(1) One of the most untrustworthy of evidences of race is language. For languages will not mix except in their vocabulary; and there it is so easy for an alien word to take on a familiar guise and a familiar word to take on an alien guise that we may be deceived by likeness or analogy. A people strips off its linguistic apparel as lightly as its peculiar dress and adopts another in its place. Races have been perpetually crossing since the beginning of mankind. But languages only displace each other or make compromise by rubbing off each other's grammatical or phonological peculiarities. And derivations are the most delusive of all lines of historical or prehistoric investigation.

(2) Yet there is something to be got at times from the general affinities and characteristics of a language. And Maori has not escaped the efforts of the linguistic investigator, as an evidence of race. For a time it was assumed to be an offshoot of Malay. And the result was the usual fallacy. The Polynesians were declared to be, like the Malays, a mongoloid race. And the misleading name "Malayo-Polynesian" still lingers as descriptive of race even in the latest histories and descriptions of the Pacific, and induces the most grievous misconceptions as to the affinities of the South Sea Islanders. The term is as misleading when applied to language. For Malay as the tongue of a mongoloid people page 82is assumed to be agglutinative, or, in other words, to express the grammatical relations by symbolic elements that are, unlike inflections, detachable from the stems. And Maori is constantly classified as an agglutinative and even as a Turanian language.

(3) But there is nothing agglutinative about it. And the remains of inflections apparent in it indicate that one of the languages that it displaced was inflectional, if not both. Now the only inflectional languages are either Aryan or Semitic, both types belonging only to Caucasian peoples. In one relic, the internal plural, it seems to have affinity with the Semitic. But in most of the remains of inflection it points to Aryan parentage. It seems, therefore, not unlikely that one or more Aryan tongues went to the making of it.

Maori Grammar, like English, is the Result of the
Collision of two Languages

(4) The true classification of Maori is with the modern European languages, and especially with the most modern of all languages, English. It has reduced inflection to a minimum, and expresses the grammatical relationship of words chiefly by their order in the sentence, and the use of auxiliaries and particles. It has the power of interchanging the significant parts of speech, the noun, adjective, verb, and adverb, as occasion requires or dictates. The number of nouns is indicated by the articles or other definitives, the case by prepositions, the gender by the addition of the word for male or female, the degree of adjectives by a separate word, and the mood and tense of verbs by a particle.

(5) It is, in short, as uninflectional or isolating a language as English, and it is agreed by most philologists now that page 83this type is by no means primitive: is, in fact, the end of a long process instead of the beginning. The farther we go back in the history of a language the more inflectional we find it; and it is the most barbarous instead of the most cultivated that has most inflections as a rule. The exception of Greek and Latin at first misled philology. And if we take English, the modern language in which the process of inflectional decay has gone furthest, we see the true cause of it plainly marked. It is the clash of grammars. It is the intercourse of two peoples in the same country, each of whom fails to master the inflections of the other's grammar. Anglo-Saxon was more inflected than any modern European tongue, and, though the process of inflectional decay had set in through the inroads and conquests of the Danes, it was not completed till the Normans settled as the aristocracy and rulers of the country. These immigrants, being in a minority, had to adopt the language of the Saxons, but in doing so they failed to understand or master the inflections, and at last did without them, whilst they introduced vast numbers of words from French into the popular speech. But traces of the old inflections have been left, especially in the pronoun and the verb, parts of speech that have so much of the relational in them that they cannot easily do without something formal in their changes. And in the noun the plural is still indicated by an external change, whilst in a few it shows internal change.

(6) Now nothing could better describe the grammar of Maori than this. The pronouns and the verbs are the parts of speech that show the remains of inflection, whilst a few of the nouns indicate their plural by internal change; "tangata," a man, has its first vowel short in the singular and long in the plural, so "wahine," a woman, and half a dozen others. But the process of rubbing out the inflections has gone much further in Maori than in English. It has page 84retained no external method of indicating the number or gender, except by the addition of another word. The internal inflection for number in those few nouns is probably the result of the same phonological law as still makes "man" in English take as plural "men," the law of Umlautthat is, the original inflection has modified the vowel of the root and then disappeared. In the Maori pronoun the disintegration of grammatical form is perhaps not quite so advanced, for we have a dual in the personal pronouns, though not in the others, whilst in the demonstrative the singular "tenei," "this," is formed from the plural "enei," "these," by prefixed inflection. In the Maori verb the old inflections have left fewer traces than in the English verb; in fact there is only one, that for voice. The passive is formed from the active by an affix, which changes to suit the termination; the original form of this was probably "ia," the third personal pronoun singular, though it often prefixes a consonant, to suit the sound, or drops the "i." This shows the formation of the passive to be similar to that in the older Aryan languages; it was what is called in Greek grammar a middle voice; in other words, a reflexive voice, formed by the addition of a reflexive pronoun to the active voice.

(7) Now, the only explanation for this striking linguistic development is that which we know fits the modern evolution of English. Two languages far more highly inflected than modern Maori have come together in the same region and mutually destroyed the formal grammar, leaving only a few traces of the inflections.

It was in Indonesia that this Evolution was accomplished

(8) But it must not be forgotten that this flexionless grammar belongs not merely to Maori, but to the dialects of all the groups of Polynesia. Not only so, but the very page 85relics of grammatical inflection are the same in all. And this reveals that the grammar, as it is, was brought with the conquerors, and being as simple as it well could be it was acquired by the conquered. Nay, there is evidence that it had already been stripped of its grammatical forms before it passed from Indonesia. For this flexionless character, along with some of the relics of inflection, belongs not merely to the Malay tongue, but to the Malagasy, spoken so far away as the east coast of Africa. The mistake made by early observers was that the Polynesian dialects, as well as the Malagasy, were all derived from Malay as spoken by the navigating race that founded the empire in the Malay Peninsula from Sumatra in the twelfth century. An even greater mistake was made in classifying the Polynesians and the Malays as the same in race because of the similarity of their languages. It was assumed by both that the Malay as a great navigator had traded into the Pacific and colonised the islands.

(9) The consideration of one fact alone would have prevented all such misconceptions. The iron age began in the Malay Archipelago about the beginning of our era; and not one scrap of iron or other metal was found in Polynesia before it was brought by the European voyagers. The historical Malay of Sumatra and the peninsula could, therefore, never have traded with the Pacific Islands; nor could his half-Mongol half-Caucasian ancestor, the Malayan, have had any intercourse with Polynesian after the beginning of our era.

(10) The true explanation is that the friction between two inflected tongues that resulted in the flexionless grammar must have taken place in the Indonesian archipelago before the Mongols appear there, and before the Malagasy dialect reached Madagascar, or the Polynesian dialect reached the South Sea Islands. In the great archipelago that stretches page 86from Sumatra to Celebes there must have been a vigorous and comparatively dense Caucasian population in prehistoric times either driving the negroid aborigines into the interior or absorbing them. Two Caucasian, and most probably two Aryan, tongues came into collision in Indonesia through the amalgamation of two Caucasian races; this and this alone will explain the evolution of so flexionless and so modern a language with the embedded relics of inflection in it.

(11) That this occurred some centuries before our era can be proved. For the Malagasy dialect contains no Hindoo or Sanskritic words like the Malay and the Polynesian. Hence it must have hived off before there was much migration from India into Indonesia; and that is supposed to have begun at least three centuries before our era. The Arabic words in the tongue of Madagascar have been shown to be very ancient, and are probably due to the voyages of the Himyarites of South Arabia, the Sheba of Solomon, along the east coast of Africa in search of gold and other precious products of that continent.

(12) The colossal mortuary dolmens and circles of Madagascar indicate that it was reached by the megalithic migrants, who, after coming to the ocean in the east along the southern Asiatic route, took to far-voyaging. They point back to Sumatra and Java and their colossal-stone monuments; whilst the comparatively fair complexion of many of the Malagasies points to a Caucasic or originally European infiltration into the African island. But there must have been a Malayan or half-Mongoloid, half-Caucasic immigration as well. For the kinship of the Malays and the Malagasies is apparent in their features, their small stature, and their general character. We may infer, therefore, that the intrusion of the Mongols into Indonesia occurred at least three or four centuries before our era. But it is equally clear that they had not overflowed the whole archipelago and modified its page 87Caucasian inhabitants. Many parts must have been un-Mongolised. For the Polynesians, who undoubtedly rested in Indonesia, carried no Mongol features or characteristics with them into their ultimate home. They remained tall, wavy-haired, bearded, and comparatively long-headed, such as no Mongol or Malay ever was; nor have they any of the recklessness or the sullen fitfulness of the Malay.

But the Phonology was modified in Polynesia, and
thus shows an aboriginal influence

(13) There was one element of the Polynesian dialects that, though probably modified in Indonesia, was still more modified in the Pacific Islands. This was the phonology, the range of sounds that the language included. The grammar it brought could not well be made simpler, and hence remained as it was brought, in spite of the new peoples that had to adopt it. But we see from Malay and Malagasy that it had in Indonesia a much larger scope in sounds. Both these languages have fifty per cent more letters to use in forming its words. Polynesian must have rejected this surplusage when it had to adapt itself to the speech organs of the conquered in the Pacific. The result is one of the simplest, most primitive phonologies on the face of the earth. Not that all the islands have the same. Every group has rejected one or two sounds that some of the others have retained, a clear proof that there were different peoples in them, with organs of speech long moulded under different climatic conditions. The vowels are all retained and are less subject to change, though in individual words they are often substituted the one for the other. They are but few, and all the dialects have rejected diphthongs, some of which Malay has retained. But the consonants, few as they are, differ greatly when the same word appears in different dialects. The dialects page 88of the islands that are nearer the equator have liberally rejected the guttural k, and that this was not solely or directly due to climate is shown by the fact that the Moriori dialect of the Chatham Islands does the same, a fact that seems to show that the Polynesian ancestors of the Morioris came from islands nearer the equator than those of the North Island Maoris. The difficult initial ng is also rejected by most of those dialects and by that of the South Island of New Zealand, the former preferring instead of it n, the latter k.

(14) It is generally the household or domestic pronunciation that determines the direction of change in sounds; the women of the conquered are taken in as wives or servants, and mould the speech of the children in their early or plastic stage, and their climatic environment is their most faithful ally in this. The new vocabulary of a language that is a compromise is moulded by the men on the battlefield and in the marketplace; its phonology and its grammar are moulded on the hearth. Now, in Maori there are only fifteen sounds, the five vowels, three liquids, m, n, r (in Polynesian 1); one guttural, k; one dental, t; and one labial, p; three breathings, h, w, and wh; and a nasal initial, ng. This is one of the simplest of phonologies; it is evidently one that has abandoned all the half-tones and variations, all the difficult sounds that may have been introduced by the more advanced culture of any conquering immigrants. If one of the ethnological strata in the Polynesians was Aryan, as the grammar seems to indicate, then the triple variety of the sonants (k, g, gh; p, b, ph; t, d, th) that belongs to most Aryan tongues has disappeared, and a set of three, one for each of the main points of contact of the organs of speech (namely, k, p, t) has taken their place. Doubtless, the aboriginals found the distinction brought by the immigrants between hard, soft, and aspirate in each order of sonant beyond their powers of page 89discrimination by ear and hence beyond their powers of speech.

This Loss of Sounds has by Assimilation brought
Incongruous Meanings under the Same Words

(15) A striking contrast to this primitive phonological scheme is the fulness of the vocabulary, which runs into tens of thousands of words, where that of most primitive and even half-cultured peoples runs only into thousands. And yet there are no abstract words for qualities of things or persons, such as are found in all languages that have reached the stage of introspection or self-reflection. The only approach to these consists of the personified names of the cosmology and the mythology, like Kore, "The Void"; Ao, "The Upper World"; Po, "The Lower World," and Whaitua, "Space," which indicate that the Polynesian mind had entered on the road that leads to abstraction and philosophy. But it went no further during the centuries of the Pacific, and instead rather tended to fall back on the particular and the individual, the mark of an uncultured people.

(16) The Maori substitute for the abstract is the use of the metaphorical or analogical. No barbarous or semi-cultured race ever indulged to such an extent in figures of speech. If figurative language constitutes poetry, an opinion not uncommon amongst critics as well as the unthinking, then there never was a people so poetical; they are a race of poets by nature; in their myths and incantations, songs and oratory, they revel in the figurative till it is difficult for the European mind to follow them. Like Hudibras, the Maori orator no sooner opes his mouth than "out there flies a trope."

(17) The result is that in the Maori dictionary there is an astounding number of synonyms for everything that the page 90people are familiar with, and there is as astonishing a development of meanings for every second or third word. But in the classification and study of these there is a great source of ambiguity and difficulty. A large proportion of the words have applications that it is impossible for the European mind to find the connection or analogy between. Take as an example the word "rere"; one set of meanings is easily put together; to run like water, fly, sail, leap, move to and fro in speechifying, rise or set like the stars; but the others are difficult to connect or classify; to be born, to be rejected, to hang, suddenly, a waterfall, a swarm, an exclamation demanding attention; in order to find the relationship here we have to strain analogy to the breaking point.

(18) The source of this phenomenon is the coalescence of different forms by assimilation; we see in English that when two words have grown like in sound they tend to become one in form, especially if they come from different linguistic origins. This occurred most in the centuries immediately after the Norman Conquest, when the native Saxons and the Norman aristocracy, with their followers, had not fully amalgamated; each moulded the words of the other into the forms that were familiar to him. The same occurs in the history of the Maori language. And what increased the confusion here was the reduction of the elaborate phonology of the immigrants to the simple phonology of the aboriginals, and especially the reduction of the gutturals to k, the dentals to t, and the labials to p. We can imagine the vast museum of meanings under single forms in English, if either the Saxons or the Normans had been unable to differentiate p from b, and b from f, or d from t, and t from th. Yet this is what has happened in Maori. The result is an amazing collection of meanings under the same sound-form in hundreds, if not thousands, of cases. And the complication is increased by the inordinate love of figurative speech in the Maori race, page 91The difficulty of the philologist in disentangling the roots is enhanced tenfold by these two tendencies,

(19) One thing, however, comes out with great clearness; it is that the Maori language is the descendant of several languages and not of one. In short, there are as manifest signs of stratification in the Polynesian phonology as in the Polynesian people and manners and customs. The character of these seems to indicate that there have been at least two immigrations into the Pacific islands.

(20) The grammar points to an Aryan-speaking race as the source of one of the elements, if not more, for it shows traces of inflection, both internal and external, both prefixed and affixed, as all Aryan languages have done, and it has developed with ease on the same lines as modern European tongues. It has thrown off almost all the formal elements that speech can dispense with, and it has attained an absence of inflection and a simplicity of system that few tongues in the world but English have reached.

The Roots of Polynesian Words are Frequently
Identical with Indo-European

(21) The vocabulary gives the same indications. When we strip cognate words of all the formative or changeable elements, we reach what philologists have called roots, fragments of from two to four letters with a fundamental sense, from each of which a series of words may be conceived to have grown. In Maori these roots are found to consist of one or two consonants and a vowel, like Aryan roots, and unlike the Semitic, which consist of consonants as the framework with vowels as the changing and formative elements. We may conclude, then, that no Semitic tongue has entered into the body of the Polynesian dialects.

(22) That the vocabulary has been influenced by one or page 92more Aryan tongues seems to be a by no means illogical conclusion when we examine the Polynesian roots; a large proportion of them are the same or almost the same as those common to most Indo-European tongues. A few examples will make the point clearer.

(23) In Maori we have the word anene (to blow softly like the wind); and the root ane (to blow or breathe) appears in many words all through the Polynesian dialects. The same root appears in most Aryan languages. It is the base of the Greek < ἄνεμος (the wind), of the Sanskrit animi (to breathe), and it takes a development natural to the Aryan mind in the Latin animus (the mind) and anima (the soul).

(24) Another widespread root in Indo-European tongues is us (to shine or burn), appearing, for example, in Latin aurora, Greek εος (the dawn), our East, and Sanskrit ush (the morning). In Maori we have ao (the dawn, the day); and that an s has been lost is shown by the cognate Samoan aso (the day).

(25) To take another letter, the Polynesian root mu (to sound, to hum), appearing in Maori mumu (to hum), is clearly the same as the Aryan mu (to utter a slight suppressed sound), appearing in Greek μύζειν (to mutter), Latin mutire (to mumble), and musca (a fly), English midge, Latin mussare (to mutter), mugire (to roar), and murmurare (to murmur), and English mourn.

(26) When we take the Maori liquids, we are faced with the difficulty that r and l have interchanged and are still interchanging all through Polynesia, the preference in the islands being for 1, and in New Zealand for r; but it is not a great difficulty, for the same interchange has taken place in Indo-European. For example, the Aryan ra (to rest, to be delighted, to love), appearing in Sanskrit ran (to rejoice), ram (to rest), and rati (pleasure), in Greek ἔζως (love), and English rest, is not far off in origin page 93from the root la (to yearn for or desire), which appears in Sanskrit lash (to desire), Latin lascivus, English lust, Latin libido, and English love. It seems much the same as the Maori reka (pleasant), rekareka (delighted), Hawaiian lea (joy, pleasure), and lealea (to delight in), and in the other sense of calm, in Maori whakaruru (sheltered from the wind), Tahitian rurua (a lull), and Hawaiian lulu (calm).

(27) Our English lull has no connection with this root, but meant originally to sing, to sleep, coming from Scandinavian lulla, and connected with Greek λάλειν (to speak). These all point back to an Aryan root, la or ra (to sound or utter sound), which appears in Sanskrit ras (to cry loudly), and ra (to resound), Latin latrare (to bark), lamentum, and loqui (to speak), and in English roar; they point also to a modified form of ru or lu, which appears in Sanskrit ru (to sound, to bray), Latin rudere (to roar), rumor, and raucus (hoarse), and Anglo-Saxon run (a mystic letter, but previously a whisper or secret).

(28) These are widely spread in Polynesian words: Maori rangi (a song or tune), Samoan lagi (to sing), Maori rango (a fly), Samoan lago (the housefly), from the sound it emits, Maori rongo (sound, noise, report, tidings), Samoan logo (to report), Tahitian roo (fame), Maori rorohu (to buzz), roria (a Jew's harp), reo (the voice, speech), ruri (a song), ruru (the New Zealand morepork).

(29) It is in the labial, dental, and guttural roots that there is sure to be most that is elusive. For in the Polynesian dialects the three varieties of each, the hard, the soft, and the aspirate, are generally reduced to one, and that is as a rule the hard; whilst Malay has retained two labials, two dentals, and two gutturals. The Samoan, the Tongan, and the Paumotan have g, and the Marquesan k where Maori and Rarotongan have ng, and Tongan has b where all the rest have p. All of them have rejected d, which Malay has page 94retained. They have kept some trace of the guttural aspirates in h, and of the labial aspirate in f or wh, but th has disappeared from all the Polynesian and Indonesian dialects.

(30) Yet even here we find roots that are identifiable with Indo-European. If we take the most elusive of the series, the guttural, there are many Polynesian and Aryan roots similar. There is the common Aryan root ki, or kei, or kai (to lie or reside), which appears in Sanskrit ci (to lie), in Latin quies (rest), and the Greek κει̑μαι (I lie), κοίτη (a bed), κωμη (a village), in Gothic haims (a village), heima (a home), and in English home and ham, the termination of so many village names. In Maori we have kainga a village, home, or place of abode.

(31) So the Maori kohu (to cook), kohua (a boiler, a Maori oven), are evidently in root to be identified with Latin coquere and our cook and kitchen. And the Maori koi (sharp), koinga (the edge), are from a root that is to be identified with the Indo-European ka (to sharpen), which appears in Sanskrit cana, Latin cos, and Scotch hone, each meaning a whetstone, Greek κω̑νος (a cone), and Latin cuneus (a wedge).

(32) A few more may be given briefly: Maori koke (to creak), and koko (the tui), may be set beside the Sanskrit kakh (to laugh), Latin cachinnare, and English quack and cackle; Maori koko (an angle or corner), may be placed with Sanskrit kuch (to bend), and English hunch and hucklebone; Maori koa (glad), may be put with Latin gaudium (joy); Maori hari (joy), and kohara (to shine forth), with Sanskrit ghri (to shine), and English glad; Maori kore (broken), with Sanskrit car (to break up), Latin clades (disaster), and gladius (a sword), and Anglo-Saxon here (a destroying army), whence comes the English harry (to ravage); Maori koro (a noose), may be placed with Latin circus, curvus, and corona, all implying something curved.

page 95

(33) These are taken almost at random from the Maori dictionary, and indicate a close connection between a large number of the primary roots in the Polynesian tongues and those in the Indo-European. And so great has been the predominance of Indo-European words and roots over those from alien sources that after reducing all the words in Maori, beginning with k and m, to their simplest and shortest common stems, one half at least of these were found to have similar meanings to those in Indo-European languages of the same form. It is noteworthy that the roots of Polynesian words are more often to be identified with roots that appear in European tongues than with roots that appear in Sanskrit only. It looks in fact as if the ancestors of one Polynesian migration had been longer in contact with the migrants who brought Aryan speech into western and southern Europe than with those who brought it into India. And if this is confirmed by careful investigation, it will prove that the Caucasian element that came from the north-west along the Japan-Ladrone-Caroline route was not only Caucasian in race, but Aryan in speech. Whilst the coincidence of Maori ruma (an apartment), used all through the Pacific in the sense of house, with the English room, of Maori poaka (a pig), a genuine Polynesian word, with Latin porcus, of hoanga (a whetstone), with Scotch hone, and Polynesian area (an open space), with European area will mean far more than derivation from the same root.

Every Indication Points to a Pre-Malayan Caucasian
Race in Indonesia Speaking an Aryan Tongue

(34) But, whatever the meaning of these indications may be, we may be quite sure that the bulk of the Aryan roots in Polynesian came the other way, through Indonesia. For the extraordinary similarity, and often absolute coincidence, of page 96words in the Eastern Pacific dialects with words in the Indonesian, that led the first investigators to call all these tongues Malayo-Polynesian, is due probably to the influx of words from India into both regions. Sanskrit and the Hindoo dialects derived from it constitute the source of many of those common words. We know, of course, that in historical times Buddhist rulers held sway in Java and the adjacent regions, that a Brahmanic civilisation ousted them, to be ousted in its turn by the Mohammedan religion. But it is not to this migration of Indian culture into Indonesia that the influx of Aryan words into its dialects is mainly due. They came long before, in fact, in prehistoric times, by the sea route from the coasts of India, probably from Ceylon as the last stopping-place.

(35) The affinity of grammatical forms, of the words for numerals, of the phonology, and of a certain proportion of the vocabulary in the dialects spoken all over Indonesia, Melanesia, Micronesia, Polynesia, and Madagascar cannot be explained by the Malayo-Polynesian hypothesis that has held the ground for so long. For, as Mr. E. Tregear, in the introduction to his excellent "Maori Comparative Dictionary" says, "the bulk of the two vocabularies," Malay and Polynesian, is not the same in origin; and he has not been able to trace more than one percent of the words in Malagasy and Polynesian as having affinity. In Fijian, which has been greatly influenced by the Tongan and Samoan Archipelagoes, nearly a third of the words, he finds, are Polynesian.

(36) The hypothesis that would fit the circumstances would be the wide spread of a Caucasian people of Aryan speech by sea in pre-Malay times over the Malay Archipelago, and across both Melanesia and Micronesia into Polynesia. It is more than likely that this race came by sea into Indonesia from India, and not by land; for, by the latter route they would have left their traces on the languages and peoples on page 97the way. The pressure of conquest by the Sanskrit-speaking immigration from the north-west of India would drive as they advanced the sea-coast peoples out along the sea routes that they had already established for trade. And when the Aryan invaders won their way to the coast either on the north-west or the north-east of India, they would follow the tribes they had driven out across the oceans.

(37) But by this time the Mongoloids had reached Sumatra, and had begun to force the Caucasians or semi-Caucasians with their Aryanised language on to the sea to seek other homes, to the south-west in Madagascar, and to the south-east in Melanesia and Polynesia. These Mongoloids absorbed the Indonesians and the forms of their language, whilst adhering largely to their own vocabulary. And when Caucasianised and turned into Malays and sailors they made Sumatra their base, and from thence they conquered the peninsula that they had originally descended and ultimately abandoned.

(38) It is to this Indian pre-Buddhist migration by sea into Indonesia, caught on its flank by the Mongoloid incursion from the north, that the final migration into Polynesia is due, The Malays followed it in later centuries only as far as the west of New Guinea; no trace of their features or headform or the iron they used is found farther east.