THE MUNDA MARITIME DISPERSAL:

WHEN, WHERE AND WHAT IS THE

EVIDENCE?

2019 version

Roger Blench
McDonald Institute for Archaeological Research
University of Cambridge
Department of History, University of Jos
Kay Williamson Educational Foundation
8, Guest Road
Cambridge CB1 2AL
United Kingdom
Voice/ Ans (00-44)-(0)7847-495590
Mobile worldwide (00-44)-(0)7967-696804
E-mail rogerblench@yahoo.co.uk
http://www.rogerblench.info/RBOP.htm

This version: Roing, 19 July, 2019

1
 TABLE OF CONTENTS

TABLE OF CONTENTS................................................................................................................................. i
1. Introduction ................................................................................................................................................. 1
2. Munda farming culture .............................................................................................................................. 3
2.1 Linguistic evidence ................................................................................................................................ 3
2.2 Archaeology: the Eastern Wetland Tradition .................................................................................... 3
3. Dating the Munda dispersal ....................................................................................................................... 4
4. Locating the Munda dispersal.................................................................................................................... 4
5. Material culture correlations ..................................................................................................................... 5
5.1 Material culture as an indicator........................................................................................................... 5
5.2 Two musical instruments...................................................................................................................... 5
5.2.1 The split-tube rattle .......................................................................................................................... 5
5.2.2 Monochord bowed mouth-resonated zither...................................................................................... 5
6. Some linguistic counter-examples.............................................................................................................. 6
7. Conclusions .................................................................................................................................................. 7
References ........................................................................................................................................................ 8

TABLES

Table 1. Paddy rice in Proto-Austroasiatic........................................................................................................ 3

Table 2. Husked rice in Proto-Austroasiatic ..................................................................................................... 3
Table 3. Chicken in Austroasiatic ..................................................................................................................... 3
Table 4. ‘Sun’ in Munda and Khasic................................................................................................................. 6
Table 5. ‘Dog’ in Munda and Khasic................................................................................................................ 7

MAPS

Map 1. Austroasiatic languages today............................................................................................................... 1

Map 2. Distribution of Munda languages.......................................................................................................... 2
Map 3. Primary dispersal of Munda speakers ................................................................................................... 7
Map 4. Dispersal of Munda speakers within India............................................................................................ 8

PHOTOS

Photo 1. Split-tube bamboo rattle, Sora ............................................................................................................ 4

Photo 2. Split-tube bamboo rattle, Sulawesi ..................................................................................................... 5
Photo 3. Split-tube rattle, Sumba ...................................................................................................................... 5
Photo 4. Sora monochord bowed mouth-resonated zither ................................................................................ 6
Photo 5. Bahnar monochord bowed mouth-resonated zither ............................................................................ 6

ABSTRACT

The Munda languages are the most far-flung and geographically fragmented branch of Austroasiatic, spoke
in a broad zone of Central and Northeast India. It is usually thought that they must have spread by land,
given the presence of Khasian in Northeast India, although it is hard to see exactly what would have driven
this dispersal. Paul Sidwell (pers. comm.) has made a proposal which goes some way to resolving this
problem, namely that the Munda, far from diffusing across the land, made a sea voyage across the Bay of
Bengal. This would certainly resolve the issue of the geography of Munda but begs many questions about
the when, where, why of such a migration. This working paper attempts to provide some preliminary
answers, exploring the linguistic evidence for agriculture, the archaeology context and a note on material
culture. It proposes that the Munda must have travelled in Austronesian shipping in the period 3500 ~ 4000
BP, leaving from the south of Myanmar of the Isthmus of Kra.

Keywords: Austroasiatic; Munda; dispersal

1
Roger Blench Did the Munda reach India by sea? Circulation draft
1. Introduction
The Austroasiatic languages are geographically fragmented, despite being a relatively coherent phylum
(Map 1). It has been argued in recent times that this dispersal is to be identified with the SE Asian Neolithic,
which would place its primary diversification at around 4000 BP (Sidwell & Blench 2011). The Munda
languages have always constituted something of a problem, since they are scattered across a broad zone of
Central and Northeast India. This geographical separation originally led researchers to believe the phylum
was divided into two groupings, Munda and the remainder, named Môn Khmer after their two most
prominent groups. This is not now generally accepted and Munda is one among the fourteen or so branches
now recognised.

However, the geography of the branches of Austroasiatic presents a problem. If the Munda languages spread
westward from a homeland in SE Asia, why do they show no specific resemblances to the nearest branches,
for example Khasian and Palaungic? What drove their dispersal and fragmentation, such that they reached
Central India? Paul Sidwell (pers. comm.) has made a proposal which goes some way to resolving this
problem, namely that the Munda, far from diffusing across the land, migrated by sea across the Bay of
Bengal, ending up somewhere near modern Bhubhaneswar in Odisha, and expanding outwards from there.
Given that the Nicobarese also reached their current homeland by sea, there is nothing inherently
implausible about this, although the journey of the Munda would be considerably further.

Map 1. Austroasiatic languages today

This would certainly resolve the issue of the geography of Munda, but in turn raises a whole raft of new
questions. These can be summarised as follows;

a) No Munda groups today are seagoing or show interest in maritime subsistence, and they are all farmers
or even partial foragers. Why did their subsistence change so radically?
b) Does the lexicon of Munda today provide clues to a past maritime lifestyle?

1
Roger Blench Did the Munda reach India by sea? Circulation draft
c) The Munda lexicon implies a rather reduced agricultural package allied to SE Asia, hence the maritime
dispersal would have been post-Neolithic, i.e. after 4000 BP. But when did this take place?
d) Where did the migration start? A ‘least moves’ version would propose modern-day Bangla Desh or
coastal Myanmar, but we have no immediate evidence for Austroasiatic presence there. Equally
plausible is the west coast of Peninsular Thailand, where there are Mon and Aslian and presumably
the starting point of Nicobarese
e) Such a migration would surely have consequences for material culture, both in the archaeological
record and in perishable items. Is there evidence for these?

The paper aims to provide some partial answers to some of these questions.

Figure 1. Revised tree of Munda languages

North Munda
Korku
Santali, Munda
Sora–Gorum
Juang
Kharia
Gutob–Remo
Gta’

From the point of view of the present argument, the exact internal structure of Munda is not significant.

Map 2 shows the present-day distribution of Munda languages. The lack of a coastal population is rather
striking as is the division into Northern and Southern populations, which undoubtedly reflects the expansion
of Orissan [i.e. Indo-Aryan speaking] in the early historic period. The scattered populations in the far
northeast, into Nepal and Bangla Desh, are known to be extremely recent, in part connected with the
establishment of tea plantations.

Map 2. Distribution of Munda languages

Source: Ho Project, Swarthmore College

2
Roger Blench Did the Munda reach India by sea? Circulation draft
2. Munda farming culture

2.1 Linguistic evidence

Most of the Munda peoples, especially the larger widespread ones, such as the Sora and Santal, are
subsistence farmers. However, among them are near-foragers, such as the Birhor and the Juang. As Zide &
Zide (1972) point out, the original hypothesis was that this was closer to the original pattern of the Munda
and the other groups adopted agriculture from their Indo-Aryan neighbours. In fact, however, the reverse is
true, not only because most of India remained in a foraging lifestyle until relatively late. The Munda must
have carried rice agriculture from the SE Asian mainland, since several rice-connected roots have good
cognates, as shown in Table 1 and Table 2;

Table 1. Paddy rice in Proto-Austroasiatic

Branch Language Attestation Gloss
Munda Sora sәrɔ paddy
Bahnaric proto Central-Bahnaric *snrɔʔ early rice
Katuic proto Katuic *sarɔɔ paddy rice
Khmeric Khmer srәw rice (unhusked)
Khmuic Khmu [Cuang] sroːɲ dry (paddy rice)
Monic Proto-Monic *srooʔ
Pearic Chong [Samre] srôː rice store-house, barn
Vietic Thavung alɔː³ paddy rice

Table 2. Husked rice in Proto-Austroasiatic

Branch Language Attestation Gloss
Munda Proto-Munda *r.ŋko rice
Munda Remo rŋku uncooked rice
Aslian Sakai rĕkuaʼ husked rice
Khasic proto Khasic *kʰaaw rice
Khmuic Khmu [Cuang] rŋkoʔ husked rice
Palaungic Proto-Wa-Lawa *rŋkoʔ uncooked rice
Pearic Chong [Trat] rәkʰәw riz blanchi
Vietic Chứt [Rục] rәkóː husked rice

None of the other roots for domestic plants listed in Zide & Zide (1972) have clear proto-Munda forms, nor
are the roots obviously connected to the same plants in SE Asia. No Munda domestic animal names show SE
Asian connections except for the chicken. The Austroasiatic root *siәr/N seems to reflect Proto-Munda
(Table 3);

Table 3. Chicken in Austroasiatic

Branch Language Attestation Gloss
Munda Proto-Munda *(g)sim chicken
Khasic proto Khasic *sʔiar chicken
Khmuic proto Khmuic *(s)ʔiәr chicken
Katuic Katu [An Diem] siem chicken

This is rather surprising, since it suggests that the population which migrated carried a very reduced
repertoire of domestic plants and animals, which perhaps points to the conditions under which they, perhaps
forcibly and certainly not as intentional colonists.
2.2 Archaeology: the Eastern Wetland Tradition
The Neolithic archaeology of Odisha [Orissa in titles of references] could hardly be said to be well
understood and even less well dated. Nonetheless, there are some indications of what Harvey et al. (2006)
call an ‘Eastern Wetland Tradition’ which is characterised by an agricultural package including rice, pulses
such as the mung bean and probably root crops such as taro. These are found at ‘impressive mound sites’
3
Roger Blench Did the Munda reach India by sea? Circulation draft
such as Gopalpur, and are given the speculative date of ‘early Third Millennium BC’, although this is not
supported by radiocarbon dates from stratified sites (Kar et al. 1998). Another possible site representing this
culture complex is Bonaigarh (Behera 2000). A lack of well-illustrated and analysed pottery and generalised
references to stone tools makes it difficult to compare these materials with the SE Asian region, but the
agricultural complex identified is not incompatible with the development of lowland rice agriculture
introduced by Munda speakers. Harvey et al. (2006) contrast this with the subsistence strategies in the
surrounding highland areas, which remain shifting agriculture, much as they still are today.
3. Dating the Munda dispersal
If the Munda peoples already had an agricultural lifestyle prior to their migration, then this must be after the
advent of the SE Asian Neolithic, i.e. subsequent to 4000 BP. However, the archaeology of the east coast of
India shows no evidence of the typical signatures of Austroasiatic expansion, incised and impressed
ceramics, etc. (Rispoli 2008). This points to cultural loss on the SE Asian mainland, which would be
common to Nicobaric and Aslian. This suggests that a core rice-growing/foraging population was
established somewhere on the southern Myanmar coast, but which had passed through a bottleneck losing
characteristic Austroasiatic ceramics. The Indian side provides no evidence for SE Asian technologies in
iron production, most notably the paired tube bellows and Munda rather shows that iron smelting reflects
contact with Indo-Aryan speakers. The Munda word for ‘iron’ itself is highly fragmented and clearly cannot
be reconstructed to proto-Munda. This means the migration was later than around 2200 BP. Plausibly, then it
occurred in the window 3800-2200 BP, but presumably earlier in this window.

Photo 1. Split-tube bamboo

4. Locating the Munda dispersal
rattle, Sora
Where would this dispersal have started? We do know that the route
between the east coast of India and the isthmus of Kra was exploited by
Indian shipping from around 200 BC (Bellina 2017). However, the maritime
cultures of the east coast of the isthmus, typically Austronesian, began
circulating in the region much earlier, from 4000 BP onwards.

Munda shows no trace of an affinity for the sea in its lexicon. Words for
‘sea’, ‘ocean’, ‘boat’ show no common roots and are clearly not
reconstructible. However, Blench (2018a) notes that various river species
and capture techniques do show cognates between Munda and SE Asian
Austroasiatic. So if the early Munda travelled in ships, these were most
likely not their own and they therefore display no culturally embedded
maritime tradition. Significantly, in the explosive Austronesian dispersal
from southern Taiwan from 4000 BP onwards, ships rapidly reached
Sumatra and also travelled between Java and Sumatra to reach the offshore
islands to the west, such as Nias. From Nias to the Nicobar islands is a
relatively short journey, and while the eastern Indian seaboard is somewhat
further, for such experienced navigators, not inconceivable.

If the Munda were not the navigators, why were they aboard the ships?
Presumably as crew, forced or otherwise. This practice would have
significant parallels in the Austronesian world, as the ancestors of the
Malagasy, the Barito, were also an inland people, not open ocean
navigators. Yet they were taken on by Malay captains and carried to East
Africa and thence to Madagascar, whereby they populated the island.
Similarly, the population of the Canary Islands, the Guanche, were
apparently Berbers who were landed on the islands in order to collect purple Source: Author photo, Tribal
dye-shells, a trade which then collapsed, leaving the settlers isolated Museum, Bhubhaneswar
(Blench 2018b). In a parallel situation, the early Munda either escaped on
reaching India, or were dropped to become traders and farmers. Their ancestral population must have been
located somewhere on the southern coast of Myanmar or adjacent Thailand. A small number went aboard an

4
Roger Blench Did the Munda reach India by sea? Circulation draft
Austronesian ship and were carried to the coast of India, from where they dispersed both north and west, to
reach their present destinations.
5. Material culture correlations

Photo 2. Split-tube bamboo rattle,

5.1 Material culture as an indicator
Sulawesi
SE Asian societies have a rich repertoire of material culture,
typically made of vegetable material, which does not preserve in
the archaeological record. Baskets, traps, mats and other
implements are highly distinctive and very dissimilar to those
made by the Indo-Aryans. As a consequence, their idiosyncrasies
can often provide indications of migration through their
distribution. This section gives the examples of two musical
instruments whose morphology an distribution points strongly to a
maritime transmission. Musical instruments are good proxies for
migration, since they are essentially arbitrary. As human
constructs, they do not necessarily converge morphologically as
natural entities do. In the case of the proposed connection, two
instruments have a specific form which occurs on the east coast of
India and in Southeast Asia.
5.2 Two musical instruments

5.2.1 The split-tube rattle

This instrument consists of a bamboo internode with one node cut Source: Author photo
away and the remaining section cut
Photo 3. Split-tube rattle, Sumba
vertically into a series of long thin strips
joined to the lower joint at the base. A
segment of bamboo is allowed to remain
beyond the lower joint to act as a handle.
The instrument can be rolled between the
hands or struck against the palm, and it
creates a crackling sound as the strips strike
one another. These instruments are typical
of ISEA and are also found in a small area
of the Odisha. Photo 2 shows an example
from the island of Sulawesi, and Photo 3
shows a similar instrument from Sumba. Source: Author collection
These examples may not be bamboo, but rather a long thin gourd, hollowed out and cut in the same shape.
Photo 1 shows a very similar instrument from the Sora people in Odisha.

It should be noted that this instrument is found only here and nowhere else in the world, i.e. it is not
something which is re-invented multiple times.
5.2.2 Monochord bowed mouth-resonated zither
Another highly unusual musical instrument found in SE Asia and among the Sora in the monochord bowed
mouth-resonated zither. There is a single string, tightened with a peg which is stretched along a wooden bar.
The string is sounded by bowing, but the sound is amplified and modified by the mouth of the player. Photo
4 shows an example of this instrument played by the Sora, and Photo 5 a similar instrument played by the
Bahnar in Vietnam.

5
Roger Blench Did the Munda reach India by sea? Circulation draft
Photo 4. Sora monochord bowed mouth-resonated zither

Source: Author photo, Tribal Museum, Bhubhaneswar

Photo 5. Bahnar monochord bowed mouth-resonated zither

Source: Author photo, Vietnamese Institute of Musicology, Hanoi

Like split-tube bamboo rattle, it can be underlined that this instrument is unique to these two areas, and does
not occur elsewhere in the world.
6. Some linguistic counter-examples
It would be unscientific not to highlight some possible counter-arguments, especially linguistic. Munda
languages do appear to share some specific links with Khasian, which would sit easily with a maritime
dispersal. Table 4 shows the word for ‘sun’ in Munda and Khasic, showing an almost exact match between
presyllable and root, where the prefix is different elsewhere in Austroasiatic.

Table 4. ‘Sun’ in Munda and Khasic

Language Attestation
Gta' sni
Bondo [Hill] siŋi
Mundari [Nijpara] siŋɡi
Ho [Ghatshila] siŋˈgi
Proto-Khasic *sŋi
Proto-Mon-Khmer *tŋiiʔ

Table 5 shows the word for ‘dog’ in Munda and Khasic, where both have a velar prefix, in contrast to the
remainder of Austroasiatic which usually has zero.

6
Roger Blench Did the Munda reach India by sea? Circulation draft
Table 5. ‘Dog’ in Munda and Khasic
Language Attestation
Gta' gsu
Remo gusoo
Gutob gusoʔ
Bondo [Hill] gu-soʔ
Bodo-Gadaba gusɔ
Sora kinso:ʈ
Proto-Khasic *ksәw
Proto-Mon-Khmer *cɔʔ

These may be chance. Other lexical items, for example ‘husked rice’ (Table 2) show exact presyllable
matching with other Austroasiatic branches and not with Khasic.
7. Conclusions
The linguistic geography of the Munda languages has long been a puzzle, since it was presumed that their
core population must have dispersed by land, passing around north of the Bay of Bengal. Munda shows no
special relationship with Khasian, its nearest geographical relative, nor does the pattern of languages suggest
such a migration. A more credible proposal is thus the maritime dispersal proposed by Paul Sidwell,
although we have no evidence for a seagoing tradition in this region at a presumed early date such as 3800
BP. This problem is resolved if the migrant rice-farmers were carried in Austronesian ships, since we do
indeed have evidence for extensive and long distance voyages from 4000 BP onwards. We do know the
Nicobaric core populations arrived in their islands by unknown means at about the same period. This would
also explain why no Munda cultures show any affinity for maritime subsistence. Map 3 illustrates this
primary dispersal schematically;

Map 3. Primary dispersal of Munda speakers

The map includes the hypothesis that the dispersal of Nicobaric was part of the same general process of
migration as a consequence of the Austronesian maritime culture, although Nicobaric and Munda show no
special linguistic relationship.

Map 4 is more speculative, but outlines a possible model of Munda dispersal within India, incorporating the
notion that South Munda is a spurious grouping. It assumes that the newly arrived population first stayed on
the coast and continued both rice production and extensive foraging. This culture was upended by the
incursions of Indo-Aryans who drive them away from this core area, at the same time transforming their
agriculture with new species and techniques.

7
Roger Blench Did the Munda reach India by sea? Circulation draft
Map 4. Dispersal of Munda speakers within India

Archaeological evidence remains weak, although we do know that a rice agriculture appears in the lowlands
at around the same time, the Eastern Wetland Tradition, which is implanted in a matrix of foragers and
shifting cultivators. It would be feasible to connect this with the arrival of the core Munda. Synchronic
material culture, although undated, provides additional evidence of the connection across the Bay of Bengal,
including at least thwo musical instruments of SE Asian provenance, which are not otherwise characteristic
of mainland India.

Much remains to be done, including rethink the dispersal of the Munda languages in the light of this new
proposal, and in particular shedding better light on the Neolithic of Odisha, which would provide the
material evidence to support the linguistic proposal.
References
Basa, K.K., 1999. Neolithic and post-Neolithic cultures of Orissa: an overview. Orissan History, Culture
and Archaeology, New Delhi.
Behera, P.K., 2000. Neolithic culture complex of Bonaigarh, Orissa. Archaeology of Orissa, 1: 222-263.
Bellina, B. 2017 (ed.). Khao Sam Kaeo: an early port-city between the Indian ocean and the South China
sea. Ecole française d’Extrême-Orient. Paris : Mémoires Archéologiques 28, Ecole française d’Extrême-
Orient.
Blench, R.M. 2011. Was there an Austroasiatic presence in island SE Asia prior to the Austronesian
expansion? Bulletin of the Indo-Pacific Prehistory Association, 30: 133-144.
Blench, R.M. 2014. The origins of nominal affixes in MSEA languages: convergence, contact and some
African parallels. In: Mainland Southeast Asian Languages: The State of the Art. N.J. Enfield and
Bernard Comrie eds. Canberra: Pacific Linguistics.
Blench, R.M. 2018a. Waterworld: lexical evidence for aquatic subsistence strategies in Austroasiatic.
JSEALS, Special Issue 3. Felix Rau & Hiram Ring eds. 174-193. University of Hawai’i Press.
Blench, R.M. 2018b. Relating linguistic reconstructions of plant names in Berber to the archaeobotany of
North Africa. Presentation at International Workshop African Archaeobotany IX University of Las
Palmas de Gran Canaria. 26-29th June 2018.
Blench, R.M. and Paul Sidwell 2011. Is Shom Pen a distinct branch of Austroasiatic? Austroasiatic studies.
ICAAL IV. Mon-Khmer Studies, Special Issue 3. 9-18.

8
Roger Blench Did the Munda reach India by sea? Circulation draft
Harvey, E.L., Fuller, D.Q., Mohanty, R.K. and Mohanta, B. 2006. Early agriculture in Orissa: some
archaeobotanical results and field observations on the Neolithic. Man and Environment, 31(2): 21-32.
Kar, S.K., Basa, K.K. and Joglekar, P.P., 1998. Explorations at Gopalpur, District Nayagarh, Coastal Orissa.
Man and Environment, 23(1), pp.107-114.
Sidwell, Paul & Roger M. Blench 2011. The Austroasiatic Urheimat : the Southeastern Riverine Hypothesis.
In: N. Enfield ed. Dynamics of Human Diversity in Mainland SE Asia. 317-345. Canberra: Pacific
Linguistics.
Sidwell, Paul 2000. Proto South Bahnaric: a reconstruction of a Mon-Khmer language of Indo-China.
Canberra: Pacific Linguistics.
Sidwell, Paul 2005. The Katuic Languages: classification, reconstruction and comparative lexicon. Lincom
Europa.
Sidwell, Paul 2014. Khmuic classification and homeland. Mon-Khmer Studies, 43.1:47-56.
Sidwell, Paul 2014. Proto Khasian: an emerging reconstruction. In: Gwendolyn Hyslop, Linda Konnerth,
Stephen Morey, Priyankoo Sarmah (eds.), North East Indian Linguistics Volume 6, pp. 149-163.
Canberra: Asia-Pacific Linguistics.
Sidwell, Paul 2015. Austroasiatic classification. In Jenny, Mathias and Paul Sidwell, eds (2015). The
Handbook of Austroasiatic Languages. Leiden: Brill.
Sidwell, Paul 2015. The Palaungic Languages: Classification, Reconstruction and Comparative Lexicon.
Lincom GmbH.
Sidwell, Paul and Felix Rau 2015. Austroasiatic Comparative-Historical Reconstruction: An Overview. In:
Jenny, Mathias and Paul Sidwell eds. 1:221-263. The Handbook of Austroasiatic Languages. Leiden:
Brill.
Sidwell, Paul and Jacq, P. 2003. A Handbook of Comparative Bahnaric: volume 1-West Bahnaric. Canberra:
Pacific Linguistics.
Zide, Arlene R.K. & Norman H. Zide 1976. Proto-Munda cultural vocabulary: evidence for early
agriculture. In: Austro-Asiatic Studies, Part II. P.N. Jenner, L.C. Thompson & S. Starosta (eds.) 1295-
1334. Honolulu: University of Hawai’i.