7. The Andamanese

Numbers

by George Weber

How many Andamanese lived on their archipelago when the British arrived to stay in 1858? And how many are still there today? These questions may sound simple enough but, like almost anything else concerning the Andamanese, a thick fog of uncertainty is blurring the answers.

The first rough estimates of numbers came from British sea captains who had sailed in the island waters before and during the establishment of the short-lived British settlement 1789-1796. Their figures ranged from 2000 to 4000 people and referred only to the Great Andamans. At that time, only the shore-dwelling Aryoto were known; the interior of the islands was thought to be impenetrable and uninhabited. The Onge of Little Andaman were not counted.

Downhill all the way 1: the Andamanese Negrito population.

Downhill all the way 2: the Andamanese Negrito as a percentage of the total island population.

 

Up, up and away: the total Andaman islands population.

After the establishment of a permanent settlement at Port Blair in 1858, estimates of the total number of natives were published but had to be continuously revised. All remained uncertain guesswork and, to tell the truth, the situation remains essentially unchanged today. In the early decades, the inland-dwelling Eremtagas were discovered and new groups kept appearing from obscurity thereby increasing estimates. Simultaneously epidemics swept through the islands, reducing the native population dramatically. The last two tribes to be discovered, the Aka-Bo and the Aka-Kora, were found in the course of work leading up to the 1901 census. The small number of Aka-Bo was explained to the census-takers as the result of a recent epidemic that had come from the neighbouring Aka-Kari and Aka-Kora. Having no contact with the outside world provided no protection from the new diseases. The Aka-Bo dealt with the problem by killing all of their own who showed symptoms. Such drastic methods did not help them survive in the long run but may explain why their numbers actually increased between the 1901 and 1911 census.

During the four decades between settlement in 1858 and the census of 1901, only two estimates need be mentioned. Mr. Man complained in 1885:

the utter hopelessness of obtaining from the aborigines any correct idea of the population of the tribes individually and collectively will be readily understood when it is stated that the only numerals in the language are those for denoting 'one' and 'two' and that they have absolutely no word to express specifically any higher figures, but indulge in some such vague terms as 'several', 'many', 'numerous', 'innumerable' which seem to convey to their minds an approximate idea of the number intended, but fail to satisfy the requirements of the statistician.

He then went on to estimate the numbers himself and came up with a total of 3,250 for 1883. Another estimate two years later gave a total of 6,600. 

Tribal population figures 1858-1995. 

Tribe	1858	1883	1901	1911	1921	1931	1951	1961	1971	1981	1995
Great Andamanese
Aka-Cari	100		39	36	17	9
Aka-Kora	500		96	71	48	24
Aka-Bo	200		48	62	16	6
Aka-Jeru	700		218	180	101	46
Aka-Kede	500		59	34	6	2
Aka-Kol	100		11	2	0	0
Oko-Juwoi	300		48	9	5	0
A-Pucikwar	300		50	36	9	1
Akar-Bale	300		19	15	4	2
Aka-Bea	500		37	10	1	0
Sub-Total	3500	2000	625	455	207	90	33	19	24	24	37
Onge-Jarawa
Onge	700		672	631	346	250	150	129	112	106	99
Jarawa/Jangil	60	1250	585	231	231	120	50	56	25	250?	>200
Sentineli	?		385	117	117	?70	?60	50	100	50?	80?
Sub-Total	1300	1250	1642	979	694	440	260	235	237	406?	379?
Grand Total	4800	3750	2267	1434	903	530	293	254	261	430?	416?

The census of 1901 tried to put things on a more systematic footing but it, too, ended with a rather unsatisfactory compromise. A few years earlier, Portman had estimated the total number of all Andamanese in 1858 at 8000. The colonel in charge of the census of 1901 disagreed and, thinking Portman's figure too high, came up with the figures of 4800. Whatever the relationship of the colonel's figures to reality, his figure became The Official Estimate.

Nothing certain is known about the numerical split between Aryotos and Eremtagas. There are a number of rather inconclusive hints, based on an analysis of food resources and the number of kitchen-midden in the interior, that imply a rather greater number of Eremtaga than Aryoto prior to 1858.

The figures of the 1901 census by male/female and adult/children.

Tribe	Number of adults (male/female)	Number of children (male/female)	Total	Children as % of total
Great Andamanese
Aka-Cari	31 (16/15)	8 (6/2)	39	20.5%
Aka-Kora	63 (31/32)	33 (14/19)	96	34.4%
Aka-Bo	31 (15/16)	17 (7/10)	48	35.4%
Aka-Jeru	178 (98/80)	40 (26/14)	218	18.3%
Aka-Kede	54 (24/30)	5 (3/2)	59	8.5%
Aka-Kol	8 (6/2)	3 (3/0)	11	27.3%
Oko-Juwoi	40 (21/19)	8 (7/1)	48	16.7%
A-Pucikwar	45 (31/14)	5 (2/3)	50	10.0%
Aka-Bea	15 (5/10)	4 (3/1)	19	21.0%
Akar-Bale	30 (14/16)	7 (3/4)	37	18.9%
Total	495 (261/234)	130 (74/56)	625	20.8%

As the figures for the Great Andamanese tribes grew more illusory from census to census, by way of compensation they have become ever more reliable. Tribal society had largely broken down by the time of the 1901 census. For some decades previously there had been a growing intermingling of survivors from all tribes. The census officers on tour for the 1911 census witnessed many inter-tribal marriages, including some matrimonial crossings of the line between the southern and northern groups of Great Andamanese. The 37 persons said to be of Great Andamanese extraction in 1995 were in fact all of mixed Burmese/Indian/Andamanese ancestry.

While the Great Andamanese tribes merged and then melted away to become mere traditional shadows, the Onge-Jarawa tribes were separated by geography and remain unmixed to this day. If the figures for this group above look a little chaotic, that is because they are. Firstly there is a historic confusion around the name "Jarawa." The early investigators, especially Mr. Man used it to cover the Onge, what we today call the Jarawas and by implication the Sentinelis although he does not seem to have been aware of the latter's existence. Side by side with the older usage, some researchers limited the name to the two hostile northern groups, calling the Onge by their own name. Since the 1950s the three groups are called Onge, Jarawa and Sentinelis although a few still insist on including the Sentinelis among the Jarawas. This explanation is necessary to clarify why the earlier figures leave it unclear what groups are included or excluded.

Then there are the figures on the Jarawas and Sentinelis themselves. The Sentineli were, whenever they were mentioned at all which was not often, counted as "Jarawa." The Census of 1911 for the first time mentioned the Sentineli as a separate group, estimated at 117 people, but subsequent counts simply repeated this figure and the 1951 census even dropped all mention of them. The official figures, British as well as Indian, relating to the Sentineli were never more than vague guesstimates. As a result, there is a confusion of numbers bandied about in literature today.

But consider the formidable practical problems in estimating, let alone counting, these inaccessible, hostile and highly mobile populations, living as they do under an impenetrable jungle canopy. The only practical way is to fly over their extensive territory and to count the number of visible columns of smoke that indicate aboriginal groups cooking dinner. It all has do be done over a considerable territory in one go around meal time in good weather and in daylight. This does not leave a lot of time. Groups can move overnight and might be counted twice if the operation is spread over two days, for example. Once the number of smoke columns has been established, multiply with the number of people each group might reasonable be expected to contain (a pure guess itself), add an additional group for safety because you might have missed one, and there is your estimate. It cannot be a work of high precision and it is expensive.

Just how much estimates of Jarawas and Sentinelis vary even when all base their reports on the official Census of India figures is shown below.

Discrepancies in estimates of Sentineli and Jarawa populations.

Year	Sentineli					Jarawa
	Mathur L.P 1968	Pandit T.N 1990	Singh K.S. 1994	anthrop. museum Pt. Blair 1995		Mathur L.P. 1968	Pandit T.N. 1990	Singh K.S. 1994	anthrop. museum Pt. Blair 1995
1951:	?	50	?	?		?	50	50	>50
1961:	50	500	50	>500		500	500	500	>500
1971:	?	100	82	>100		?	250	275	>300
1981:	?	100	100	>100		?	250	200	>250
1991:	?	80-100	?	>150		?	200	?	>250

The tenfold increase in the number of both populations between 1951 and 1961 is not due to a unique upsurge in fertility but apparently to a re-evaluation of the 1951 census figures by some anthropologists. If sources other than those given above are consulted, it will be found that everyone has a favourite figure but that few agree.

Recent Onge population figures.

Age Group (estimated)	1963	1965			1969
	total	male	female	total	male	female	total
0-14 years		23	16	39	20	19	39
14-24 years		8	7	15	9	8	17
25-34 years		15	16	31	14	11	25
>35 years		18	26	44	18	22	40
Total	132	64	65	129	61	60	121

Living children per woman: 1.55

Even accessible and cooperative Andamanese groups can be difficult to count, as the Italian anthropologist Cipriani found in the 1950s. Counting the number of bedsteads, he did not take into account the fact that the Onges were nomadic and that many if not most had several bedsteads to lay their weary heads. Two new counts carried out during the 1960s by Indian anthropologists using indelible inks to avoid multiple counting produced the first reasonably reliable figures for the Onges.

When the Aka-Bea living around the coasts of southern Great Andaman and Rutland island died out and the Jangil (Rutland Jarawas) living inland vanished during the last years of the 19th century, some Onge from Little Andaman sized the chance and moved into the vacated territory. The same epidemics that had dispatched the previous owners then forced them to retreat. The population of this short-lived Onge community has been estimated at 80 persons in 1911, 61 in 1921 and nil in 1931. All they left behind was confusion in scientific nomenclature between Jarawa and Onge.

There is a rather colourful pointer at the unreliable nature of figures concerning the 19th century Great Andamanese. These are the figures that are so uncritically bandied about in literature regarding the attackers involved in the Battle of Aberdeen. There were three major attacks on Port Blair (of which Aberdeen is a district) within the space of 5 weeks in 1859, with only the final known somewhat melodramatically as the Battle of Aberdeen. In the first attack on 6th April 200 Andamanese are said to have taken part, no less than 1500 in the second of 14th April and 250 in the third of 17th May. The fighting took place only one year after the arrival of the British when the traditional tribal structure was still intact and numbers had not yet been reduced by disease. This meant that the various tribes had very little to do with each other. The attacking tribe was the Aka-Bea and they are said to have numbered around 500 persons in 1858, including women and children. Something is not quite right here. Males did all the fighting so that even when counting very generously and even if many women came along for the plundering, the Aka-Beas could not possibly have got more than 400 people on the warpath. With 1500 warriors we are looking at a substantial part of the fighting power of all Great Andamanese tribes combined - and this at a time when the southerners were not even aware of the existence of the northern tribes, and vice versa. Although it seems unlikely and we hear nothing about it, there may have been some fighters from neighbouring tribes, the Aka-Bale and A-Pucikwar, helping the Aka-Bea. But 1500?

Anthropologist Radcliffe-Brown, doing research in the Andamans 1906-08 but writing up his results more than a decade later, took a middle course as regards the native population of 1858. He found Portman's 8000 too high and the census figure of 1901 of 4800 too low. He himself estimated a total of 5650 people of which 3750 Great Andamanese, 1200 Aka-Bea and Jarawas and 700 Onge and Sentinelis. The lumping together of Aka-Beas and Jarawas is odd but Radcliffe-Brown was interested in the population densities on South Great Andaman, not in the number of members per tribe. The Aka-Bea and Jarawas hated each other but they did share the same island. Radcliffe-Brown explained that his estimates were based on what the Andamanese of 1908 were able to tell him of the conditions prevailing 2 or 3 generations before, around 1860. He qualified his figures by stating that his estimate could only be of relatively small value and that if anything he had under- rather than over-estimated.

If we must pull a 'most likely' figure out of the hat, we could do worse than accept a fairly stable total population of around 7000 for the time before 1858 and perhaps 400 today. However uncertain the figures, they show with devastating clarity the slide of the Andamanese race towards extinction. How dramatic the situation really is can best be seen in graphic form. At the beginning of the slide were devastating epidemics, then there seems to have been a psycho-cultural collapse with the will to survive severely impaired among all tribes in contact with outsiders. Perhaps there was a small chance until the early decades of the 20th century for the Great Andamanese to climb back but it was not to be. The ever-growing tide of government-sponsored Indian settlers from the late 1940s finally put paid to any such hopes. The Onge on their isolated island have fared slightly better so far but the trend of their population figures is quite unambiguous, too. 

 

[ Go to HOME ] [ Go to ANDAMAN Table of Contents ]   Click on icon for Andamanese video clips

 

Last change 20 March 1998