Art. XIX.— The Migrations of the Polynesians according to the Evidence of their Language.
[Read before the Wellington Philosophical Society, 6th September, 1911.]
In the “Transactions of the Royal Scientific Society of Gottingen” for 1909 there appears a long paper on this subject by the late Professor Finck, of Berlin. It attempts, as its title implies, to point out some of the distinctions between the various languages of Polynesia, and by this means to indicate the lines of migration that peopled the islands in which they are spoken.
The gist of the arguments and conclusions is given in the last two pages, and is somewhat as follows: From the southern Solomons a really united people shifted to the northern fringe of Polynesia on their eastward trek. Before the expedition turned southwards to Samoa the ancestry of the present-day Ellice and Tokelau people branched off. The speech of that time possessed all that marks Polynesian as contrasted with the related Melanesian, especially the use of the old trial as plural, and the employment of separate possessive pronouns where once only a suffix was used; it was, in fact, probably the fundamental Polynesian tongue. The use of afc for “a thousand” does not contradict this, although it appears in this sense only in Fakaofa, Futuna, Samoa, Tonga, Uvea, and Niue; for the word is, as the Maori awhe shows, common to Polynesia; but it was extruded in the other dialects by mano. There was a long rest in Samoa, as is shown by the use of tokelau for “north” and tonga for “south” in a majority of the groups, words taking this sense from the direction of the Tokelau and the Tonga Groups from Samoa. After a small colony had swarmed off westwards to Futuna, the great eastward-going expedition went southwards to the Tonga Archipelago, as is shown by the use of h in all the groups to the south and east for s in Samoa and its immediate neighbours, and by the use of toko as a personal prefix to words implying number and quantity in all to the south and east for toka of Samoa, Fakaofa (the Tokelau Group), Vaitupu (Ellice Group), and Futuna. After a short rest in Tonga the expedition went off eastwards, leaving a contingent which sent branches to Niue and Uvea. In the Cook Group it made a long sojourn, and there formed the ground speech of eastern Polynesia; it changed l into r and f into h before o and u, brought the adnominal particles na and no into use beside the older a and o, and abbreviated the old possessive tou into to.
From this point various expeditions set out. One went to New Zealand and the Chatham Islands and developed h for f before other vowels than a and o; it left before the counting by pairs arose that characterizes the other eastern Polynesian dialects. A second went off south-east to Mangareva; thence a branch hived off to Easter Island, farther in the same direction, before the birth of the linguistic neologisms that unite the dialects of the Marquesas and Hawaiian Groups with that of Mangareva, the formation of adverbs by prefixing ma or mo to a noun, and the change of tokerau into tokorau. It was long before this northern expedition set out—long enough to develop these peculiarities. The Marquesas Group developed as linguistic characteristics the pronominal form toia and the further duplication of numeration by pairs in the case of rau (there equal to 400) and mano (there equal to 4,000) before sending off the Hawaiian branch. Meantime from the Cook Group another colony hived off to Tahiti, whose
dialect seems to be closely akin to that of Rarotonga, as is shown by the common use of the plural and dual prefix pu'e. From Tahiti the Paumotu and Manahiki Groups were colonized.
At the end of the article a sketch-map is given of these branching migrations. But the limitations of the linguistic method are revealed by the accompanying sketch-maps, one made by Horatio Hale in the “forties” of last century on the “Wilkes Expedition,” another by Gerland for Waitz's “Anthropologie” in the “sixties,” and a third by Weule for Helmolt's “History of the World” early this century. Hale brings the expedition first to Samoa, with offshoots to the Ellice and Tokelau Groups, then to Tonga, and thence direct to New Zealand and the Chatham Islands; from Samoa, also, one goes off to Tahiti, whence one goes to the south-east Marquesas, a second to the Tubuai Archipelago, and a third to the Cook Group. A third colonizing expedition leaves Samoa for the Cook Group, the Tubuai Archipelago, and Mangareva. Besides the branch to New Zealand, Tonga sent off one to the north-west Marquesas and on to Hawaii. Gerland, like Finck, brings his primary expedition through the Ellice and Tokelau Groups to Samoa, thence, like Hale, over Tonga to New Zealand and the Chatham Islands, whilst, as in Finck's, a Samoan offshoot goes to Futuna and one Tongan offshoot to Uvea and another to Niue. He also sends a main expedition, like Finck, over the Cook Group to the Tubuai Archipelago, and one to the Marquesas, a third to Easter Island, and a fourth to Hawaii. Weule, like Hale, brings his expedition first to Samoa; thence one colony goes direct to Hawaii and another by way of Tahiti; a third goes direct to the Cook Group, and thence to the Tubuai Archipelago and Mangareva. From the Cook Group a colony goes to New Zealand, whilst from Tahiti one goes to the Cook Group and another to the south-east Marquesas, and the north-west Marquesas are peopled from Tonga.
There is no better criticism of the linguistic method of finding lines of migration than the presentation of these differences. The fact of the matter is that these pure philologists isolate a few small phenomena that each belongs to several groups, and ignore hundreds of others in which the groups thus united disagree. One instance will be enough: Finck gives a table of the sounds of each group, and then he proceeds in his sketch to ignore some of the more striking variations. He gives ts (the English missionaries make it ch) as a variation of t in Futuna, Uvea, Tonga, and the Chatham Islands before the vowel i; all the other dialects have only t; yet he brings no migration from any one of these direct to the Chatham Islands, skipping New Zealand. So wh is given as a variation of h and f not only in New Zealand and the Chatham Islands, but in the Tokelau Group; and the same groups are united by using w for v. Yet he ignores this community of linguistic phenomena, and brings no migration from the Tokelau Group to the southern groups, or the reverse. These are quite as important as the break (') for k, on which he bases the linguistic community of the Ellice, Tokelau, Samoan, Tahitian, South Marquesan, and Tubuai Groups; or the variation of r from l, on which he bases an eastern Polynesian Group, consisting of New Zealand, Chatham Islands, Tahiti, the Paumotus, the Cook Group, Mangareva, the Tubuai Archipelago, and Easter Island.
The radical mistakes made by these philological ethnologists are the attempts to draw inferences from the language without the culture, and the assumption that there was but one colonizing expedition. The extraordinary similarity of the dialects (Finck seems to acknowledge “dialects” as the proper term, for when he says “Sprachen” he always adds, “that
is to say, ‘Dialekte’”) as contrasted with the countless variety of not merely dialects, but languages, in the Melanesian region and the Malayan region, if properly considered, might have saved them from the latter mistake. Even the few centuries which they seem to have in their minds as covering the history of the human race in Polynesia would have developed languages as distinct as, say, French and Spanish, or English and German. If we were to take into account the marvellous similarity of the Polynesian dialects not only in phonology and grammar, but in vocabulary, spread over an oceanic region as wide as Europe and Asia combined, we would not be far wrong in concluding that there have been thousands of migrations from every island to every other island; in short, a new sketch-map of the Polynesian migrations should so completely cross-hatch the central Pacific that it would look black. In other words, for centuries at least intercourse must have been almost unbroken amongst all the groups. If this means anything, it means that for a prolonged period all the Polynesians must have inhabited a large island or archipelago centrally situated, and also quarantined from other regions under a social, if not political, system that was practically a unity. The minute dialectic differences that arose must have been kept in bounds by the constant social intercourse that a single administrative system would allow—a system absolutely different from that of Melanesia or of Malaysia. The differences are no greater than those that separate the dialects of, say, Yorkshire and Somerset, or Scotland and Middlesex.
The consideration of the culture conveys the same impression; the ethnological differences are as negligible as the linguistic when placed beside the points of agreement. One can find as wide variations of culture and dialect in the purely German part of the German Empire. They seem to have arisen in the presence of each other, as well as of the predominant community of culture. In other words, they must have slowly developed during the immense period of time that certainly was taken to produce the practical identity of culture and language. This identity would have been shattered into strongly contrasted fragments had it been compelled to run the gauntlet of the limitless variety of Malaysia and Melanesia, not to speak of having to sail right in the teeth of the south-east trades, the only fairly constant wind on that route, the contrary wind being brief, fitful, and cyclonic.
There is, of course, a striking similarity between the languages of Polynesia, Melanesia, and Malaysia that makes many speak of them unitedly as the Oceanic language. But there is a phonological gulf between the Polynesian dialects on the one hand and the Malaysian and, still more, the Melanesian languages. Each of these two regions has its own range of sounds, with considerable community; but Polynesian has the peculiar and distinguishing sounds of neither—it has the simplest range of sounds that ever language had, all easily pronounceable by Aryan and, one may add, by Japanese organs of speech. It has a similar contrast in vocabulary: with any one of the Malaysian or Melanesian languages except Fijian it has never more than 20 per cent, of common words. It is the grammar that has led to their classification as one language; for none of them have practically any formal grammar—they all move in an atmosphere of particles, and there is a very considerable resemblance in the particles used. But this absence of formal grammar is the commonest characteristic of crossbred languages—i.e., languages that have resulted from the permanent or continuous settlement of a masterful people amongst a people linguistically different; the formal grammatical peculiarities of both are
gradually dropped, and particles take their place, or variations of order of words.
The distinction which Max Müller drew between languages, classifying them into isolating, agglutinative, and inflective, according as they had no formal grammar, formal grammar with forms detachable from the stems, and formal grammar with forms undetachable, is no real distinction. There are few languages that have not at least traces of all three—isolation, agglutination, and inflection—either as vanishing habits or as neologisms. It is the phonology, or range of sounds, that really distinguishes languages. This cannot change—i.e., the organs of speech cannot change, except by change of environment—i.e., by change of climate or change of educative influences in the formative period of the organs of speech. The grammar and the vocabulary are constantly changing by loss, or addition, or development. Within the same zone of climate and physical environment the sounds do not change except by change of mothers—i.e., by intermixture of races linguistically different.
But in the languages of the three regions referred to — Polynesia, Melanesia, and Malaysia—there is a considerable similarity of particles. This undoubtedly means that one language has saturated the languages of all three regions. The great variety of languages in Malaysia bars that as the region from which this language came; the still greater variety in Melanesia still more effectually bars that. There is an easy solution when we turn to Polynesia, which has only one language, though it has many dialects.
But were this in conflict with the racial and cultural phenomena of the three regions it would have to be abandoned, or considerably modified, or conditioned. It is not, however. A visit to the Solomon Islands soon convinces even the superficial, untrained observer that the fundamental race of Melanesia is negroid: the woolly, tufty hair, the thick lips, the flattened nostrils, the projecting muzzle, and the absence of calves on the lower limbs are to be seen on all sides, quite apart from the dark colour which gave the region its name. The predominance of the round head and the low stature indicates the negritoes or pigmies as the branch of the negroid race that first peopled Melanesia. But there is a considerable infusion of tall stature, straight and wavy hair, light-brown and even auburn hair, European features, and light-brown colour; especially in the eastern islands of the Solomons are the last three apparent. In the western Solomons and the Bismarck Archipelago, though the colour is close to black, the hair is often straight or wavy, and the profile is what we call Semitic, whilst tall stature is not infrequent. There can be no hesitation in homing this peculiar western Caucasianism to the west—i.e., to Malaysia or the Asiatic Continent—and in homing the light-haired Caucasianism of the eastern islands to Polynesia. In Malaysia, again, we have, as the name implies, a strong admixture of Mongoloidism with the primeval negroidism and the secondary Caucasianism. When we turn to Polynesia we find the purest racial elements —fundamental Caucasianism, with a slight admixture of negroidism.
The culture exhibits similar phenomena. Polynesia is the realm of the patriarchate; the pivot of relationship is the father. Right through Melanesia and Malaysia the matriarchate is the system; the mother is the pivot of relationship: there is therefore no history, no preservation of the records of the past, no tradition, the mother being only a private person, and having no public events in her life to hand on the memory of to posterity; the sons as well as the daughters belong to her and her kin, and do not count any relationship with the father and his relatives. The
patriarchate is at least thousands of years in advance of the matriarchate, for it makes history and tribal and political unities; the father hands on to the children, and he is the warrior and event-maker; hence, under the patriarchate, tradition accumulates into chieftainship and kingship. There is no broad realm of the patriarchate westwards from Polynesia till we reach India. That the Polynesian social system should have travelled tens of thousands of miles in frail canoes in the teeth of the trade-winds, and run the gauntlet of two matriarchal realms, has a touch of the miraculous in it or, in other words, seems contrary to the laws of nature.
It seems more in harmony with the possible, if not the probable, that whatever kinship lies between the cultures and the languages of these three regions has gone westwards out of Polynesia. And this is borne out by facts. Fiji, the nearest part of the two regions to Polynesia, has had its social system transformed from the matriarchal to the patriarchal; chiefship and tribe and tradition have arisen in the group. It is highly Polynesianized. When we get to the Solomon Islands, the nearest part to Fiji in the eye of the trade-winds, three islands have gone in parts through the same transformation—Malaita, Choiseul, and New Georgia; and their natives show a larger percentage of European features and light-brown hair than those of any others of the group; they are also most warlike, and go back furthest into the past with their genealogies and traditions. The influence of the patriarchate tapers off as we go farther west into Malaysia.
The purpose of this excursion into ethnology is to show how close to the absurd those philologists, like Finck, go who make the starting-point of Polynesian colonization the south (they should say rather the east) of the Solomons. The basis of the conjecture is a name often given to San Cristoval, the most easterly of the Solomons. Hale identifies Bulotu, the paradise and probable original home of Tongan and Samoan tradition, with Bouro, one of the most easterly islands of Malaysia. German ethnologists prefer, as a rule, to identify it with Bauro, the name referred to as applied to San Cristoval. But Bauro is only a district on the north-east coast of the island, and the natives prefer to call the island, if they have any name for the whole, Makira.
We get into the region of the miraculous when we start a patriarchal, tribal, genealogy-loving, chiefly Caucasian people from a matriarchal, kin-divisioned, short-memoried negrito island; and still nearer the miraculous when we start off, for nearly ten thousand miles of open oceanic wandering, a canoe expedition right in the teeth of the only constant winds, the trades that blow eight or nine months of the year, from an island that had only shallow shells of canoes, unfit for crossing anything but fairly narrow straits in calm weather or a favourable wind. The Polynesians were the only people in the world that learned oceanic navigation before the use of the compass. And it needs some exceptional, if not catastrophic, goad of nature to explain the exception; that we have in the subsidence, probably often slow, but probably as often sudden, of the central island zone of the Pacific that stretches south-east from the southern end of Japan across the Equator, even as far as Easter Island. This manifestly went on for hundreds of thousands of years; and any humans that got on to the islands of this zone would, time and again, have to go off the best way they could find in search of other standing-places in the great flux of waters. Nowhere else in the history of our world has such a goad been held by nature to the backs of human beings. We may be quite certain that the regions to the west would get flooded with migrations from water-logged Polynesia.