Volume 49 1940 > Volume 49, No. 193 > The physical anthropology of the Maori-Moriori, by H. L. Shapiro, p 1-16
PHYSICAL ANTHROPOLOGY OF THE MAORI-MORIORI.
- ii Page is blank- 1
THE PHYSICAL ANTHROPOLOGY OF THE MAORI-MORIORI
WHEN the Maori and their Chatham Island neighbours, the Moriori, first appeared on the expanding European horizon, the problem of fitting them into the frame of known mankind immediately suggested itself. To discover the origin of the Maori and the Moriori and to unravel their relationships with other Oceanic stocks not only furnished appropriate subjects for the speculative interest aroused by the discovery of these Polynesian outposts, but also provided a goal for the classificatory quest which seeks to understand by systematization. In the beginning the discussion of these problems was based mainly on impressionistic data derived from the accounts of travellers, but later much of it emanated from more specialized studies. From both these sources there developed a general body of opinion which embraced a wide diversity both of view and of method, and this composite tradition, in turn, has tended to guide and influence research in the restricted and specialized aspects of the general problem. There has come about, therefore, a confusion between cultural and biological origin, and between the methods suitable for the solution of these distinct questions.
It is not my purpose here to weigh and to summarize the varied data which have contributed to the elucidation of Maori-Moriori classification, although that of necessity is my main theme. Instead I propose to survey only the results of research in physical anthropology upon this problem. I - 2 make this distinction not because evidence from linguistic or ethnographic sources has no bearing upon the subject, but because it is necessary first to clarify the problem and to solve it in consistent terms before bringing in the corroborative testimony derived from other methods designed for data of a different nature. I wish, therefore, at the outset to make it clear that the present summary is concerned with the physical origin and relationships of the Maori-Moriori and with physical evidence which has been put forth. And since a survey of the physical anthropological literature devoted to these groups reveals little or no concern with any other problem than that of origin and classification, these will furnish the only subject of discussion.
Although interest in physical anthropological studies of the Maori and Moriori dates from the early part of the 19th century and somewhat later, such distinguished names as Zuckerkandl, de Quatrefages, Turner, Weisbach, and Davis, are associated with published data in this field, it was not until 1894 that the first thorough-going study of Maori craniology appeared. The series of crania published at that date were studied by Dr. John H. Scott, and consisted of eighty-three individuals. Since then no work comparable in richness of material has been issued, although significant additions have come from von Luschan (1907), Mollison (1908) and recently Wagner (1937). Although I have not attempted to make an exhaustive catalogue of every skull published, a rough estimate of the principal work discloses more than 250 crania described.
Similarly the Moriori were known only by a few specimens which had wandered into the anthropological cabinets of Europe until Scott published his study based on fifty crania. Since 1894 additions to the published Moriori crania have come from Duckworth (1900) Poll (1903) and Thompson (1915-17).
These represent the principal modern sources for the raw material on Maori and Moriori craniology. In addition a number of studies have appeared which have employed these data in part or in whole for the purpose of the respective authors, the most significant for our purpose being Volz. Before Scott the craniological studies published - 3 were invariably based on totally inadequate series, and their conclusions were arrived at by methods not entirely clear when they were not obviously insecure. Consequently in the ensuing discussion the major emphasis will be laid on Scott and his successors.
Although it can not be said that the Maori and Moriori are exceptionally well represented by craniological publications, nevertheless it would appear that enough has been published to permit some degree of certainty in defining their respective types and in determining their affiliations. But in spite of this accumulation of data, only one conclusion has received universal acceptance. All authors agree that both the Maori and Moriori are basically Polynesian in origin, but there the unanimity ceases. Scott in his classic work regarded both Maori and Moriori as closely related with enough overlapping of craniological traits to render it difficult, if not impossible, in a number of cases to distinguish one from the other. Both these groups were, in his opinion, derived from a mixture of Polynesian elements with an antecedent Melanesian strain. The relative proportions of these intermingling strains varied, according to Scott, in the various districts of New Zealand, so that it is now possible to discern greater Melanesian influence in some tribes than in others. In general the Melanesian component is more apparent in the North than the South Island.
This view of Maori and Moriori relationship with other Oceanic groups has supporters not only among physical anthropologists but also among cultural anthropologists. A close study of Scott's monograph, however, reveals an extremely tenuous basis for his conclusions. His comparative data are far from impressive, and his statistical procedures lack conviction. One cannot help being impressed by the probability that Scott absorbed a strong bias from the environment of current theory concerning Maori-Moriori origins.
The next considerable work to appear on the physical relationships of the Maori-Moriori came from the pen of Wilhelm Volz in 1895. Although Volz's original data consisted of Easter island skulls, he made an extensive survey of previouslcy-published Polynesian crania including - 4 a composite of New Zealand crania. Volz's thesis that the Australian represented a stock once widely distributed in the Pacific was supported by his analysis of Maori crania. He professed to be able to isolate among these latter skulls individuals of Australian type, as he defined that type by a combination of cranial indices. Thus this indicial dissection of Maori crania led him to add a third component to the Maori melange. In his view the Maori represented a mixture of Polynesian, Melanesian, and Australian.
In 1900 Duckworth published a small series of Moriori crania together with a brief analysis to the effect that the Moriori originated from the “two great Pacific stocks,” presumably Polynesian and Melanesian. He appeared to approve Scott's belief in the close resemblance between Maori and Moriori in spite of certain craniological distinctions between them. At the same time Shrubsall commented on the difficulty of unequivocal identification of Chatham island skulls as Moriori, since the Maori were known to have invaded the islands in the 1830-s. Inexact reference to the Moriori of transplanted Maori or mixed Moriori might thus conceivably confuse the craniological characters to be properly assigned to the pure Moriori.
Three years later another study of the Moriori was published by Heinrich Poll. This author recognized three distinct types of crania among the Moriori: 1, mesohypsicephal, 2, mesoorthocephal, and 3, dolichoorthocephal. All of these were found among the Maori as well. Poll, therefore, agreed with Scott in linking the two groups craniologically, and he repeated Scott's observation on the difficulty of identifying individual skulls as either Maori or Moriori. Despite this similarity, however, certain differences did exist between the two groups. The Maori contained a fourth type, the dolichohysicephalic, not found among the Moriori. This proved to be a vital difference. In Poll's scheme the various skull types, which he isolated by indices, are compared with certain Oceanic stocks and identified with them, but the basis for the order of their respective invasions of New Zealand and the Chatham islands was not made clear. The original type, now associated with the mesoorthocephalic and dolicho-orthocephalic skulls, was Australian (especially Tasmanian) - 5 in origin. The dolichohysicephal, found in New Zealand, but absent among the Moriori, corresponded to the Melanesian type. The mesohpsicephals present among both groups was introduced last by the invading Polynesians. Consequently the Maori and Moriori share Australoid (Tasmanoid) and Polynesian components, but differ by the presence of Melanesian blood in the Maori complex and its absence among the Moriori.
Rather similar in certain aspects to Poll's interpretation was Th. Mollison's analysis of Maori-Moriori relationships with the major Pacific groups. In this view the Maori as well as the Moriori represent mixed groups. The Maori, on the one hand, are mainly Polynesian with the admixture of what Mollison calls an Australoid-Melanesian element. The Moriori, on the other hand, are by inference also of part Polynesian origin, but are regarded as even more similar to the Australian type than was the case for the Maori. Mollison, like many other physical anthropologists, accepts the tradition of a dark pre-Polynesian population in New Zealand, and uses it as evidence of the existence of an Australoid-Melanesian element. The mixture of this aboriginal population with the invading Polynesians varied according to locality, thus producing the present tribal differentiation among the Maori.
Up until the publication of Miss Thompson's paper on Moriori crania in 1915-17 the dominant view of the Maori-Moriori pictures these groups as a Polynesian stock mixed with one or more alien elements, Melanesian or Australian, or both. This hypothesis seemed to fit certain cultural data suggestive of a Melanesian influence, and had more or less replaced an earlier conception of the Maori-Moriori as a purely Polynesian group. Miss Thompson's biometric study, however, definitely eliminated the Australian from any part in creating the Moriori type, and since her comparisons led to the conclusion that the Moriori were closer to the Maori than any other group known to her, this absence of Australian mixture may by inference be taken to be characteristic of the Maori as well. Miss Thompson likewise concluded that the then current view of Melanesian admixture in the Maori-Moriori complex received no confirmation, - 6 although this opinion was based only upon indirect evidence. The Moriori were not compared directly with Melanesian groups, but with African negro series which presumably Miss Thompson regarded as equivalent to Melanesian. This, however, is an assumption which requires more proof than she presents. Having thus denied both the Australian and Melanesian from any role in the Maori-Moriori complex, Thompson proceeded by statistical comparison to align the Moriori with the Long Barrow crania of England, although the affiliation was recognized as only suggestive and based on inadequate samples. This connection of the antipodes, however, was somewhat loosened by the much closer tie-up between Moriori and such marginal Pacific populations as the Ainu and especially the Fuegians. The statistical resemblance of Moriori and Fuegians was surpassed by the resemblance between Maori and Fuegians. Employing Miss Thompson's data and adopting her method I discovered that the Maori are even closer to the Ainu than to the Fuegians.
We are presented, therefore, in this work with a quite startling departure from all previous speculation concerning the relationships of the Maori-Moriori. Although Thompson nowhere offered any comparative data, I assume that she considered her Moriori, as well as the Maori, to be basically Polynesian. Beyond that she found their affiliation not with any of the neighbouring Australian or Melanesian people but rather with fringe groups of the Pacific, especially with the Fuegians, and concluded that it is difficult to avoid the conviction that a link exists between them.
Twenty years elapsed between Miss Thompson's study and the next craniological investigation of the Maori-Moriori, which is also the latest. In 1937, K. Wagner published an extensive survey of the crania of the entire Oceanic area and included in it a series of Maori skulls hitherto unpublished. Supplementing this original series with comparative groups critically examined for similarity of technique, Wagner by the use of the Pearsonian coefficient of racial likeness (CRL) attempted to align all the series which he takes into consideration. The method is purely statistical. I shall summarize here only his conclusions referring to the Maori-Moriori.- 7
Comparing the Maori and Moriori with other reliable Polynesian groups, Wagner concludes that the Maori form together with the Marquesas and the Society Islanders a rather closely knit unit, within which slight local differences occur to differentiate them from each other. The Maori, furthermore, reveal a clear cut deviation from the Hawaiians.
To test the traditional belief in Melanesian and Australian links with the Maori, Wagner calculated the CRL-s between the Maori and various Melanesian, Australian, and Tasmanian groups. The Australian and Tasmanians show according to this method no statistical grounds for assuming any racial connection with the Maori. Similarly the Melanesian groups with one exception can hardly be said to resemble the Maori. The exception is the group composed of Loyalty Islanders who exhibit some degree of affiliation with the Maori, although their connection is much closer with the Easter islanders than with any other Polynesian group. Morant in an appendix to von Bonin's study of the Easter islanders had previously indicated this linkage. It should be borne in mind, however, that the Loyalty group is not especially characteristic of Melanesians, and their relationship with certain Polynesians may as well be explained on the basis of Polynesian admixture with the Loyalty islanders as on the assumption of a Melanesian addition to the Polynesian groups.
In spite of possessing certain cranial characters suggestive of Australian and Melanesian origin such as the fossa praenasalis, lophocephaly, hypsistenocephaly, the receding massive brow, and long, narrow palate, Wagner concludes that the various Polynesian stocks represent the result of independent development, thereby eliminating the necessity of assuming Australian and Melanesian miscegenation.
It is also interesting to note that Wagner obtained a low CRL, indicative of craniological similarity, between the Maori and Fuegians as well as between the Maori and the Ainu. This linkage, however, is not developed to the extent to which Miss Thomson carried a similar conclusion.
Up to this point the discussion has been confined to craniological studies. The living, however, also provide a - 8 significant field of investigation, although by contrast with craniological activity the anthropometry of the Maori and the Moriori has been sorely neglected. In fact, with the exception of stature and cephalic dimensions cited by Deniker, and the few subjects examined by Weisbach, the only anthropometric inquiry of which I am aware is Dr. Peter H. Buck's examination of Maori troops returning to New Zealand in 1919. His series of full-blooded Maori consisted of 424 subjects. Unfortunately the Moriori became extinct before anthropology could be employed to record their physical characteristics, and the only records of their appearance in life are preserved in purely descriptive and impressionistic terms.
Dr. Buck subjected his material to a Dixonian analysis by plotting the combination of cephalic and nasal indices. By omitting the intermediates on the ground that they represented “blends” rather than pure types, Buck obtained four extreme types out of a possible nine combinations. These were dolichocephalic-leptorhine, dolichocephalic-platyrhine, brachycephalic-leptorhine, and brachycephalic-platyrhine. He then equated them with Dixon's types in the following fashion: the dolichocephalic-leptorhine with the Caucasian, the dolichocephalic-platyrhine with the Melanesian, the brachycephalic-leptorhine with the Alpines, the brachycephalic-platyrine with the Palae-Alpine. The last type, however, does not occur as a pure type but is deduced from the distribution of the “blends.” The basis of this kind of type analysis employed by Dixon is of very dubious validity. It assumes without biological evidence that all intermediate forms represent mixtures and that the statistical segregation of the combinations of various indices corresponds to primary types. Genetically this is not necessarily true. Many traits are well known to behave independently and to recombine in various ways, yielding a greater range of types than originally entered into the cross. Moreover, from our knowledge of the variation-pattern of human traits, it is obvious that even within a pure type arbitrary sub-groupings may be plotted by combining the independent variations of specific characters. It is, therefore, extremely doubtful that the Maori racial components may be determined by such an unbiological reasoning.- 9
Buck also sought to align his types as defined by Dixon's method with the basic Polynesian elements which Sullivan isolated. The dolichocephalic-leptorhine type (Dixon's Caucasian) is taken to agree with Sullivan's Polynesian; the brachycephalic-platyrhine (Dixon's Palae-Alpine) with Sullivan's Indonesian. In addition Sullivan also recognized a Melanesian strain among Polynesians which Buck supposes to be equivalent to his dolichocephalic-platyrhine type.
It would, therefore, appear that Buck considers the Maori to be essentially a branch of the Polynesian stock in which four distinct types are to be discerned. The manner in which these four types co-mingled to form the Maori population was wisely left in abeyance.
The diversity of opinion disclosed by this survey requires, it seems to me, an explanation which may be discussed under two heads: the nature of the evidence and the nature of the methodology. The obstacle and difficulties embraced under these rubrics are unfortunately not confined to the investigation of the Maori alone; and although a detailed analysis of them would not be appropriate here even if space permitted, some mention is necessary in order to clarify the position in which the physical anthropology of the Maori-Moriori finds itself.
In the earliest studies, which have received but slight attention in the above discussion, the inadequacy of the material is so obvious that it is unnecessary to stress the difficulty of drawing conclusions from it, even though some brave spirits were undeterred from doing so. Perhaps this very paucity lent a clarity and simplicity to the problem which more abundant data has dissipated. But even though the Maori-Moriori crania on records have now reached numerical adequacy, each investigation is still sorely handicapped by the absence of complete and reliable evidence for a number of crucial comparative groups. Often where data has already been published they can not be utilized because of differences of technique. Polynesia aside from the Maori-Moriori is but poorly represented by reliable series suitable for comparative purposes. Melanesia, much more complex racially, is also relatively ill-known. In addition to these tribulations arising from the inadequacy of the material and - 10 the obsolescence of earlier data, much confusion has been injected into the discussion of the problem by the lack of definition of the regional types pertinent to the inquiry. It was, for example, once thought that Melanesia could be characterized as a whole by a prevailing type. In practice its isolation was extremely unsatisfactory and its distinction was based on only a few traits. Frequently, as was the case with Scott, a single Melanesian group such as the Fijian was taken as representative of the whole are. It has now become increasingly clear with the accumulation of data that “Melanesian” is no longer adequate as a descriptive racial term since an enormous variety of types are embraced within it. Some Melanesian groups, the Fijian possibly among them, are even coming to be regarded as impregnated with Polynesian blood. Consequently it is easy to see that uncritical comparisons may result in a game of tail-chasing. Moreover, the importance of establishing the nature of the variations within each stock has not been properly recognized. For example, the dolichocephaly of the Maori has on a number of occasions been referred to Melanesia on the grounds that it varied from the standards of the Samoans and Tongans. These latter were taken as typical for Polynesia, principally, I suppose, because data from other Polynesian groups were not at hand. As a matter of fact dolichocephaly is quite common in Polynesia. Lophocephaly similarly has been attributed to Australian influence, although it may be observed in Marquesan skulls. These traits and others like them may well have entered the Polynesian complex through Melanesian admixture or they may be of even older origin. But in any event it is highly significant to know whether they are characteristic of the Maori-Moriori alone of Polynesians or are common to other Polynesian groups as well. Their presence in New Zealand and the Chatham islands may not necessarily be the consequence of direct migration of people from Australia and Melanesia but may have been introduced by the Maori and Moriori themselves from Polynesian sources. Before attributions of characteristics found in the Maori-Moriori are made to non-Polynesian populations and elaborate migrations are hypothesized to account for them, it would seem to be the wiser course to ascertain the distribution of these traits in the rest of Polynesia.- 11
Curiously enough I have been unable to discover that the bearing of chronology has ever been seriously considered or even mentioned in these craniological investigations, and yet it may well have an extremely important part to play. The tendency has been to look upon Polynesia as two-dimensional and to omit its past; mainly, I venture to surmise, because stratification has not been a significant feature in Polynesian archaeology. Nevertheless Polynesia has a past and much has occurred in it. Ignorance of this has, indeed, led to certain misconceptions which have long been embalmed in the literature. Until the recent anthropometric studies carried out by the Bishop Museum were published, Tahitians were described as dolichocephalic, whereas in fact the recent Tahitians are strongly brachycephalic. Similar mis-statements have come into print for other archipelagoes, the reason being that cranial material usually was collected from caves and other ancient burial sites, and these were taken to represent the recent populations. In my own experience, I have found in the Society, the Hawaiian, and Tuamotuan islands a marked change in the head-form of the recent population. This alteration is not the result of admixture with Europeans or Orientals, neither of whom have as brachycephalic a cranium as the modern inhabitants of these Polynesian groups. It may possibly be the effect of selection and environment, but it seems more likely to be the consequence of population replacement through migration and invasion. Unfortunately it is difficult to date the skeletal material from most Polynesian islands, but nevertheless the changes which have occurred on some of them are chronological phenomena and demand consideration in comparative studies.
The questions of methodology are, on the other hand, much more difficult of resolution. They involve questions of the utmost significance for all physical anthropological problems of classification. That they have not been solved is all too apparent in the literature as a whole, leaving aside the problem of the Maori-Moriori. To my mind the safest rule is the criterion of resemblance. Close and intimate resemblance in a large number of traits implies in all systems a relationship, and the degrees of relationship may be roughly arranged according to the degree of resemblance. It is - 12 obvious that difficulties even in this simple approach will be encountered, but at least it offers an avenue of attack to which everyone will agree. The arbitrary systems, however, of analysis for type are too often purely statistical and unbiological. Such statistical figments are moved about in an arbitrary manner to account for variations and to establish relationships with equivalent figments similarly arrived at in other areas. The schemes devised by Dixon and Volz are typical examples of this form of procedure. Without knowledge of the genetic consequences of miscegenation it seems peculiarly hazardous to attempt to isolate from a complex population its contributory types.
It thus becomes apparent that the problem of Maori-Moriori affiliations has become unnecessarily complicated by inadequacies of data, by lacunae in comparative material, by incompleteness of definitions, and by variations in methodology. It is, of course, impossible to predict that order will emerge from chaos by the application of correctives suggested by the above deficiences, but it is difficut to see how secure progress can be achieved in their absence.
Before concluding this essay, it may not be altogether inappropriate to refer briefly to my own work on the Polynesian problem even at the risk of adding still further to the confusion which I have already deplored at some length. In a recent work on the Cook Islanders I drew the conclusion from anthropometric data on living and therefore contemporaneous Polynesian populations that the Polynesians despite certain common traits also revealed a local differentiation which could be correlated with geographical distribution. 1 On the basis of the available data I felt justified in distinguishing three geographical variants of the Polynesian type. These were intended to be tentative and possibly to be expanded in number, with additional information. First, the central Polynesians typified especially by the Society islanders, but also characteristic for the Hawaiian Islands, the western Tuamotuans and the Austral islanders. Second, the western Polynesians best represented by the Samoans and Tongans. Third, the marginal Polynesian as found - 13 among the Maori. With this third Polynesian type I associated the Mangaians in the Cook group, since they closely approximated the Maori. Less striking than this similarity, but nevertheless suggestive of a former relationship, appeared the resemblance between the Maori and the Manihiki-Rakahanga natives. Evidence of a close affiliation between Maori and Marquesans also emerge from the comparison of their respective anthropometric traits. Since publishing these opinions, I had the opportunity of collecting records of the Mangarevans, and I am inclined to group them together with the Maori and their associated islanders as belonging to this marginal group. The Easter islanders require a special discussion and I shall perforce omit them from present consideration. We have, therefore, a marginal continuity connecting the Marquesas, Mangareva, and New Zealand with remnants in central Polynesia. Such a distribution conforms to an hypothesis, which supposes Central Polynesia (Society islands) to have been a centre of migration. 2 At least one and probably several waves of population movement spread from this central area in which the migratory populations were replaced by the existing type. According to this view branches of the stock which moved to New Zealand also drifted to the Marquesas and southeast to Mangareva, and probably left the traces which now exist on Mangaia, Manihiki, and Rakahanga.
I am for these reasons inclined to agree with Wagner's conclusion that the Maori are essentially Polynesian with their closest relationship with the Society islands and the Marquesas. It is necessary to recall that the Society islanders to which Wagner refers are not the present-day inhabitants, but a population of crania most of which undoubtedly date from prehistoric times and represent according to my belief the earlier population which gave rise to the Maori.
The frequently-cited Melanesian component in the Maori population I questioned. If a Melanesian population had mixed with the Maoris, it seems remarkable to me that so - 14 little trace remains of such traits as do generally characterize the Melanesians. Buck found no more frizzly hair among the Maori than may be discovered throughout Polynesia, yet frizzly hair even if it were a recessive should be more common here than elsewhere in Polynesia, or at least as common as it is in Tonga where relatively recent admixture with Melanesia has occurred. Similarly prognathism and heavily-pigmented skin should be excessive among the Maori on the hypothesis of Melanesian miscegenation, whereas in fact they are no more frequent among the Maori than in Polynesia generally. This does not, of course, dismiss the existence of Melanesian elements in the Maori or the Polynesians as a whole. Such elements may well be distributed throughout the entire Polynesian population. But it does call into doubt the thesis that the invading Maori met and mingled with a pre-existing Melanesian population in New Zealand.
If succeeding investigations sustain this view of Maori relationship, the problem of their ultimate origin then becomes bound up with the general problem of Polynesian origins. At present, very little direct physical anthropological evidence on this subject exists. The discussion which has ranged around this problem is almost entirely speculative, and it would require considerable space to do justice to its ramifications.
Finally, I can not close without a last mention of the Moriori who have apparently fallen by the wayside in this summary. The reason is apparent from our earlier comments on the craniological studies, and until the gaps in our knowledge are repaired it is wiser to postpone judgment.
- 16 Page is blank
1 See discussion on modification of physical characteristics due to environment in H. L. Shapiro, Migration and Environment, 1939.
2 Dr. Buck suggests the same, the eight migration arms of the wheke of the Pacific radiating out from Tahiti. See his Vikings of the Sunrise, 1938—Ed.