Of the hundreds of tribal groups still found in South Asia, only a few dozen are represented on our map. Among those the Gonds, Bhils, Santals, Khonds, and Oraons in India proper and the Shans, Karens, and Kuki-Chins in Burma are most ex- tensive in distribution as well as most populous. It should not be thought, however, that the existence of a common designa- tion for a particular tribal group implies a sense of unity within that group. Typically, the functioning tribal unit, led by a local chieftain, accounts for only a small fraction of the total pop- ulation with a given tribal name. There are many named sub- groups (dozens, for example, among the Nagas), often with mutually incomprehensible languages with which our map does not deal. Further, it should not be supposed that particu- lar tribal territories are mutually exclusive. This is often not the case, especially in the Chota Nagpur region of northeast- ern India, where different tribal groups are greatly intermixed, several of them commonly living together in a single village.

Sources

For India

India, Census (1911), (1921), (1931).

For Ceylon

Same as those listed for plate X.C.1.

X.D.1. Culture Realms, Regions, and Areas, c. 1961

Here we seek to provide a spatial synthesis, within a hier- archically ordered system of culture "realms," "regions," and "areas," of the principal elements depicted in the religious, linguistic, and other ethnic maps in sections X.A, B, and C of this atlas. Although at the time of this writing religious data for 1971 are available to us for India, they are lacking for other areas; so we have utilized the data of 1961. The year 1961 is also the latest for which linguistic data are available. For data on the composition of the population by caste, tribe, or comparable ethnic division, we have necessarily had to rely primarily on data from the 1931 census, the nature and reli- ability of which is discussed in our introduction to the text for subsection X.C. There too we discuss the sorts of adjustments one would have to make in the data in applying them to the contemporary ethnic scene, as we have tried to do for our present purposes.

In calculating the estimated population of our culture areas, as named in the table below our map, we have aggregated the 1961 census data, wherever available, for the component dis- tricts, making estimates for the portions of a given district falling within each of two or more areas when it lies astride one or more of our boundaries. For Ceylon, 1961 populations were interpolated from those of the 1953 and 1963 censuses. For noncensused areas, as of 1961, we have utilized a variety of estimates and occasionally made our own. Our figures are given to the nearest million where they are over 10 million, to the nearest hundred thousand for areas with populations be- tween 100,000 and 10 million, and to the first significant figure for areas with fewer than 100,000 inhabitants. The estimates of the proportional strength of particular religious, linguistic, and caste, tribal, or other ethnic groups, as indicated by the typographic conventions explained in the notes at the end of the table, are within broad percentage ranges and can normally be considered reliable. Where we are in doubt, however, in the case of castes, tribes or other ethnic groups, we have indicated that doubt by a question mark.

Sources

The sources used for plate X.D.1 are those utilized for pre- vious plates of section X that it synthesizes. For a theoretical regional viewpoint, see J. E. Schwartzberg (1967). For alter- native perspectives on regions and regionalism in South Asia, see R. I. Crane (1967), in which the aforementioned essay appears.