Mayon. Besides, there are common traditions between Tamil Nadu and Sri Lanka that he took his wife Valli from the Jungle folk, Veddas. This proto-

Dravidian god was later syncretised with Brahmanical skanda and considered as the nephew of Visnu. He is also mentioned as Cewel (the red coloured one) connecting him to the tribe wno are again common to Sri Lanka and i'amil Nadu, The kartika festival which was originally a neolithic pastoral rite and was traditionally associated with this god, is men­ tioned as one of the earliest festivals of the Island (Parana- vitane 1929).

The evidence for the existence of nuruga cult in South

India in proto-historic times is well attested in both literary and arciiaeological records. This is quite evident from the remark of Nilakanta Sastri (1963 t 21<-23) who observed that

"The undoubted antiquity of this cult among the Tamils is attested by the discovery at the pre-historic um field at

Adiccanallur of bronze cooks, iron spears and mouth pieces of gold leaf similar to those employed by modem worshippers of

Murugan when they are on a pilgrimage carrying the kavadi in fullfilment of a vow. The oldest stratum of Tamil litera­ ture mentions Vel^Jiadal, and ecstatic dance by the priest possessed of Velan (iiuruga)" • The discovery of an iron spear at Pomparippu In ^ri Lanka on the opposite coc;St of Adichcha- nallur shows its ecarly migration into Sri Lanka, The form 33;j

o f worship of tnis god In most of the temples even today is with no iconographic representation excepting, the spear. He is offered animal sacrifice and has non-Brahmanical officia> ting priests. These are the survivals of the old form of worship associated with Velan priests. The traditions asso­ ciating the arrival of the god in different places of the island is suggestive of his proto-historic migration and popularity (Parker 1909).

The third god ^^atha could be identified with Siva who is also a red coloured god so that often he was identified with Ceyon and finally became his father. The brahmi records of Sri Lanka refer to a number of persona£,e6 bearing the name Siva. So is the nahavamsa. according to which many of the early kings bore the name Siva. While referring to the reign of Hutasiva, the father of Devanampiya Tissa it is stated in the riahavamsa (Ch. XIII : 1-2} that Mahinda delayed his

;}ourney to Sri Lanka because the time was not propitious enough due to Mutasiva's rule. This is an indirect admission on the * pait of the author about the prevalance and popularity of ttds cult in proto-^iistorlc times. The colt of Siva figures in the

Panuukabhaya legends as well (Paranavitane 1929). So is the

Banyan tr^e in the earlier legends of Fandukabhaya. Trisula. the weapon of Siva, is a graffiti mark on megalithic pottery

(Ch. VIII), So is i^iandipc.da in the earliest epigraphs (Fig.

49). In the Sangam literature Siva is often associated with the Banyan tree. (Subrahmanian I96bb : 336). 34u

i‘he last, guardian god Pattini must have originally iirrived in the iorm of Korravai, the mother goddes who rules the deseit land. As Ceyon is mentioned as the son of Korravai, similarly god of Kataragama is also associated with Pattini.

However, in its present form the word Pattini is connected with kannaki of the Cilapuadikaram and literature in both countries confirm that the i>ri Laxika king Gajabahu I (2nd century A..D.) popularised this cult, The association of this godaes with Veddas ooth in Sri Lanka and in I'amil Nadu points to similar trends. X'he Veddas of bri Li^nKa worship her as

G iri Anuna and consider >^iyya;.an/Kiyanar, another minor god as her son (Pai'ker 190y).

The culmination of this process o. syncretism can best be observed in the festival of i'ooth uelic, popularly known as

Kc^ndyan Perahara, Sarachchandra (1966 : 5) observed that **It is also well- inown that a large nuiaber of tiindu rituals were incorporated into the worship of the iooth itelic. A thing, of recent occurrence .s the transformation of the Perahara at

Kandy into a buddhist ceremony. Originally celebrated in honour of the local deities Natha, visnu, Kataragama and

Pattini, the Perahara consists of a ^^roup of ceremonies connected with the folk belief of tne people, rhis group of ceremonies were amalgamated with buddhism by kin^ Kirti Sri Kaja* simha in 1775, wneii he ordered the sacred tooth to be carried in procession alon^, with the insignia of the Gods” . 34±

Hov>ever^ scanty all these evidences may be, when they are all put together they reflect a cofflmon religious tradition of both the regions of far south of the sub-continent and ibri

Lanka, fhey substantiate the literary and archaeological findings by indicating a longer time span for their origin and development, many centuries before the introduction of Buddhism.

While some of these cults were luter absorbed into Buddhism, others survived in the folk religion of the Sinhalese. Very soon the places associated with these cults were over snadowed by the .:>ymbols of Buddidst worsliiip. o^his really accounts for the existence of early Buddhist btupas in Kataragama, lissa- maharama (in south eastern Sri Lanka) Fomparippu (in north western Sri Lanka) Kantazodai, Vavuniya and >muradhapura (in northern Sri Lanka) and finally Dighavapi (in Amparai district in the eastern Sri Lanka).

LINGUISTIC mPx:^>lC£ '

In this section we have analysed the character of the present day languages of the Island. In our view this would give a clue to the languages spoken by the descendancs of mega- l lith builders. Since the Dravidian affliation of Tamil is well A known, in our analysis discussion on Tamil is avoided. Instead our main concern would be with the Sinhalese language. Because on the basis of its characteristics today it is taken as a evidence for the migration of Aryan language speakers from

Jlorth India. 34^

As the theory of early Aryan colonisation has no ob;)ective histoiical or archaeological evidence, that of ihe North Indian origin of the ^inhe^ese language needs a fresh approach from a purely linguistical angle. i.e« through a historical analysis of the Sinhalese language* At this juncture it may be observed that ..yen the schol^.:rs who grouped Sinhalese within the Indo-

Aryan group, found it difficult to assign a proper place to it in the Indo-Aryan faiiiily, Geiger (1936 s V I), an authority on

Sinhalese language, states that "it is very difficult to fix upon one of the middle Inaian dialect- as tne mother tongue of the Sinhalese or to assign to Sinnalese a definite place among the modem nryan languages of India," Very recently Dissanayake

(1978 ( 51-57) has not only accepted the peculiarity of ’the

Sinirialese language amon^, the North Indian Prakrits but has also drawn tne attention oX scnolars to the existence of a language or languciges in Sri i^anka which may nave influenced the course of development of the Sinhalese language*

I'he main objective here is to find out wnether there was one or more languages in the proto-historic period which helped the evolution of the Sinhalese. 'Xhe approach to the study of the subject could be initiated with the analysis of the names used to denote Sri i-anka, Ihis might perhaps show the linguistic affiliations of the original sei^tlers of Sri Lanka and their role in the development of the Sri Lankan culture. The continual usage of these few names bear testimony to the role of the 343

original settlers in the development of Sri Lankan civilization.

The names that were commonly used in the i^ali Chionicles and the otiier sources to denote the Island are Taibapanni and Lanka.

Sihala (Simhala) came into vogue only in the 3rd century A.D,

The earliest reference to Taunbapanni (Taprobane in Greek) appears in the writings of the Greeks as well as the Pali literature

(Nicholas 1959 I 6-10). The Pali chronicles give a fanciful explanation of the word Tambapanni. xhe Dipavamsa (Ch. IX t

30, 40-42) explains tiiat when Vi^aya and his companions landed here their hand and feet were besmeared with red-coloured dust of the ground, hence it was then named as i'ambapanni. The rtahavamsa (Ch. VII } 40-42) adds tnat since taeir hands were reddened by touching the dust of the red-earth, that region and also the Island were named Tambapanni.

Both the names Tamrapami and Tambapanni which mean copper- coloured leaf or copper coloured hand, have no real meaning unless they are taken as variants of the i^anskrit lorm Tamra- vami(which means copper-coloured flow of water). Tamrapami is a reputed river of the Pandyan coast and the cradle of Pandyan culture as testified by a number of um burials sites like Adi- chchanallur on its banks, which are identical with Pomparippu.

Moreover the pearl industry in this delta region was of remote antiquity in which Sri Laruta had a share. However^ the Tamil sources mention that Porunai or Tan Porunai is the original name of this river, iiccordingly Tamrapami is actually a Sanskrit 344

rendering of Tamil Tan Porunai (Chempaklakshmi 1ffS " *120)

Naloney (1968 i 57) identified Tainbapanni along the Malvattu-Oya

river in the region around Mantai. Others identified it near

Pomparippu. Anyhow according to the Chronicles, the earliest

capital of Sri Lanka founded by Vijaya was known as Tambapanni. ffhis was later shifted to Upatissagama on the banks of the

Malvattu Oya and finally it was shifted to ^nuradhapura during

the time of Pandukabhaya (f^Jicholas 1959)

The Pali Chronicles testify to the presence of the name

Tambapanni when the first wave of the mythical nryans arrived.

The Greek references (Ta|>robane) are for the entire island.

So are the inscriptions of Asoka, We have already seen how the

Pandyan river Tan Porunai was later Prakritised as i'ambapanni

and banskritised as lamrapami. The existence of a common

mebdlithic cultiire has already been establisned. All these

observations prove beyond doubt that there was a cultural

stratum common to Pan iyan country and Sri Lanka which was

named as Tambapanni.

Another parallel with Tamil Nadu too reinforces the

above inference. The ^angam literature mentions a people called

Palalvar (old ones) (iiubrahmanian 1966a i 582) living in the

Tamrapami delta around I'-ladurai and Coimbotore area. According

to Megasthenes people of Sri Lanka were known as Palaeoaoni as

early as 4th century B.C. and the Island produced more gold and and pearls of £,reater size ttian India (Nicholas 1959 : 9) As we 34o

know Megasthenes never visited iiri Lanka and he had written in a North Indian coiirt from what he had heard possibly from i^orth Indian sources. It is equally possible that the people of Sri Lanka w re known as Palaivar. which was rendered by the

Rorth Indian speakers as Falagana and Megasthenes rendered this as FalaeoKonl. t,ven as late as the mediaeval Cola inscriptions the Islands to the south of India (excludini^ Sri Lanka) were kiown as Palantivu (old Islands), However/it is ratner interest­ ing to note that the woid Palava occurs in the early Brahml inscriptions of Sri Lanka (Paranavitane 1970b. Nos. 58 & 159)*

Finally this indicates that the Tamrapami region played a vital role in the early human settlement of Sri Lanka, caldwell (1881 :

10) indicated this possibility. Thus he observed that "it is

•equally possible that the name Tambapanni or Tapiobane applied to Sri Lanka in pre-Christian times was first given to it by settlers in the Island from the Tinnevelly coast i.e. the Tamra- pami region even before Vijaya’s conquest of it.*

Another common name of the Island used in the Pali Chroni­ cles is Lanka which persists even today as Sri l.anka. This name is more popular than Tambapanni in the Pali Chronicles, In Sri

Lanka it appears in the later inscriptions as'Lanka' and'Lamka'

( E . Z .19)2^^ I9i 2.-^7 11333 £ 194-5 )• Here it should be pointed out that the traditional Tamil name for the Island is Ilam which may be an aboriginal term. It appears in the Tamil Brahmi inscriptions as well as in the Sangam literature, and continues 34G

to Che present day. In all probability the word X^anka is derived from Ilam —>>Ilamka — — > Lamka — ?»’Lanka. It may be noted here that triis name Ilam is not foimd in the Chronicles except in one solitary instance as a proper name of a king called I la Na^a (W.V. Ch. XXXV i iwho ruled in Anxaradha- pura in the early part of 1st century A.D. The fact that the form 11a Ma^a appears as Jblun iMa> a in the Sinhalese Chronicles

(Rajavaliya : 33) in the form of the above term clearly indicates that it is the name Xlam/Ila that is meant. That Xla and Klu were treated as equivalents is also borne out by the above evi­ dence. We find a further reference to Ilam/Ila in the Anuradha- pura Tamil House nolders Terrace inscription (dated to 3/2nd century B.C.) which was wrongly read as Elu by faranavitane

(1970b.No. 94)

Paranavitane (Hay 1959 * 90) thinks that Sihila is the intermediate form of bimhala from which Tamil Ilam was derived. « On the otherhand Gm n^prakasar (19^' * 137) observed that iSihala is the later form of the original Ilam combined with a qualifi- cative Si (Sri) means illustrous. Kasanayagam (1926 t 102) derived Sihala from Ilam through Silam» Sihalaro and Simhalam.

It is now opportune to see how the name Sihala (Simhala) came to be applied to ^he Island about the 3rd century A.D. We have seen already that Ilcjn is earlier than Sihala, A deeper analysis shows thot the form Ila which occurs in the earliest

Brchmi inscriptions of Sri Lanka has the variant forms of Hel/ 34.

Hela/and £lu In the Sinhalese inscriptions (E.Z, 1912 t 170»

238; E.2. 19^3 » 54) of the period after the 8th cent\iry A.D,

and in the Sinhc^lese literature which begins to appear after

the 9th century A.D. While Paranavitane and others have derived

Hela, the ancient name for the Sinhalese people from Sanskrit

Simhala (Pali Sihala •> Hela - Hel) no explanation^ has been

supplied for the elision of the initial syllable (Si) which

should be treated as the most important element in the name.

This iactor raises serious doubts as to the validity of the

very base of the derivation of Ht.la from Sihala, It appears to

be proper to explain the element (S i) as a later addition to

the original name Ila/Hela, bein^, derived from Sanskrit (Sri)

just as in the case of the name Lanka we find the addition of

Sanskrit Sri in later times. Considering these factors it

seemE probable that the form Sihala (Simhula) has been derived

through a process of back formation from hela and subsequently

a legend seems to have been evolved to explain the name Sihala,

Finally our contention that Simhala was derived from Ilam tollies,

with ijurrows (1947) derivation of Simhala from Ilam through the

following process* ilam —^Cilam —^ Silam Sihala —^ Simhala,

We may now turn to the study of sources relating to the

evolution of the Sinliala language, xhe epigraphical records

appear in plenty with the introduction of Buddhism. Pali, the

canonical language of buddnism, also came to be introduced during

this period. The dominance of Pali as the linixixu fri-nca in the 34b

country prevented the development of the local languages. This

really accounts for the slow development of the Sinhalese

language whose literature begins to appear only in the 9th century

A.D, Thus the hold of Pali on native languages woiCLd have

given rise to an ideal situation of langu^^ges in contact and

bilingualism and in course of time, the languages of the indi­

genous people would have passed into oblivion. A study of the

vocabulary of Sinhala for example will shed some light on the

nature of the impact of the native languages on contemporary

Sinhalese. The earliest extant gr^unmar of the Sinhalese

language, namely 6idat Sangarava ^Dissanayake 1978) dating back

to the 13th century A.D. records three categories of words

into which the vocabulary of the oinhala could be grouped:

(a) Tasama (Skt. Tatsama)

(b) Tab|iva (Skt* Tatbhava)

(c) Mipan (Skt. Nispanna)

The above two categories of words namely Tasama and Tabava * designate words that have been either borrowed or derived from

some other language. The example cited by the author of the

Jxdat dan^arava to illustrate these two classes of words are

all of Indo-Aryan origin. The thiiM category of words referred

to as Nipan embraces words that are of indigenous origin. Here A one could see the deliberate attempt on the part of the author » -

Sidat San/^arava to treat the words of Aryan origin as foreign

to the oinhala language. Presumably there v/as a grammatical

tradition of treating Indo-Aryan words as foreign and Sidat

% ■ian^yarava only reflects that tradition. If Sinhalese is an 343

Aryan language, one cannot explain a3 to why the author of

the aidat dangarava should differentiate between indigenous

and Aryan words. This shows the original non-Aryan character

of the Sinhalese language. This is amply borne out by the study

of the ancient Sinhala script.

The Sinhalese alphabet is divided into two groups namely

iJLu alphabet (PI. XIX A) and the mixed alphabet (misra ) (PI.

XIX B). xhe iilu is the earliest form of the alphabet, which

is given in the Sidat oan/rarava. (Gunasekera 1962 i 3-6). The

mixed alphabet comprises letters of the £lu and the Sanskrit

alphabets and is more perfect than either of thaTs. It contains

54 letters of v^ich 18 are vov/els and 36 consonants and they

are used in writing both Slu and foreign words naturalised in

the language. At this juncture it may be relevant to note

the absence of mixed alphabet in the early Sinhalese records,

although the engravers were well conve sant with Pali and

Sanskrit, This shov/s the aversion of the users towards the

additional letters in the mixed alphabet, which were ^>vidently

foreign to them. This reflects the tradition as preserved in

the Sidat San/carava.

i'lany scholars who have worked on the origin and develop­

ment of the Sinhalese language have overlooked an important

element in the Sinhalese language, that is the structure. They

have been preoccupied with the analysis of the vocabiilary of

the language and its phonology and morphology but have paid little 350

attention to its grammatical stmcture and to any comparative

study of the language with the neighbouring group of DravidianQj languages. This ./as largely the result of starting ,'/ith the premise that the riahavamsa legend about the North Indian colonisation of the Island holds the key to the origin of the

Sinhalese language. As long as one sticks to the above pre­ mise, one is not very likely to go very far out of the Indo-

Aryan borders and this is what actually has happened in the

case of many scholars \iho wrote on the origin of the Sinh=ilese.

It is therefore necessary to br ak av/ay from this premise and

start with an analysis of the structure of the language as a

means to its identification. Of the scholars who were able to

leave the premise prescribed by the Mahavamsa and attempted

an analysis of the Sinhalese language from another angle, Guna-

wardhana and to some extant Kumaratunga, stand out prominently.

Gunawardhana (191B) vdio stressed in the manner of modem

linguists the importance of the stmctura of the language for

the identification of its place in any language family. His

knowledge of Tamil infact equipped him, unlike Geiger, to make

a genuine comparative study of the two languages. To quote

his words, “language, it may be premised, is the medium for

the communication of our thoughts, and thoughts are comnuni-

cated not by isolated words, but by means of sentences. Lang­

uage taken essentially is, therefore the sentence and gra'mnar ^>^*1

is that science which analysjfes and explains the construction of

the sentence. Scientifically, therefore, the deteraining

factor of a language is not its vocabulary, but its structure, 3 5 j-

viz. that aspect of it v/hich Is concerned with the arrangement and mutual adjustment of words in the expression of thought; and in this respect, it must be said that Sinhalese is essen­ tially a Dravidian language. This is not all. Its evolution too seems to have been on a Taiiil basis. And so we seem safe in saying that while in regard to its word equipment, Sinhalese is -aie child of Pali and Sanskrit, it is, with regard to its physical features and physical structure, essentially the daughter of the Tamil l;anguage» (Gunawardhonsi 1918 i 13). Thus

Gunav/ardhanct has clearly shown th it structurally there is a wide divergence between Sinhalese and the middle Indo-Aiyan languages.

It is interesting to note that the leading exponents of the theory that the Sinhiilese is an Indo-Aryan language, like

Geiger and Paranavitane, have conveniently fihosen to avoid meeting the arguments put forward by GunawirdJianfii.

The researches of younger linguists in recent yaars have revealed important characteristics of the Sinhalese language which are closer to those of Tamil and other , than to middle Indo-Aryan languages, Karunatillake (1974 :

151-153) for instance has dram attention to some of the points of similarity. "As far as the phonemic systems are concerned” he says ”a significant point of simil.irity between

Sinh .lese and Tamil is the absence of the aspirate stops which puts Sinhalese into sharp contrast with the rest of the modem Indo-Aryan languages". At the level of synt;ax, Karunatillake points out an instance of similarity, namely the formation of the 35^

relative clause in both the languages by using a verbal adjective instead of 'th t - who/which type of (co) relative pronominal sets'. He'J^f^phasises the fact th t "this feature which is typical of almost all the Dravidian languages is strikingly absent in ail the modem Indo-Aryan languages other than the

Sinhalese’ , Karunatillake adds to this an other instance such as the formation of higher ordinals involving 10 'and -Uie alter­ native construction type formed by repeating the fUyi3tor 'Hari'

( Tamil cari) after every item that is put into alteration".

(Karunatillake 1974 : 151-153).

In analysing the vocabulary of the Sinhalese Ian iiage, one no doubt finds that a majority of words are of Indo-Aryan origin. This should not lead one to conclude liiiat Sinhalese belongs to the Indo-Aryan family. As Geiger (1938 : IV) has emphasised "the loan words do not touch the character of a language". The occurrence of such a large amoimt of Indo-Aryan words can easily be explained. In the 3rd century 3*C,, when

Buddhism came bo be widely accepted in Sri Lanka, there v/as a necessity to use a common written language, to serve the purposes of the new religion, Pali therefore became ih e obvious choice and inscriptions came to be written in that language all over the Island, Pali, the canonical language of Buddhism, also came to stay permanently along v/ith Buddhism, These factors were to determine the character of the ne\7 language that was to emerge in the Island over the next few centuries out of the one or more non-Aryan dialects that were prevalent in the Island in the proto-historic period. 35u

Reverting to our linguistic analysis of the Sinhala language, Gunavardlume (1973 * 12-13), who studied the voca­ bulary of the Sinhalese language in some detail, enumerates a list of about two hundred words, which he considers as belonging to the Jipan (local) category. According to him most of these words relate to day to day life particularly of the simple

•village foli. Among the stock of words -ttiat he lists are those that refer to various animals and objects, plants and trees which formed an integral part of the natural and domestic environment of the earliest inhabitants. The occurrence of such important non-Aryan basic words definitely indicates the influence of a pre-Aryan language s tratum in the Sinhala language.

As mentioned earlier several of our linguists have admitted -tiiat there were certain non-Aryan languages in the Island in the pre-

Buddhist period andjthat they had excerted tlieir influence in the development of 3inhala. But unfortunately hitherto no research has been done in a scientific manner either to identify the earliest languages or*to asses the extent to which these lanijuages contributed to the evolution of the 3ihhala language.

An analysis of the basic words as well as the structure of the iiinh:-lesa language indicates the presence of at least two important layers before it was finally influenced by Pali. These are the AuslTro-Asiatic or Munda and the Dravidian elements.

However, it should be remembered th. t we have not attempted o a throuijih linguistic analysis of the material available ana therefore it is not our intention to make any definite claims about the nature of the affinities of the earliest languages of the Island and the iilu which emerged out of it . However^ on the basis of the foregoing discussion a suggestion is ventur d,

iifith regard to the Austro-Asi'^.tic element, it may be Siid tliat this was the langua e of the prehistoric late stone age people and today it survives in Vedda and Rodiya dialects (Ch. Ill),

Paranavitane (Kay 1959 s 96) a great c lampion of North Indian colonisation of Sri Lanka, admits the existence of a langua-e in the Island prior to this colonisation* According to him the vast majority of the people who today speak Sinhalese or Tamil must ultimately be de cended from those autochthonous people of

hom we know next to nothing. iVhatever the case may be, r,he mix­ ture of the Austro-Asiatic speaking people with the later arrivals of the Dravidian speakers shows that Austro-Asiatic element has played an important role in the composition of proto-i2Lu v/hich later emerged as a single language in the beginings of the early A, historic period. This view can be substantiated by correlating some of the proper names of Sri Lankan Brahmi inscriptions with the tribal people of South India. Thus it could be definitely asserted that thure were more than one non-Aryan dialects in use during the proto-historic times. The Tolkappivain, the ear'll eat extant speaks of 12 divisions of Nadu each having a dialect of its own (Meenakshi^Sundaram 1965 :194).

For example in the old Pandyan country itself there were tv'/o dialects in vogue. (Fig. 52). They are the dialect of Tenpan-

tina"^ (southeinmost^ortion of Tinnevelly district) and Patina^. ''IV

(Pandyan coxintry proper). So is tiie axae of the dialect of

Pulinatu, the old name of Velnatu (now comprising the Kanyakumari • • • district). This viras the region vriiere the old chieftains such as and Vels, rei^^ed supreme, whose influence in colonising

Sri Lanka is discussed in the epigraphical section. ^

Kanapathippillai (1948 t 216-228) observed that there is a strong tradition in the Tamil'literature that Sri Lanka was once a part of Old Tamil Nadu, What v/e v/ant to point out here is the old Tamil tradition mentioned in the Tolkaopivam about a nuclear zone of perfect Ta.iil (Cen Taiail) and peripheral derivations

(Kotum Tamil) around it. It consists of an encircling of peri- pheral dialects and another encircling of peripheral languages.

The Pandydin country vrtiero the traditional three Sangams took place was considered as the nuclear zone. The patterns of derivations in the peripheral languages like Kannada, Telugu, Tulu and Mala- « yalam show similarities with the Siniiala derivations of Tamil words which oLLtimately identifies Sinhala as a peripheral lang­ uage of Tamil. It is interesting to note here that the worlfs of poets from Sri Lanka are found in the Sangam anthologies which show even in the peripheral zone the studies of perfect Tamil was prevailing (Subrahmanian 1966a : 127). However^ in the case of Sri Lanka, the peripheral dialects of Tamil Nadu seem to have excerted a greater domin.inc|r ttian the dialect of the nuclear zone.

Going back to our analysis of the languages of the proto- 35u

historic times the Pandyan elements of the legends could be equated with the TarabapcUini region, (Fig, 50) although their influence was extended to Nagadipa, Anuradh ;pura and the south eastern co ists as well. The megalithic u m burials probably represent the earliest element of this wave moving out of the

Pandyan country. It is also posoible that the peripheral dialects arrived in Sri Lanlca slightly later than the ori.-inal nuclear zone dialect. The evidence of the rule of early chiefs such as Vels, Ays, in Sri Lanka clearly substanti tes this view.

Their admixture wi-tti the Austro-Asi; tic elements is also reflected in the xakka legends of the Mahavamsa.

At this time there \;as another definitely identifiable wave of Tamil spealcers corain^ from T^unil Nadu speaking more deve­ loped fora of Ta;nil and this took pl:ce around the middle of the first millennium B.C. or slightly later. This group v/as sufficiently different from the earlier Dravidian speakers and this seems to have been the reason vjhy the members of this group were desi^ated by the epithet DamedayPemela in the 3rd centuiy

3.C . Brohmi inscriptionsof Sri Lanlca (Appendix I ) . The ne\-/ wave of Tamil influence and the strong influence of Pali language beginning from the 3rd century B.C. were considerably felt on the earlier dialects, thus paving tha /ay for the emergence of i:.la, proto-Sinh ii.ese,

Indo-Aryan words in r-lu largely borrowed through Pall rather than directly from Sanskrit, indicate that it v/as the 357

predominance of Buddhism and Pali, that was responsible for the

occurrence of large number of •.■forda of Indo-Aryan origin in the 1 £j.u language. In iiie case of the Sinhalese language the impact

of Pali \-raa understandably even greater, for the imposition'of

Pali took place at a very early stage in the evolution of the

Sinhalese language. The influence of Pali continued for over

2000 years. Consequently the vocabulary of the Sinhalese langu­

age could only be expected to be largely drawn from Pali, This

really changed the original character of the Sinhalese language which undoubtedly belongs to the family of Dravidian languages

and more closely allied to the peripheral languages of the old

Tamil. Pali being the language of the State religion gradually

changed the character of the local language, namely Elu. This

really accounts for the peculiar position of Sinhalese well

as its similarities wiiii the North Indian dialects.

The above feature i .e . bilingualism is not without any

parallels and there are cases when the original character of

the language had been chr^nged or displaced through contact or

bilingualism when a language is supported by political power and

religion. This actually happened in Sri Lanka as i^ell. A few parallels from outside world are being cited to demonstrate

how a change can take place. Today less iiian half of the popu­

lation of the world speaks the family of Indo-i:Airopean languages

(Rulen 1975 J 58). This dominent position is of recent date,

being a direct consequence of the colonial polices launched by

the xAiropean nations around 1500 A.D. From this time onwards 35d various i^Juropean powers i .e . Spain, Portugal, France, Great Britain

and Russia began the policy of dividing the rest of the world

amongst themselves. In South America although it v/as populated

by North i^erical Indians when the ii.'uropeans arrived there were

only 10-20 million speakers there, speaking large number of

dialects. Today Indian language speakers are only 11.2 million.

(Rulen 1975 * 135).

Two thousand years ago there v;ere number of Italic languages

spoken in what is now called Italy, ■lilch the establishment and

expansion of the Roman iiinpire to the Atlantic ocaan in the west

and the black sea in the ±uast, the principal Italic language,

Latin submerged the other Italic tongues, which died out soon

after the beginning of the Christian lira (Rulen 1975 » 69). Prior

to the establishment and expansion of the Roman empire, Celtic

languages were widely spoken throur;hout iJurope from the Iberian

Peninsula in the west across burope to the Balkans and into Asia

ainor in the £ast. As Latin spread over much of this area,

however, the various Celtic groups were first submerged and

later assimilated to the Latin speaking populations. Only in

a more inaccessible regions of the British Isles did Celts

manage to preserve their language down to the present day. On

the jcXiropean continent itself Celtic languages became extinct

before 500 A.D. (Rulen 1975 * 70).

Afro-Asiatic group of languages are spoken in North Africa

by 170 million people. Of these 100 million speak Arabic. These languages are subdivided, into 6 branches, namely Semitic, Berber,

Chadic, Cushitic, Omotic and hgyptiein. The last became extinct by

the I7tli century A.D« In terms of both number of speakers (125

million) and geographical spread (from North west Africa to Asia-

minor), the Semitic br;inch outweighs the other five branches in

importance. Its best known languagiss are Arabic, Heb- ew and

Amharic. The latter is the national language of iithiopia. The

Berber languages, with seven ;aillion speakers, v;ere formerly

,/idespread throughout North Africa. However, the Arabic expan­

sion eastward from the Arabian Peninsula in the 7th century A.D.

largoly submerged various Berber languages, which are today

sopken chiefly by peoole bilingual in Arabic (Rulen 1975 * 6 l ).

Another parallel that could illustrate this point is the case

of Iranian language. Al-iiough this langua^^e belongs to the Indo-

Ira:iian grou.:) and was considerably developed at the time of the

expansion of Islam 6/7th century A .D ., the impact of Arabic

through Islam brought into that Iranian language a very large

•amount of Arabic words which resulted, in modem Iranians voca-

bul. ry bein j 70% Arabic.

The above few examples cleaily show how a language accompained

by political jower and in certain cases religion could either

submerge another or change its character while coming into contact.

In Sri Lankan context Pali, the language of Buddhism allied with

the political power, while coming into contact with local language

iilu by the process of bilingualism, changed its character. This

process has been clearly observed by linguists as well. According

to Sridhar (1974 : 409) the prevalance of bilingualism may also affect the language concerned, i .e . it may lead to linsuistic diffu­ sion wherein a mo el in language A has been reproduced in language

B as part of acculturation process resulting in the Ian uage in­ fluence taking place wii^iout bilingualism i.e . merely through the introduction of cultural items - the resultant language displacement being only partial.

* Lastly it should be mentioned that the study of place names is another valuable source of information for this period. But as

Perera (1953 * 2425-250) observed it is a pity liiat the place names of the Island, though a valuable source for liie student has still remained unexarained. But due to chronological problem the study has not been pursued here in the analysis of the proto- historic language. The same is true of Palaeolinguisties,

./hich is also another branch of research vrhich could yield valuable data for this period, for in the Jaffna Tamil dialect one still finds the survival of the words used in Sangam literature, in their daily life,

£^IGKAPHICAL LVIDi^Ci^

The analysis of the Brahmi inscriptions of Sri Lanka,

«/hich are the earliest episraphical records is relevant here for chey represent the ethnographical and linguistical residues of the megalithic settlers, besides pointing to the geographical distri­ bution of the settlements. According to Paranavitane ( 1970b : XXIII)

Brahmi script appears in Sri Lanka along with the introduction of Buddlaism aroiind 3rd century B.C. (Pl. XX) thou^^ he n-ivar d.:nies 36i

the possibility of an earlier introduction which v/as postulated by scholars like Buhler (1904), Fernando (1969 * 19-24) and

Karunaratne (1960a, 1960b, 1962a & 1962b), yfe share the view with Fernando and Karunaratne in idaitifing pre-Asokan and post-

Asokan stratums in Sri Lanka Brahrai script on the evidence of the survival of the archeiic forms foimd both in Sri Lanka and in

Tamil cave Brahmi inscriptions of Tamil Nadu.

It is true that, we have not got all the forms of writing of the pre-Asokan script in the Brahmi records of Sri Lanka. But the very survival of these forms does shov/ their .iwirlier occu­ rrence and suggests that 'tiiey were submerged by the wave of

Asokan Brahmi, This is also confirmed by the historical ana­ lysis of the Sinhalese language, 03ie existence of cave inscrip­ tions in Tamil Nadu (the first discovery goes as far back as

1906) has enlarged the horizon of the study of the southern variety of Brahrai. Nagaswam^ (1968 i 410-415) equates this script with Damili mentioned in the Jaina works. Here it may be pointed out that the dates of the early Brahmi records of Sri

Lanka too are the same as that of Tamil Nadu,

Though Buhler putforth the famous hypothesis that the southern variety of Brahmi developed from Bhattiprolu, Maloney

(1968, Ch, VIII) on the evidence of Palaeography and geographical distribution of the earliest cave inscriptions of the old Pandyan country, has convincingly proved that the Tamil Brahmi is inde­ pendent of the Bhattiprolu tradition of writing. However^ his ■ yt. ’ . ( . . 1 r, 362

contention that this script has been introduced from Sri Lanka to Pandyan country cannot be accepted for reasons discusoed above.

The Tamil Brahmi inscriptions stand in close geographical proximity to the Sri Lankan Brahmi inscriptions. Like the earliest

Brahmi inscriptions of Tamil Nadu, the earliest Brahmi inscrip­ tions of Sri Lanka are also clustered around the Anui^dhapura region (north-western Sri Lanka). The architectural features 1 of the caves of both the countries are similar. Bo1di were used by the non-Hindu sects for rainy season. In both the places the inscriptions are foxind on hard granite rocks protected from the rain by drip-ledges and in positions beyond the reach of those who would have caused damage to than. Such occurrence in other parts of the subcontinent is rare. In both 'tiie places tiie inscriptions are generally short, running into one or two lines only. In some of the caves are found small bed steads for the use of those who dwelt in them. In general appearance these re­ cords are so alike that these can be easily mistakai for those of Tamil Nadu (Fernando 1969),

The analysis of the internal evidence, namely the script, alBo su jports the above identical feature. In the earliest examples of Brahmi writings in Sri Lanka it has been noted that there are some features not noticed in tBie Brahmi script employed in the inscriptions of Asoka, the earliest extant examples of Brahmi writing in India, More remarkable affini­ ties are noticed in i, n^, 3^, ra, ga, te, u, & e (Fernando 1969 19-24), 36o

Palaeograohlcal Similarities (Pl. XX)

(a) ( H ) type a with the curved left limbs arc the common­

est in Asokan Brahmi in which the two curves have in some types

been reduced to two slan:ing strokes Joined to the vertical in

the same place at its middle. But in the earlier Brahmi inscrip­ e tions of 3ri Lanka one could see a type of a in which there is

I some space between the junctions of the two left linbs. This

is a peculiar feature of both Tamil and Sri Lankan Brahmi, (Pl.XX).

(b) (•!•) 4 Initial vowel employed in the inscriptions of Asoka

consists of three dots placed in such a manner as to form the

comers of a triangle, whereas in the earliest extant Brahmi

records of Sri Lanka this vowel is represented by a vertical stroke

with two dots on either side. It is interesting to note that the

same syiabol is found only in the cave inscriptions of Jamil Nadu

other than Sri Lanka. The modem Sinhalese ( 6^) as well as the

Tamil letter ( ) for i is obviously derived from this symbol

' (PI. XX & Fig. 53).

(c) There are several forms of the letter ( ^ ) ma used in -ttie

inscriptions of Asoka, but the form that is employed in the

early records of Sri Lanka does not resemble any of them but

belongs to a type 1diat is generally called the ( W ) t3^ e ,

formed of a tube like o u w e with a horizontal crossbar termi­

nating at the middle of the vertical arras of the tube. This

is the symbol for '||||[^ in the Tamil Brahmi records of Tamil Nadu

(Fig. 53). 3G4

Commenting on the occurrence of symbols _£ and ma in the

early Brahmi records of Sri Lanka Fernando (1949 * 283) observed

that 'i f as is generally believed, the Brahmi alphabet v/as

introduced by the Buddhist missionaries who came over from

India, it has to be explained how the Ceylon records which, as

far as is ascertainable, go as far back as the time of king

Uttiya, come to have forms of the i, and the which are al­

together absent in Indian records set up at the same time the

Ceylon records are said to have been inscribed. The earliest

occurrence of this i in North India is in the 2nd century A,D,

The has not been noticed in any records of North India of

the time. Perhaps these letters may have belonged to a diffe­

rent school of vnriting, that h=d separated itself from the

Northern school and had remained confined to the South” .

(d) ( t)' ) la This is the symbol for palatal to the right • • vertical of which has been attached a short horizontal stroke

with a short horizontal stroke hanging down from its end.

Paranavitane ( 1970b x XXII-XXIII) read this as lu. Whereas

Parker (19^)9 : 436) raad this symbol from the inscriptions at

iirupotana (No. 33) and Periya-Puliyanlculam (No. 353 and 355) as

la. iVhile coming to this conclusion he observed that ’’with • regard to the characters used, it is interestin-r to obsc^rve

in no less than four of these inscriptions (No. 13» 26, 27 and

33) a letter ^a, which in India I believe, is only found in the

southern ins^-riptions. I am not aware that it occurs in early

cave inscriptions in other han the northern parts of the Island...

I am afraid that it would be unsafe to assume that the names

given in Nos. 26, 27, 28 and 31, may be those of Dravidians; 365

there m i ^ t be such chiefs in northern Ceylon whose families

were Buddhists." It may be pointed out here that the study

of this symbol occurring in inscriptions shows that it was not

fonfined only to the northern Sri Lanka as Parker observed but

it occurrs in other pc«rts of the Island as well. The comparison

of this with the Indian Tamil Brahrai ^ sho./s that both are

identical and hence may be from one source. In Tamil Nadu K,V.

Subrahmaniya Aiyer (1925 * 275-300) adopted the vie^.'/ of Parker

and read this as According to Sivaramainurti (1952 : 152)

the modem Tamil la is derived from this Brahmi symbol and this

was confirmed by Mahadevan (1968) and Mahalingam (1967). In

Sri Lanka also Karunaratne (1962b) disagrees with Paranavitane

and deciphered this symbol as la ("^ ). m

Another significant symbol is ) which occurrs • • in the Periya Puliyankulam inscription at Vavuniya (Paranavitane

1970b. No. 356) It is interesting to note that this letter

exhibits a form (PI. XIX C) ttiat is even earlier than that of

the la so far seen in the inscriptions of Tamil Nadu (Fig, 53).

In this respect She Sri Lanka inscription is of great signi­

ficance. The retroflex cerebral 1^, a sound very distinctive 9 0 and peculiar to the Dravidian languages is not known or used

in the Brahrai script of Asoka. The occurrence of this symbol

la in reference to a Tamil merchant is significant in more • » ohan one respect. The Archaic form of this symbol not only

shows its antiquity in the Island but also prooves that there

v;ere Tamil speaking people who used it. 3GG

Besides the above mentioned forms, even the signs lik e ra . e ga, ^ ^ and u deviate from Asokan Brahmi but have close affinities with -tiie Brahmi of the Pandyan country (Fig, 53). Another feature that points to the greater antiquity of the Brahmi scrip t in Sri Lanka is the employment of the aspirated form of the letter ja in places where the de aspirate is normally required. (Fernando 1969). For exaaple the word i\a.1a (king) is spelt invariably with ttie aspirated fora of the letter which occurrs in the inscriptions of Asoka as well as in other inscriptions. This practice is not noticed anywhere else, and quite clearly shov/s that the Brahmi script used in Sri Lanka in the earliest extant records does not belong to the same tradi­ tion as the sci-ipt employed in the inscriptions of Asoka. The absence of the symbol for the long vowel a ( ^ ) is another feature of both Sri Lanka and Tamil BraJomi. ./herever a v/as nectissary it was represented by idie sign of mere a ( H )• This fact should oe correlated by the Tamil practice of using only a dot to differenciate the medial vowel from the long vowel. Hence a symbol of Asokan medial vowel should be taken as a Iona vowel in Tamil and Sri Lankan Brahmi. Thus -aie word I'ead by Paranrivitune (1970b) as AYA will become ^ which w ill be diseussed later,

Fernando (1969 : 21^ further sees a remarkable a ffiriity of the script of the cave records of Sri Lanka v/ith the gr .ffiti • incised on pottery discovered at Arikamedu in Ta^iil Nadu (l/2nd century A.D,), where the peculiar forms i and ma occur. Accord- .ing to him ”the inscriptions of Tamil Nadu, Brahmi records of 36

Sri Lanka, and the graffiti marks on .he pottery at Arikamedu (although belonging to a later period) represent perhaps the same tradition. Probably these records were carved by scribes of one and the s:ime school and i f so, it*has to be assumed that a school of scribes, differing in several respects from those who carvcd the.inscriptions of Asoka, was existing in Joulii India and Sri Lanka and was practicing its art in these regions even before the time of Asoka". Finally he says th it "th is school of Jtirahmi must have migrated to South India and Ceylon, at least a century or two earlier, for by the time of Asoka the North Indian school had forgotten the South Indian forms i and ma."

As said earlier, Buhler, Fernando and Karunaratne accepted the possibility of the Brahmi script being introduced much -arlier. It is reported that Saddhamangala Karunaratne (1962^) has claimed that before the introduction of Buddhism there had been a use of Brahmi alphabet in Sri Lanka. In fact it is further reported that Karunaratne (1960b) basing his arguments on some of the Brahmi in scriptions, has stated that Sri Lanka had been invaded by Tamils from India before the coming of Vi;}aya.

The above analysis clearly shows th^^^t there was an earlier form of Brahmi prevalent in the Island, before the introduction of Asokan Brahmi. Coming to the introduction of pre-Asokan Brahmi into Sri Lanka, the existence of early Dravidian form of writing clearly shows that the introduction of Brahmi took 36o

place along with the other cultiiral elements from the Pandean coast. This is strengthened not only by the reference to Tamils in the epigraphical and literary records but also by the linguistic evidence of the inscriptions.

The pre-Budahist social stratum becomes more clear when one analyses the inscriptions from a linguistic angle. Une of the most important words is ^arumaka (Far^-navitane 1970b t Ixxiv* Ixxxvi). This is us

the Parumakas themselves had ceased to be referred to in inscriptions, (Paran-vitane 1970b).

Various scholars have hazarded diiferent oijinions about the etymolotiy of th is word. B ell (1892b i 69) argued that the word signified a chief and was applied to kings; he also draws our attention to the similarity of this word with Sanskrit Pramukha and Jilu (Sinnalese) Panok and with the Tamil word Perumakan used for a prince and a nobleman. On the other hand Paranavitane (1970b } Ixxiv) nolds that triis word is the old t>inh-lese form of bansKrit Pramukha and P ali Pai.ukha or Pa:aokkha . which in olten used to denote the President of a guild or corpo­ ration. He further says that equally untenable is the effort to derive Parum^^ka from iam il Perumakan. which it s e lf has to be taken as derived from Sanskrit Pramukna. According to him the word Perumakan is attested in Tamil literature for a period which is at least two centuries after that of the latest of the records,

Paranavitane's farfetched etymological efforts to connect Parumoka with bans r i t and P ali become a miserable fa ilu re when we examine the word phonolofcically and morpholOt;,ically, Accord- infe to Wickramasinfehe (E,Z 1912 t 17, 35) i>anskrit Pramukha. Hindi Parmukha, P ali Pfamukho and PamokKho. Sinhalese P.oaok. a ll mean ^chiex * and phonetically also the derivation of the Pali and the Sinhalese forms is correct. But Sanskrit Pra will be­ come Pai‘ or but not Paru in the P rakritic languages, as in 370

the case o i banskrit urlva becoming uiva. and cnandra becoming Chanda. Hence Paranavitane*s derivation of this form is phone­ tically not possible. This proves that the Sinhalese Famok and the P ali Pamukho and Pamokkho are rea lly the derivations from the iianskrit word t-ramukha. It is quite surprising that Paranavitane should derive Parumaka from Pi*amukha when there is already a parallel available in Pali and Sinhalese. Further the woi d Paru.aaka has been written invariably without using the aspirated kha. though this kha is available in Sri Lankan Brahmi which proves the Peiumakan origins of this word. It is interest­ ing that we do find a feminine term Perumakal (Paranavitane 1970b) which Paranavitane reads as Parumakalu by mistaking the la (nh ) for lu ('V’ )• fCiiiinine term Perumakal revtals without any doubt that i t originated from the Tamil term as th is gender ending is not found in any of the Indo-Aryan languages. Kven i f i t i s irarumakalu. i t contradicts Paranavitane * s contention as Pramukha is a neutor word and i t cannot take a feminine ending, as Parumakalu. irerumaKans were the lo ca l cnieitain s who lived in various parts of Tamil Nadu as revealed in the Sangam litera­ ture (I.L.T•A.1970 I 1097-98). Even the word ^arumakan gives the same meaning as Paru and Peru are of same origin and of same meaning.

A closer analysis of the meaning of this term snows that Parumaka/Parumakan/Perumakan in both S ri Lanka and Tamil Nadu referred to the personalities who had similar functions. Tamil lexican (1932 * 1881-2882) gives three meanings of th is 87i

terra. They are (1) Great man (2) chief (3) kin^. Peruman or DGmman. a variant of this, also meant a nobleman, great man, elder, king and finally lord Siva. Again Perijmakkal too meant greatmen. Perumal. a title used by the Cera kings had an ori» iin a l meaning elder, ^reatmen, noblemen, kin^,, and fin a lly the God. The secondary meaning of this term as a reference to god t shows the divine origin of kingship as in the case of the terms and Utaivar in Xajail.

Two other words which occiir in the inscriptions are Vela and Vel. Paranavitane (1970b t 122) in Jriis characteristic * manner traces the origin of Vela to

Paranavitane (1970b : 122) derives the terra Yel from Sanskrit (va ilva ). As mentioned e a rlie r th is again is a result of reading the symbol lb as lu. ihe term Vel figures in the Tamil Brahni records of Tamil Nadu (l^ahalingara 1967) and hence it could be taken as Vel rather than Velu. There is thus both lite ra ry (Subrahmanian 19^6l9i } 801) and epigraphical references for the CVels) who were local chiextains ruling in 372 different parts of Tamil Nadu and their country was called as

Vel Nadu, which is located between Madurai and Trivendram. The # Vels are one of the earliest a^^ricultural conununities of Tamil Nadu whom Champaklakshmi (1978 > 110-122) identifies with the Black-and-ited ware usint megalithic people. They were known as

Velir in plural and mentioned as Tonmudii- Velir and mutiukudi

Velir in the Sangam literature which means grand old agricultural % community. Even the present Ta;ail term Velalar for the agri- cultural community, which is the dominant caste in Tamil Nadu, is derived from the term Vel (Subrahmanian 1966b), The other proto-historic tribes such as i^^alavar and ays also referred to in the Lankan Bra]:Mi inscriptions are actually the branches of th Vtlir community.

The word ^ and its feminine form iibi are frequent and widely di:.tributed names in the bri Lankan Brahmi inscriptions which disappear alter the first century A,d ., alon^ with

Parumaka. Paranavitane (1970b) reads them as Ava and as usual derives tiiis word from Jianskrit iirva. But Mahalingam (1967 i

160) on the basis of the parallels from Tamil Nadu says that this word resembles very much the term ^ in Tamil indicating the line of petty ruleis who h-uld sway over the /el Nadu in the early days. The word Avimara appears in four different inscriptions. Though Paranavitane (1970b* 103) derives this term from a vedic deity (S nskrit Arvaman) . it will be more convincing if we equate it with Avmaran who could have been a Kandyan chieftain. 373

The I’eminine form of ^ is mentioned as a h i in inscriptions wO which f'aranavitane (1970 * 102) traces the origin from the Sanskrit Amba.Actually the term Abi is a derivation of iiwai (a.R.A.S.M. 1929 » 58) which could be a feminine form of Aiw an. In early Kannada inscriptions which is a peripheral Dravidiijn language, tliis tenaa is men­ tioned as Aboe (A.K.A.b.M, 1929 * 79)» obviously due to the interchi-oige of jjra and Similarly it has become Abi in proto-binhi-lese, that is £lu,

Anoti.er relevant term we come across in the 5ri Lankan inscriptions is the term falava (Paranavitane 1970b Nos. 58 and 159) which denoted Palavar. another proto-historic tiibals of Tamil f'tadu (6ubrahm;inian 1966a t 539) who are also connected to the Velir. iralayan maran is a name 4f a Pandya King, and riatiavamsa (Ch, XXXIII t 56-61) also mentions neunes as Palava mara and Pan:-va mara which testifies the presence of Palavar in Sri Lanka. The verm Pulava (Paranavitane 1970b s 115) occurs in inscriptions twice and this could be taken back to a Tamil tribe known as Pulavar. ruentioned in the Sangam litera­ ture (Subrahms.niari 1966a i 582), despite the vague effort of

Paianavitane to truce it to Sanskrit sa^e rulava whose existence is doubtful.

Another common name af p«aring in the inscro-ptions is barata which Paranavitane (1970b t 116) interprets as a title and gives a meaning as Lord. But one could see the occurrence * I 374

of the parallel term referring to a coimr.unlty called Paratar

^aratavar figuring in the Sangam I'iteiature (J:>ubrahmanian

1 9 6 6 a t 532) who lived in the coastal areas and took initia­ tion in the seafaring activities. Probably the present day

Paratar community of coastal Sri Lanka and Tamil Nadu are the

descendants of this tribe^ The above mentioned social set up

clearly indicates that the early Sri J..ankan society had close

affiliation with the proto-iiistoric society of the iandyan

country and was free from any Aryan iniluences before the * introduction of BuddnismX

The presence of Prakrit inscriptions does not mean

that the entire local population was well versed in it.

Even in many of the ^>outh hiast Asian colonies» by the dawn of

Christian era we iind pure poetic Sanskrit inscriptions which

obscure the nature of the local languages. But after few

centuries the local vacabulary comes to the surface to form a

cr.ol language. Cieiger and Paranavitane and others have

assumed that the Prakrit we find in the inscriptions was the

language of the people and proceeded to analyse the evolution

of the Sinhalese language on the basis of this conclusion.

This basic misconception has naturally led to serious defects in their analysis.

% However modern linguisticians do not deny the existence

of the non-Aryan languages in the Pre-buddiiist period.v^ Even 375

Paranavitane unwittingly admits this (Kay 1959 » 96). Already in the Prakrit inscriptions, one could see the occurrence of words from the local non->iryan langut^ges, a feature which becomes increasingly significant as time ^oes on. By the time we get to the 5th and 6th centuries A.D. there is a considera­ ble dis^.lacement of early Prakritic words by words from local languages. It is at this point that the two elements begin to fuse. Unfortunately scholars like Geiger see this as a 0 break with the earlier form of the language and found it diffi> cult to explain the transition satisfactorily. According to him ”to explain tliis break we may assume that in the course of the fir s t five or s ix centuries of our era the ,;.poken language of Ceylon gr^idually differed so much from the style, phraseology, and grammar of the inscriptions owing to the conservative character which is peculiar to such documents, that they became more or less u n in tellig ib le to the general public. A radical change was unavoidable, even in the outer form, by the introduction of a new type of script derived from the Grant ha of i:>outh India which probably had come into use in the daily life" (Geiger : 1938 : iv). Rather than treating this as a break, one has to take it as the point at which the alien Prukritic language and the native languages began to meige.

t i’he above discussion highlights the controversial ujspects of Sri Lankan Brahmi inscriptions which have to be reanalysed and correlated with fresh archaeological evidence, i^vhat 37 u

happened in iirilankan lin^juistics during the proto-historic and early his(,oric periods was syncretism, assimilation and fusion between the then existing Dz^vidian cum Austro-Asiatic language elements and the waves oi canonical Pali« This view is substantiated by the presence of a pre-Asokan stratum of ai- chaic Brahmi and by the presence of certain megalithic corniou- nities who are common to la 11 Nadu and Sri Lanka (Fit, 54 & iippendix I). Ihe very presence of the archaic form of Brahmi script only in the Pandyan coimtry and in ^ri l«anka is of great significance. Karunaratne (1960a, 1960b, 1962a & 1962b) and

Nagaswamy (1968) see a common source independent of Asokan

Brahmi and tCrishnan (1974 i 26-30) confirms tids by saying that this a different source which remains to be explored.

We may now conclude with Mahalingam*s (1967 l 160) obser- vauionsi "it is very likely that traces of the also are found in the cave Brahmi inscriptions in Ceylon. A study of the Ceylon inscriptions from this point of view, which does not seem to have been seriously attempted so far, may possibly bear ample fruit".

Our survey, however, clearly shows that the interpretation of the geneiis of a particular culture in the ab ence of ablolute dates needs a different approach. The study of the various phases of this culture based on arcriaeoio«;,ical comparison with the mainland of India and letter a correl^^tion of common socio­ cultural eluents of an earlier substratum which survived the later influences, clearly siiows that a time lag of a couple of 37 V Ij - ^ I

centuries before the Introduction of Buddhism would have been

required for their migration and survival, ^ence as Homila

Thapar (1975) indicated, a date of 7th century B.C. could

stand as a reasonable date for its introduction. Yet it is

plausible that the mej/alithic culture may have been intro­

duced even one or more centuries earlier than 7th century B.C.

This is supported by the legends of the piroto-historic period

as wej.1 as by the evidence of other elements of the early

civilization alon^ the Pandy&n coast.

'■ . I