CHAPTER XIX
CORRELATION OF CLASSIFICATIONS
We have now given a descriptive account of the facts constituting our subject, and with this preliminary task ended, may turn to a few more serious problems. We have succeeded in classifying the minor social units, according to culture data, language, and somatic characters. Also, it has been possible to group geographically the known archæological artifacts resulting in a distinct culture classification. The general impression prevails that all of these classifications are independent of each other. The facts in the case are that each system developed in its own way and largely at the hands of specialists. The result has, on the whole, been detrimental to anthropology, because the tendency has been to diverge into uncoördinated sciences, as linguistics, archæology, ethnology, and physical anthropology. Such differentiation developed partisans for each classification, whose controversies have obscured the problem rather than otherwise. Thus, the linguist asserts that no one can do anything until he learns native languages, because they are the keys to the whole subject; the archæologist insists that his is the main road to travel; the ethnologist belittles the claims of both; while the physical anthropologist looks down upon all. Fortunately, this condition is passing and the coming generation of anthropologists is facing the synthetic problems upon which the future of our science depends.
Such broader synthetic work must, in the nature of the case, begin with the correlation of the four great groups of data we have so far discussed independently. Each has developed a classification based upon geographical distribution and in this direction lies our first task. As may be anticipated, any such attempt to correlate these four great systems of classification will meet with opposition, for there is a deep-rooted belief that there are no such correlations. But let us ignore this proposition for the present, and examine the case on its merits.Let us begin first with the historic and prehistoric culture classifications. Our experience with Old World archæology arouses two expectations. In classical problems, we anticipate finding a long and continuous occupation of definite areas where we find the successive remains of what is, in its fundamentals, a single expanding culture. As all of these areas have well-established historical periods, we take, as a matter of course, the close correlation, or superposition, of archæological and historical culture localizations. On the other hand, when we deal with the archaeology of cultures that do not readily connect themselves with historical data, as the Paleolithic cultures of western Europe, and even some of the Neolithic, we are accustomed to find a series of cultures resting one above the other, but between which the connections are broken. It has sometimes been proposed that this difference is, in the main, one of interpretation, but again it is defended as a real difference due to the relative antiquities of the two groups. Thus, it is considered that cultures arising early enough to run their courses before the dawn of history, have time to vanish completely and leave room for the establishment of something entirely new and independent. Without dwelling upon the respective merits of these two somewhat opposing views, we may turn to the objective correlation of the archæological and cultural classifications for the New World.
In the first place, the reader should fix in his mind the nature of these distribution areas. Our analyses show that each is a more or less illy-defined area in which there are many trait differences, but that these show a gradation outward from a center, or nucleus. While, by comparative study, one arrives at the generalization of a type for the various social groups making up such a center, the result is, nevertheless, strongly grounded in empirical methods, and is not the work or interpretation of a single individual. We have seen that in this respect the historic and archæological areas are essentially similar, indicating that the data of both apply to analogous, if not identical, cultures. We should note, also, that the difference between the two classifications is one of time. One deals with observations upon living natives, the other with what is found on the ground. Since the former begins with 1492 and thus arbitrarily cross-sections the cultural career of the natives, we may expect a large part of the readily accessible archæological data to refer to the historic cultures. On the other hand, we have just grounds for assuming that the prehistoric period was of much greater duration than the historic. One of our problems, therefore, will be to discover if the archæological data can be analyzed so as to reveal earlier independent cultures in any of these areas, if such there were. Unfortunately, this task must await more searching field-work, as our discussion of chronology has suggested; so all that can be hoped for now is the mere formulation of the problem.
The most direct approach to this will be a comparative study of the respective distribution maps. For convenience, we have superimposed the archaeological and culture area maps (Fig. 101), the historic areas in Arabic, the prehistoric, or archæological areas, in Roman enumeration. At first glance, the respective boundaries may seem bewildering, but closer inspection shows many areas to have common centers. Reading from north to south, we have the following coincident pairs: XI–5, X–4, IX–2, VIII–3, VII–9, VI–1, XVIII–15, XXI–12, XXIII–14, XXIV–13. Thus, ten of the fourteen historic culture areas coincide with prehistoric archæological centers. Of the remaining, XII–6, may well be added to the above, since it is rather a graphic factor that gives the impression of divergence.
Turning now to Area 10, we find it embracing archæological centers for XIV, XV, and XVI. Area XIII lies between or on the margins of 9 and 10, but need not be considered here, for reasons previously stated (p. 260). Area XVI is the old Maya region and is, therefore, somewhat earlier than the type culture of historic Area 10. In like manner, XV is the center of Zapotecan culture, which also seems to be older than the type of Area 10, or XIV.
Likewise, for Area 11 we have three archæological centers, XVII, XIX, and XX. While we cannot yet state the chronological relations of these, yet, from the artifacts alone, we see 
Fig. 101. Superposition of Culture Areas: The Historic Tribes in Arabic; the Prehistoric Tribes in Roman Notation
Perhaps the most complicated is the superposition in eastern North America. Thus, in Area 8 we find two archæological centers, and in 7, three, if not four. In Area III, the home of the Iroquois, stratified deposits have been reported showing Iroquois remains over and above what are regarded as Algonkin. The latter have affiliations to Area I and V, and since the ancestral home of the Iroquois seems to have been farther south, we assume that the prehistoric Area III is due to this intrusion. In the discussion of culture Area 7, we found the Iroquois to be intermediate to 8 and hence, not typical. That their archæology seems more sharply divergent than their historic culture, may be due to the leveling effect of contact with the surrounding tribes. In Areas I and V we have no very satisfactory chronological determinations, so that further discussion is useless.
Looking back over this comparison, we see, first of all, that the greater disparities between the two classifications fall in the regions of higher culture. In the case of 10 and 11, it is suggested that the chronological factor is the chief disturbing element. For 7 and 8, we are less clear, but some students attribute the divergence to shiftings in the population, as the data for the Iroquois suggest.[1] To generalize, we may say that on the whole, the tendency is for the historic cultures to coincide with those of the prehistoric classification, but that in the region of intense culture and the eastern maize province, the archæological areas are more numerous than the historic classifications would lead us to anticipate. A natural inference would be that this is explicable as an expression of relative age, those areas in which the people had lived through only one culture cycle, being the ones in which the two maps most nearly coincide. This, however, does not seem consistent with other data, and may be disregarded at present.
LINGUISTICS AND CULTURE
The reader is, no doubt, quite familiar with the idea of no correlations between culture and linguistic type, which is regarded as a kind of truism. But when we look carefully into the case, it is not clear that every kind of correlation is absent.[2] Our attention, so far, has been fixed upon the conventional stock grouping, but it is now becoming clear that this is not the only possible classification. Under the appropriate heading, we have noted that in California, positive similarities between a large number of stocks have been worked out and new groupings proposed. Further, we have shown how the great variety of languages spoken in the California culture area fell into groups strictly parallel to the culture grouping, and noted suggestive data from other areas pointing in the same direction. In all this, there appears a definite tendency for language to correlate in certain ways with the culture grouping. Yet, this correlation may be an expression of tribal contact rather than genetic relationship in speech, for it certainly does not follow that similarities in culture parallel identities of linguistic stock.
Perhaps the reader should be reminded that we are discussing unity in stock and not identity of speech. Though the English and German languages are of the same stock, they are far from being mutually intelligible, and this well illustrates what we meet with in native American stocks. It is, in fact, a safe assumption that real identity of speech in the New World will be accompanied by political and cultural unity. On the other hand, we are dealing with a distinction of another sort, for a stock relation is an expression of common origin. Even the most cursory comparison of culture and linguistic maps will show that a single stock may spread into several culture areas, as, for example, the Athapascan, which is found in Areas 1, 2, 3, 6, and 9. This simply means that a language can travel independently of culture, or, at least, outlives it. The point we made in the preceding is that the representatives of different stocks gathered into a culture area often show intra-stock similarities by which also they can be grouped and segregated. These similarities suggest that languages have a grouping similar to, and largely coincident with, the culture grouping. This may, or may not, be the result of genetic relations, but rather of long social contact. Our point is that the languages of a culture area, even when regarded as of independent stocks, still show a grouping that tends to be coincident with that for cultural characters. It is a reasonable expectation that a distribution of phonetic types alone would show a similar correspondence. The peculiar point is, however, that the unity is in the stock affiliation and not in practical speech, for the linguistic differentiation within a stock is often very great. Neither is there political unity, and it may be doubted if the latter can long exist without the other. The suggestion is, therefore, that the similarities of languages within a geographical, or culture area, are due to the expansion of the early parent stocks within their habitats, and to the long association made possible thereby.
SOMATIC CORRELATIONS
GENERAL RELATIONS OF CULTURE, LINGUISTICS,
AND SOMATOLOGY
If now we take a comprehensive view of anthropology as a whole, we see no reason for revising the general assumption that no necessary relation exists between the known types of culture, linguistics, and somatology. Neither are certain types of linguistics associated with certain types of culture, nor certain types of somatology accompanied by special forms of culture or language. The particular combination of these which we encounter are, in the main, historical facts. Yet, we have seen that, while this is essentially true, there are ways in which these classes of data do correlate. This correlation is seen in many of our culture areas. One of the best known examples, that of the Eskimo, has been cited in a previous discussion. Here the language, culture, and somatology are all distinctly different from those in other areas. Again, in California the languages, though of many stocks, show a curious tendency to possess common distinctive characters, while the somatic type is claimed as uniform throughout.[3] We can, therefore, safely summarize our discussion by stating that each distinct culture area tends to have distinctive characters in language and somatology. However, the reversal of this formula does not hold, for people speaking languages of the same stock do not show a tendency to common culture characters unless they occupy a single geographical area. An analogous negation holds for somatology. It seems, then, that culture is one of the primary factors in this association, and that, due to causes we have not yet perceived, both languages and somatologies are differentiated after culture's own pattern.
THE MIGRATION FACTOR
Our next task will be to seek out clues to the identity of these contributory causes. Early in this discussion, we saw how closely the prehistoric areas agreed with the historic, suggesting that a type of culture was in some vague manner firmly fixed in a locality. In some respects, such a conclusion is disconcerting, because the traditions of our subject call for a veritable and constant flux of migrations. While our previous discussions have made it apparent that migration is exceptional rather than universal, still, the logical necessities of the case require some movements of population. One can scarcely conceive of the peopling of the New World except by the expansion and spread of its population gradually from one area to another. It may be, however, that this is not the important point here. Thus, when we consider the best known or most probable cases of migration, they all seem to have one common character, since they are circumscribed movements in a single area. For example, the Eskimo, whose first appearance in the New World must have been in Alaska, spread only along the Arctic coast belt to its ultimate limits. Yet, since they are the sole possessors of this territory, they offer a far less suggestive example than can be found elsewhere. We now have satisfactory data for the northern movement of the Iroquois, but if we superimpose a linguistic map and one showing topography, we see clearly that whichever way the tribes of Iroquoian stock moved, they kept close to the Atlantic Highlands. On the other hand, the Algonkin expanded in lands of less elevation, 500 to 2,000 feet, but chiefly in a wooded lake and portage region. In like manner, we may follow out the Siouan, Athapascan, Salishan, and Muskhogean, Shoshonean, Mayan, Arawakan, Cariban, and other stocks. It cannot be an accident that all the Muskhogean peoples lived at an elevation less than 500, that the Algonkin were, with few exceptions, between 500 and 2,000, that the Athapascan are chiefly inland between 2,000 and 5,000, while the Shoshoni-Nahuatl stock occupies land above 5,000. If, then, we turn from linguistic to culture characters, we find similar agreements. The Southeastern culture (Area 8, Fig. 67), is below 500 feet, the Eastern Woodland area (7) is, in the main, from 500 to 2,000, the Nahua area (10) above 5,000, as was also the home of Inca culture (12).
It is also suggestive to take a topographical map as our point of departure, and note the correlations between elevation and the variations in cultures. Thus, the Algonkin of the Atlantic Coast plain from Maine southward are below 500 feet, or on a level with the Muskhogean group, and it is just here that we find certain northern traces of Southeastern culture. In the Mississippi Valley these same lowlands reach up to the Ohio and the Missouri, and here also, we find the margin of Southeastern traits. The two types of culture which we find in the Bison area (1), the western and the eastern, line up along the north and south divide of 2,000 feet. Further, it is distinctly among the eastern tribes that we find Eastern Woodland traits, the elevation of both being below 2,000 feet.
These few illustrations must suffice, but the reader can follow out others by referring to suitable maps. We are not contending for a direct correlation between elevation and culture, for these numerals are but convenient geographical indices to climatic, faunal, and floral areas. As boundaries, they are just as arbitrarily chosen as those for culture areas, but for that very reason should be strictly comparable. The fact, then, that they do so correspond cannot be dismissed as a logical error. We are thus brought to the conclusion that the phenomena of our subject manifest a strong tendency to expand to the limits of the geographical area in which they arise, and no farther. Language and blood seem to spill over the edges far more readily than culture, from which we must infer that their dispersion is a by-product of migration, but that these migratory groups seem unable to resist complete cultural assimilation.
From this point of view, it is conceivable that the Muskhogean and Algonkian stocks, for instance, could have exchanged habitats without changing the cultures localized within the two areas, provided the shifting was by successive small units; or that all of the Shoshonean peoples could have become Pueblos and the Keresan and other stocks have scattered out on the plateaus to the north, while the culture values of the two areas remained relatively the same. That such extreme transpositions ever occurred is, of course, improbable, but their possibility is demonstrated in the Plains and particularly in the Pueblo area itself.
Yet, the real problem in the case remains unsolved. What became of the culture brought into an area by the first settlers? That the new environment was harsh enough to stamp it out promptly is beyond belief. The idea that all initial cultures were uniformly simple at the first dispersion over the New World is equally unsatisfactory. No doubt part of our trouble arises from the fact that we are best acquainted with the centers of lower cultures. When we have fuller knowledge of Mexico and the Andean region, it may be expected that different underlying cultures will come to light. It is impossible to guess what insight will thus be gained into the early history of the Americas, though it is a fair assumption that even the earlier cultures of these higher centers will possess individuality comparable to other areas.
On the other hand, we have everywhere encountered evidences of fundamental unity of origin, in both culture and somatic types. The basic elements of culture in the area of most intense development have their fading parallels as we go outward, and the somatic grouping shows, also, unmistakable evidences of expansion from a single ancestral stock. This may account for the close correspondences we have observed between archæological and historic classifications, for, in contrast to Europe and western Asia, there were no sudden eruptions of new races until the white occupation of the New World, whereas in the former, we find evidences of many such successive invasions.[4] It is fair to assume that, had the road from Asia to America been broad, direct, and open, we should find here a complex archæological condition quite comparable to that of the Old World, for then wave after wave of new peoples and cultures would have poured into America as they did into Europe.
THE INFLUENCE OF ENVIRONMENT
We are now prepared to take a closer view of the controlling factors in the formation of these geographical types of man and his works. It is natural to suspect the environment to be one of these, since the fauna and flora of a locality are certain to leave their stamps upon many traits of culture. From the very first, we saw how the location of food areas laid down the general lines of culture grouping. One striking characteristic of such an area is the tendency to specialize in some one or two foods. Reference to the initial chapter of this work will make this point clear. For example, we have, centering in California (3), the use of the acorn; in the Plains (1), the bison; in parts of both continents, maize; and in the Amazon country (14), manioc. At all these culture centers we find more or less elaborate processes of preparation involving technical knowledge, for example, the making of acorn flour and bread, the roasting of camas, etc. These processes tend to spread throughout the area of supply. Thus, the acorn industry extends well up into Oregon far beyond the California center; the roasting of camas (2) to the mouth of the Columbia, and also to the Blackfoot of the Plains, etc. Again, we note certain specializations of manufacture: California (3), baskets; North Pacific Coast (4), boxes and plank work; the Plains (1), rawhide work (parfleche, bags, etc.); Mackenzie (6), birchbark (canoes, vessels, etc.); Plateau (2), sagebrush weaving; Southwest (9), textiles and pottery; Southeast (8), cane and fiber weaving; the Eastern Woodlands (7), knot bowls and bass fiber weaving, etc. Types of shelter present similar distributions, and so do many other traits. All of these traits are seen to reach out far beyond the borders of the respective culture centers, and such extensions can, in the main, be correlated with faunal and floral distributions. Yet, not even all of the more material traits can be considered dependent upon the fauna and flora, for example, pottery. Also, art and ceremonies are no less distinctive and, as we have seen, are also localized in these same centers. In this case, the influence of the environment could be but remote. In any case, we find that the people in a culture area have chosen but a few of the possibilities and specialized in them, leaving many other resources untouched. It is, of course, plain that, if the directions of this specialization varied, many different kinds of culture could successively occupy the same geographical area.
Apparently, the chief explanation of this phenomenon lies in man himself. A group of people having once worked out processes like the use of acorns, maize, manioc, etc., establish social habits that resist change. Then the successful adjustment of one tribe to a given locality will be utilized by neighbors to the extension of the type, and to the inhibition of new inventions, or adjustments. Therefore, the origin of a culture center seems due to ethnic factors more than to geographical ones. The location of these centers is largely a matter of historic accident, but once located and the adjustments made, the stability of the environment doubtless tends to hold each particular type of culture to its initial locality, even in the face of many changes in blood and language.
As to actual movements of tribes from one center to the other, we have no good historical examples. Our discussion of migrations has shown that the tendency was to move about in the same geographical area, yet these wanderings must have sometimes led to other culture areas. The most probable case of this kind is to be found in the history of the Cheyenne Indians. According to Mooney,[5] they left Minnesota, where they had an intermediate Plains culture and joined the typical tribes at the center, where, in the course of a century, they became typical in culture. On linguistic grounds, we assume that the Blackfoot and Arapaho did the same at a much earlier date. In California and the Southwest we see Athapascan-speaking remnants who are most likely immigrants from the North, yet this does not show itself in their culture. The fact seems to be that these migrating units were not strong enough to prevent their people from following the lead of neighbors wherever they happened to tarry. It is clear, however, that such migrations must have been but occasional and exceptional, for, unless the distinctive tribes remained in force at the culture center, the type would have been destroyed. Hence, we must assume very great stability of these centers and a fair degree of stability for the New World as a whole.
Now reverting to the correlations of language, culture, and somatic type, we are prepared to see that while the environment does not produce the culture, it furnishes the medium in which it grows, and that when once rooted in a geographical area, culture tends to hold fast. The somatic type, while not closely correlated with environment, yet shows a tendency to conform to it. This would be consistent with our conclusion that the population of a center was relatively stable, so that an occasional immigrant group could be gradually leveled by intermarriage. But the immigrant group need not change its language so long as it maintains its social independence, nor does the environment appear to put any pressure upon its members to so change it. What uniformity of somatic type we then have is merely the accidental result of sexual contact, while the lack of linguistic uniformity is due to the former history of the group. It follows then that about all that survives in an immigrant group is its speech, for both its culture and blood gradually disappear in the new environment. But blood being a biological character, in contrast to culture, it is subject to certain very definite laws of inheritance and the particular forms the mixture will take are likely to be various. Culture on the other hand, can be entirely displaced even within a single generation.
Finally, with a few generalizations we may summarize this long and tedious, but we trust not unprofitable, survey of aboriginal man and his varied characters. It appears a fair assumption that so long as the main sustaining habit-complexes of life remain the same in an area, there will be little change in material culture. This may be in part an explanation for the lack of close correspondence between the historic cultures and archæology in the several parts of the great maize areas in contrast to identity elsewhere. The bison, salmon, wild acorn, and guanaco must have been in their respective habitats for a long, long time, and a culture once developed around them could be displaced only by a radical change, such as the introduction of agriculture or pastoral arts. Now the regions where maize was found in use at the opening of the historic period are just those in which archæology shows the most disparity. It seems then, that the environment as a static factor conserves the types of culture and because of this weighting of one of our three great groups of characters, breaks their unity, so that the same language and likewise the same blood may be found in association with different cultures according to the laws of historic accident.