Linguistic Families of American Indians North of Mexico

Linguisitic Families of American Indians - Powell
Powells Map showing the Linguisitic Families of the Indian Tribes of the United States

From the date of its first appearance in 1891 the Powell map of “Linguistic Families of American Indians North of Mexico” has proved of the widest utility. It has been reissued several times and copied into numerous publications. There has, however, been almost equal need of a map giving the location of the tribes under the several families.

To one familiar from his readings in early American history with the names and locations of our prominent eastern “tribes,” such as the Delaware, Iroquois, Cherokee, and Choctaw, the preparation of a tribal map would seem to be simple, and it would indeed be so if all Indians had been grouped into bodies as clearly marked as those mentioned. But even in the eastern United States the term “tribe” is quickly found to have no uniform application. The Creeks were a confederation of a few dominant tribes and a number of subordinate bodies, each formerly independent. The name “Delaware” is commonly said to have covered three tribes or sub-tribes, but while two of these seem never to have been independent of each other, the third, the Munsee, is often treated as if it were entirely separate. The name “Powhatan” was applied to about 30 tribes or sub-tribes which had been brought together by conquest only a few years before Virginia was settled, and the term “Chippewa,” or “Ojibwa,” is used for a multitude of small bands with little claim to any sort of governmental unity. In the case of the Iroquois, on the other hand, the tribe was only a part of the governmental unit, the Iroquois Confederation, or Longhouse.

The northern Plains tribes present a certain coherence but farther south and west our difficulties multiply. An early explorer in Texas states that in that region, by “nation” was to be understood only a single town or perhaps a few neighboring villages, and in fact the number of tribal names reported from this section seems almost endless. In the governmental sense, each Pueblo community was a tribe, and if we were to attempt a complete list we should have in the first place a large number of existing, or at least recently existing, tribes, little and big, and a still greater number known only through the early writers or by tradition. In California, Kroeber (1925) states that there were no tribes in the strict sense of the term except among the Yokuts of the San Joaquin Valley and their immediate neighbors. Elsewhere in California, and in western Oregon and Washington as well, tribe and town might be considered convertible terms. As the number of these was continually shifting, it would be impracticable to enter them in that capacity in a work of the present kind.

North of the International Boundary, conditions are, if possible, worse, except in the southernmost section of Canada where lived tribes similar to those in the eastern parts of the United States, such as the Huron, Chippewa, Assiniboin, and Blackfoot, though the Chippewa, as already mentioned, require a somewhat elastic extension of our common concept of a tribe. On the north Pacific coast, however, the conditions noted in western Oregon and Washington are continued. We have numerous local groups associated into several major divisions on linguistic grounds alone. Still farther north and east, among the Algonquians, Athapascans, and Eskimo, we are confronted with a bewildering array of bands and local groups, usually confined to one town and taking their name from it or from a certain territory over which its members hunted, and the numbers and names of these are uncertain even at the present time. Nothing remotely resembling scientific accuracy is possible in placing these bands, if we aim at chronological uniformity, and we must either enter great linguistic groups, embracing sometimes almost an entire stock, or make an arbitrary selection of bands with the idea of including those which we esteem the most important.

Northeastern México and some parts of Central America may also be defined as band areas, but most of North America below the Río Grande was occupied by well-recognized tribal divisions. From all of the West Indies except Haiti, Cuba, and Puerto Rico nothing like a complete list of tribes has survived, and even for the best documented of these, Haiti, it is impossible to say how many of the caciquedoms mentioned should be given tribal status.

A short study of the conditions above outlined shows that only two alternatives are open in a work like the present. Either one must, in effect, alter it to a town and band map, entering the most minute recorded subdivisions and setting his results forth, not on one map but on dozens, or he must be satisfied with a relatively conventional classification, having in view popular convenience rather than scientific uniformity, and making the best grouping he can of those peoples which did not have real tribal organizations. In the present undertaking the latter plan has been followed, but clues to the more scientific study have been given by including lists of “subdivisions” and “villages.” There is no profession that these lists are complete; a perfect presentation of them would demand an investigation for which there is as yet no opportunity. The rest of the accompanying text has been devoted to certain items of information likely to be called for first by the general reader, including: the origin of the tribal name and a brief list of the more important synonyms, the linguistic connections of the tribe–it has not seemed feasible to try to include the physical and cultural connections–its location, a brief sketch of its history, its estimated and actual population at different periods (based mainly on Mooney’s (1928) study and the reports of the United States and Canadian Indian Offices), and the “connection in which it has become noted,” particularly the extent to which its name has been perpetuated geographically or otherwise. I have also included references to the more important sources of information. Two works have been used as basal authorities. One, the Handbook of American Indians (Hodge, 1907, 1910), is general in scope and may be assumed throughout except for the tribes of México, Central America, and the West Indies. The other, Kroeber’s Handbook of the Indians of California (1925), is the basal authority used in treating the Indian groups of that State. In the Gulf area I have utilized the results of my own studies, published and unpublished.

As far as possible each tribe, or group has been treated by itself, but in Washington, Oregon, California, and Alaska, to avoid needless repetition, the history of the tribes is considered as a whole. The section on México, Central America, and the West Indies represents an afterthought. Both map and text material were drawn originally from the “Indian Languages of Mexico and Central America” (Thomas and Swanton, 1911), and Dr. Lehmann’s (1920) monumental work on “Zentral Amerikas,” but they have been made over thoroughly in the light of the classification and map of Dr. J. Alden Mason (1940) and Frederick Johnson (1940), and no attempt has been made to take up the history of the several tribes or indicate other authorities.

A brief history of the present undertaking will perhaps enable the reader to obtain a better understanding of it, appreciate the difficulties encountered in the compilation, and in consequence view its short­ comings, of which as the compiler I am keenly aware, with due charity. It represents an evolution both in method of procedure and in the extent of territory covered. In the beginning I was governed by the older tradition regarding map work of the kind, the idea of entering a tribe in the place where it was first encountered by Whites, but an attempt to carry out this plan soon presented difficulties because neighboring tribes were often encountered a century or more apart and their relative positions may have changed utterly in the interval. There is no certainty, for instance, that the Indians outside of the narrow strip of territory opened to our vision by De Soto’s army in 1539-43 were in the same relative position when Carolina was settled about 1670 and Louisiana in 1699. It is particularly to be noted that, while De Soto found eastern Arkansas full of towns, it was almost deserted when Marquette and La Salle visited it in 1673 and 1682. We also know that great alterations took place in the St. Lawrence Valley between the voyages of Cartier in 1534-43 and Champlain’s appearance there in 1603.

In view of these difficulties, I gave up this plan and tried the device of putting each tribe in the region with which it was most closely associated historically. But with what region were the Shawnee, Cheyenne, Arapaho, Kiowa, Comanche, and some other tribes most closely associated? The Middle West or the Plains are rather too general terms. Moreover, tribes acquired this close association with certain sections at very different periods and, if this plan were carried out, the map as a whole would be historically inaccurate. Thus the Delaware upon the whole were associated most closely with the valley of the river which bears their name, but when the Foxes had reached Iowa and the Dakota had occupied South Dakota, where they are best known, the Delaware had removed many hundred miles from this region. The Abenaki were most closely associated with western Maine but were uprooted in the middle of the eighteenth century and moved to Canada. The Huron are most closely connected historically with the region of Lake Simcoe, Ontario, but they were driven from there in the middle of the seventeenth century, and a hundred years later under the name Wyandot they, or at least part of them, came to be “closely associated” with Ohio. Thus we have here two associations of the same tribe.

For a time it seemed as if some of these inconsistencies were unavoidable and that any attempt at chronological accuracy was out of the question. Such is indeed the case if we insist upon absolute, documented accuracy, because Alaska, western Canada, and the northwestern part of the United States were almost wholly unknown until the latter half of the eighteenth century and there is no authentic information regarding many tribes until the beginning of the nineteenth when many eastern tribes, and some of those on the Plains, had been displaced or destroyed. But on experimenting along this line I discovered that if we select the year 1650, or rather a few years prior to that date and assume a fairly static condition for 30 or 40 years afterward, we can determine the location of most of the tribes of the eastern and southern United States and eastern Canada in a fairly satisfactory manner, and this arrangement was finally decided upon. Up to 1649 the Hurons were still in Ontario; the Erie, the Neutral Nation, and the Susquehanna had not been destroyed by the Iroquois; and King Philip’s War, which was to scatter the New England Indians, did not break out until 1675. The Virginia Indians had suffered very much as a result of their risings in 1622 and 1644 but continued to occupy the same general territories in which the colonists found them. By 1650 the Gulf region had been traversed by Spanish expeditions and Florida had been settled nearly a hundred years, but there had been little displacement of the aborigines even in Florida, and between the accounts of the Spanish chroniclers and the later narratives of Virginia traders, and the South Carolina colonists after 1670 we are able to get a fair idea of the position of the principal Southeastern peoples at that date. Meantime the French penetrated into the Ohio Valley and as far south on the Mississippi as the mouth of the Arkansas by 1673, and to the ocean by 1682, and they founded Louisiana in 1699. La Salle’s Texas colony, established in 1685, however unfortunate for himself and the other participants in the venture, gives a more than fair view of the Indians of that great territory, soon supplemented by the reports of those who accompanied the later Spanish expeditions. Moreover, this data may be checked in some measure by the much earlier reports of Cabeza de Vaca bearing on the years 1528 to 1536 and the chroniclers of Moscoso’s invasion of east Texas in 1542. Moving still farther west, we find that New Mexico had been occupied by Spaniards long before the date selected, that Coronado had crossed the southern Plains, and that travelers by sea and land had visited southern California. In the meantime eastern Canada had been penetrated by two European nations from two directions–by the French along St. Lawrence River and the Great Lakes and by the English Hudson’s Bay Company through their posts on the body of water which gives them their name. Moose Factory was founded in 1671, Fort Nelson in 1682, and Fort Churchill in 1688. From these as bases explorers and traders soon worked their way far inland, and on the other hand the commandants collected considerable information from the natives them­ selves regarding the regions from whence they came.

As has been said, there was beyond a great tract of country which remained unvisited by Europeans until well into the eighteenth century, but over much of this area there is no evidence of recent tribal movements, and some movements are known sufficiently well to justify an attempt to reconstruct the earlier conditions. Thus the migration of Haida from the northern end of the Queen Charlotte Islands to Prince of Wales Island evidently occurred in recent times, not earlier than the eighteenth century, and it is clear that they replaced the Tlingit there since the names of their towns in the invaded country are all derived from Tlingit. Whether the movement of the Tsimshian to the coast of British Columbia and the, probably contemporary, removal of a part of the Tlingit northward, happened before or after 1650 we shall never know, but it seems to have taken place long be­fore the Haida emigration just mentioned. It was formerly believed that mass migrations of impressive character took place in the Columbia River Valley about the beginning of the nineteenth century. This idea was perhaps set in motion by George Gibbs (1877) in speaking of the migrations of Klikitat Indians, and was suggested in some particulars by Mooney (1928) but elaborated by James Teit (1928) and adopted and amplified by Berreman (1937). This involved the assumption that before that time both banks of Columbia River from The Dalles to the mouth of Snake River were in possession of Salishan tribes, that south of them lay the Cayuse and Molala, and south of them again the ancestors of all of the Shahaptian peoples except the Nez Percés; and that about the beginning of the nineteenth century the Shoshoneans of the interior moved northward, pushing the Shahaptians ahead of them; and that these in turn, after disrupting the Cayuse and Molala, expelled the Salishans from the valley of the Columbia in the region just indicated. More recent researches by Ray, Murdock, Blyth, and Steward (1938) seem to indicate that this is entirely erroneous and that, except for a displacement of the Molala and a relatively recent expansion of Shahaptians toward the south at the expense of the Shoshoneans, the tribes and stocks seem to have occupied substantially the same areas in the earliest times of which we have any record as they did when the reservations were established. At any rate, supposition of stability in tribal location makes the work of the cartographer much simpler, and we will accept the tribal distribution shown by Ray in his paper published in 1938 as being as near the probable situation in 1650 as can now be deter­ mined. From the fact that he indicates the northern boundary of Shoshonean peoples in the eighteenth century, it is assumed that he regards the rest of his map as valid for that century.

For the position of the interior Athapascan tribes before they were attacked by the Cree, I am indebted to Dr. Diamond Jenness, formerly Chief of the Department of Anthropology of the National Museum of Canada, who was also kind enough to go over most of my Canadian section and has made many valuable suggestions and amplifications.

The scope of the work has also been expanded territorially as it progressed. Originally it was intended merely as a convenient guide to the tribes of the several states of the American Union and Alaska, demand for such a work being considerable. But since the original linguistic map of the Bureau had included the Dominion of Canada and Greenland, it was later determined to make this of the same extent. And finally, owing to the representations of a leading anthropologist, it was amplified to take in México, Middle America, and the West Indies.

The method of treatment for Canada and Greenland has been practically identical with that for the United States, but it was thought best to represent on the map not merely the tribes but the band divisions of the larger northern tribes, such as the Chippewa, Cree, Algonkin, Montagnais, and several of the Athapascan groups, including the Kutchin and Khotana of the far Northwest and Alaska. Many of these band names are English and wholly modern, but it is highly probable that some of them correspond to more ancient divisions and, since they have found a place in literature, the identification of their locations will be convenient. For the placing of those in the Northeast I am particularly indebted to the late anthropologists Dr. Frank G. Speck, of the University of Pennsylvania, and Dr. John M. Cooper, of the Catholic University of America.

Objection has been made to entering the names of Eskimo tribes or bands on the map, since almost all refer simply to “people living at such-and-such a place,” most of them had little permanence, and there was an enormous number of them, the ones I have mentioned being merely a selection. On the other hand, it may be urged that some groups, notably those in Alaska, had considerable continuity, that most of them probably owed their existence to certain natural food supplies which would tend to reproduce other tribes at the same spots even though these were broken up, and that finally most of the tribes here entered have obtained a place in Eskimo literature and it is convenient to know where they lived even though they may have been no more important than other tribes not mentioned. Besides, if this were not done, the map would have little more value, so far as the Eskimo country is concerned, than the linguistic map. In the text I have indicated the relative lack of importance of the Eskimo tribes by treating all under the one head “Eskimo,” and their names, like the band names of the northern Indians just mentioned, are in different type. The West Greenland names are, of course, quite modern but are thought to represent the principal bands of an earlier date.

As already stated, that portion of the map south of the territory of the United States is based on the map of México and Central America published by Dr. Thomas and myself (1911), on the work of Lehmann (1920) mentioned above, but particularly on the papers of Mason (1940) and Johnson (1940). Although European influence in this region goes back to the early part of the sixteenth century, relatively little tribal displacement had taken place by 1650. On the West Indies, however, it was very different, and, if we were to note only the tribes extant there in 1650, little could be inserted. However, it has seemed best to submit to the anachronism here by giving the tribes in occupancy when Spaniards first came among them at the end of the fifteenth century and beginning of the sixteenth. In this part of the map I have followed Lehmann except in Jamaica and Haiti, but I have omitted several of his Jamaica names which seem to be merely those of towns. The tribal distribution in Haiti is the result of my studies of Peter Martyr’s “De Orbo Novo,” and I have increased the five “provinces” given by Las Casas (1875-76) because it seems to me that Marien in the northwest and Maguana in the center should have independent status. Probably the caciquedoms here and in the other islands were in a constant state of flux.

In treating the linguistic stocks, considerable compromise has been found necessary. Since the publication of Powell’s map (1891) the investigations of various students have rendered certain changes necessary, but other proposed changes have not been accepted by all students, and some are violently opposed.

The connection between Shahaptian, Waiilatpuan, and Lutuamian, first suggested by Hewitt (1897) and recently confirmed by Jacobs (1937), has made it necessary to put these three groups of languages into one stock which is here called Shapwailutan, a name made up of the first three syllables of the original stock names and in that form suggested by Hewitt many years ago. The connection of Natchez with the Muskhogean family, originally proposed by Brinton and confirmed by me, has been recognized in the present classification. I have also placed the former Tonikan, Chitimachan, and Attacapan stocks under the stock name Tunican in accordance with the results of my own researches though the inclusion of the first mentioned is not entirely beyond question. Dr. J. P. Harrington’s studies (1910) have made the relationship between Kiowan and the Tanoan tongues so evident that they have been placed in one family and given the name Kiowa-Tanoan. There no longer seems to be any excuse for keeping the old Shoshonean, Piman, and Nahuatlan stocks apart, and I have followed Buschmann (1859) and Brinton (1891) in uniting them as Uto-Aztecan. Kiowa-Tanoan is probably related to this but the fact has still to be demonstrated.

In California we are confronted by some puzzling questions as to relationships, which have been made the basis of violent differences of opinion. Some of our ethnologists have been very skeptical regarding the Algonquian connection of Yurok and Wiyot but I let it stand as on Kroeber’s Handbook (1925) pending exact determination. On the other hand, the validity of the so-called Penutian stock seems to be recognized by all of those who have had the best opportunities to study the languages composing it and is admitted here. The relationship between some of the languages of the other great stock created by

Dixon and Kroeber (1919), the Hokan, is also allowed by other students. A doubt still remains whether all of the languages classified under this head, even in the original and most conservative usage of the term, should go with it. Or rather, it seems doubtful whether our information is sufficient to justify the erection of this stock over against the Penutian. Mr. J. P. Harrington (personal information) is of the opinion that the distinction between Hokan and Penutian is artificial and that the languages of both groups and of various others not as yet brought together are probably related. But since the name Hokan has received literary recognition, it seems best to continue it provision­ ally for the forms of speech first placed in that category. Kroeber’s confirmation of Brinton’s suggestion regarding the Serian and Tequistlatecan stocks has served to add them to the Hokan family through Yuman, and Sapir proposed extension to Subtiaba and Coahuilteco. I am favorably disposed toward very considerable ex­ tensions of the present family boundaries but feel that more unanimity of opinion is desirable before including the more radical suggestions in a general work of this kind. Personally, I am convinced that a very large part of the vocabulary and structure of the Siouan and Muskhogean languages has had a common origin and believe that it will ultimately be found best to consider them as branches of one stock, but adequate proof has not yet been presented. The Tunican stock also shares certain well-marked structural peculiarities with Muskhogean while having connections also with the ancient Texas stocks, but the meaning of this has yet to be determined. It is plain that the structural parallelism between Athapascan and Tlingit is not accidental, and some striking similarities extend to Haida. Whether the somewhat similar parallelism between Salishan, Chimakuan, and Wakashan means genetic relationship is another problem, but the answers to these are not as yet sufficiently assured to incorporate any changes from the older classification in this work. It is evident that a future map devoted to the distribution of languages in North America must give something more than stocks or supposed stocks. It must show the degree of relationship between languages as well in­ side as outside of stock boundaries.

No doubt the positions assigned to certain tribes in the present map will surprise many ethnologists. This will be particularly true of the placing of some of those of the Plains like the Arapaho, Kiowa, Kiowa Apache, and Arikara. In fact, some of these locations are extremely speculative but they are governed by the necessity of harmonizing them with the locations of other tribes at the time selected as standard, 1650. In the case of certain tribes removed from their original seats before 1650, or whose locations were learned only at a considerably later time, the date of known occupancy is indicated in parentheses.

The present work was well under way before it was learned that something similar was being undertaken by Professor Kroeber, and Kroeber’s work has since appeared (1939) as “Cultural and Natural Areas of Native North America.” This magnificent publication will undoubtedly continue to occupy a place all by itself for a long time but it is evidently intended mainly for the university student, though its usefulness will by no means be confined to such students, and in other particulars the purposes of that study were quite distinct from those which the present writer has entertained.

“It aims,” says Prof. Kroeber, “first, to review the environmental relations of the native cultures of North America. Its second purpose is to examine the historic relations of the culture areas, or geographical units of cultures.” My own compilation has no such ambitious purposes. It is merely intended to inform the general reader what Indian tribes occupied the territory of his State and to add enough data to indicate the place they occupied among the tribal groups of the continent and the part they played in the early period of our history and the history of the States immediately to the north and south of us. It attempts to be rather a gazetteer of present knowledge than a guide to the attainment of more knowledge.

The preparation of this manuscript extended over several years and some new material was added indeed until my retirement from active membership on the staff of the Bureau of American Ethnology in 1944. It is admittedly defective in the use of material published during the years since that date.

In the synonymy only those forms have been given which differ so much from the popular designation of the tribe as to make identification difficult.

Although I have usually leaned very largely on Mooney’s population figures (1928) in my over-all estimates, my own for the Southeastern tribes, as shown by those on map 3 of Bulletin 137 (Swanton, 1946), would generally be considerably smaller.

The work has been done from the point of view of the United States, and therefore the Chippewa have been treated under Minnesota, the Huron under Ohio, and the Assiniboin under Montana, although their centers were rather north of the International Boundary.

On the maps the boundary lines between modern political nations and states are indicated by long dashes; those between linguistic stocks or major divisions of that type by short dashes and divisions between smaller tribal or group bodies by dots. 1

  1. This has not been consistently carried through on the maps.– J. R. S.[]

Language, Map,

Swanton, John R. The Indian Tribes of North America. Bureau of American Ethnology, Bulletin 145. Washington DC: US Government Printing Office. 1953.

Search Military Records - Fold3

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Pin It on Pinterest

Scroll to Top