Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Ethnology of Tongareva


page break



This study of the atoll of Tongareva, commonly known as Penrhyn Island, is based on field work conducted under the auspices of the Bernice P. Bishop Museum in 1929.

Tongareva is served by two trading schooners which sail from Rarotonga. These schooners do not make their first trip until late in March or April, when the hurricane season is over. In November they sail to Tongareva to lie up for the hurricane season within the lagoon at Omoka, the only safe port in this part of the Pacific. I arrived at Tongareva in the end of June by the schooner Tiare Taporo, which left seven days later for Tahiti. Through the kindness of Judge Hugh Ayson, resident Commissioner for the Cook Islands, arrangements enabling me to leave Tongareva with the British sloop H. M. S. Veronica, which called there and left again on July 13, were made by wireless. Seventeen days were spent on Tongareva, all too short a time to do justice to the special problems of that extremely interesting island.

My thanks are due to Sir Apirana Ngata, Minister in charge of the Cook Islands in the New Zealand Government, for his Polynesian welcome conveyed to the group by letter. I am under deep obligation to Judge Ayson for assistance in arranging that my first six days of work should fit in with his official holding of the Land Court in Tongareva. Mr. S. Savage, Registrar of the Court, assisted with the Tongarevan genealogies, and Mr. H. Williams with the recording of anthropometrical measurements. To Mr. Wilson, resident government Agent, I am indebted for accommodation during my stay and, together with W. Phillip Woonton, for invaluable assistance in locating and mapping out the maraes in the various islands. To Captain Viggo of the Tiare Taporo, Mr. Wilkenson, trader, and others I am indebted for many kindnesses. Pa, the oldest inhabitant, with Ma, his wife, Tupou Isaia and others of the Tongarevan people contributed field information for this work, and by recognizing my Polynesian kinship, enabled me to adjust my Maori background to that of Tongareva. To Commander Robertson of H. M. S. Veronica my thanks are due for hospitality and transport to Raiatea, where I was able to connect with a mail steamer to page 4 Rarotonga and resume work in the Cook Islands. My thanks are due to Mr. J. A. Campbell of the Cook Islands Trading Company for promptly clearing up by correspondence some obscure points in ceremonial observances.


Tongareva, an atoll situated in latitude 9° S. and longitude 157° 10″ W., is the largest and farthest north of the lagoon islands under the Cook Islands administration of New Zealand, but it is not, geographically, one of the Cook Islands. It is composed of a ring of islands spaced along a reef about 40 miles in circuit with a contained lagoon of about 108 square miles.

The name, Tongareva, means “Floating Tonga” and was given to the atoll by the Polynesian discoverers. Gill (6, p. 11)1 gives Fararanga as an older name, which he interprets as “land.” As the Tongarevan dialect contains no f sound, Gill probably confused it with Raroranga, which is mentioned in one of the genealogies collected, “The growth of the human stock of Raroranga, Tongareva.”

Figure 1. Map of part of the central Pacific showing the position of Tongareva (Penrhyn) Island.

Figure 1. Map of part of the central Pacific showing the position of Tongareva (Penrhyn) Island.

No complete survey of the atoll has been made; figure 2 is a rough approximation. The sizes of the islands are only indicated, but their relative positions and correct native names were checked with the assistance of Tupou Isaia, who acted as the guide during various expeditions to locate the maraes. Smith (23, p. 90), from data given by Lamont (15), gives a list of 14 islands, with one small island. Here confusion has arisen because Lamont did not make it clear that some of the names he gave were of divi- page 5 sions or districts. Thus, the first four names on Smith's list are all on the one large island to the southwest; of these, Mangarongaro and Hakasusa are the two main divisions of the island. Sararak, which figures prominently in Lamont's account, cannot be located, but it was probably the district in which Opaka, the leading chief, lived. Tahiti was also a district in which Opaka lived, and it is definitely located. Similarly, Omoka and Motukohiti are not separate islands but two divisions of the large island which lies on the west. Te Puka in the southeast and Motu-unga (“Motunga”?) to the
Figure 2. Map of Tongareva showing islets, passages, land divisions, villages, and maraes. Based on a British survey of 1881. (Additions by K. P. Emory and Te Rangi Hiroa.)

Figure 2. Map of Tongareva showing islets, passages, land divisions, villages, and maraes. Based on a British survey of 1881. (Additions by K. P. Emory and Te Rangi Hiroa.)

north of Omoka and Motukohiti are separate islands. Tokerau in the north is undoubtedly correct as the name of an island which is divided into a number of districts, but Ruahara, also in the north, is part of an island of which the other division is Torea. Tautua and Motu-nono (“Motumuno”) are the two divisions of one island in the east, and Te Mata (“Tamata”) is one of the three divisions of another island on the east, the other divisions being Naue and Patanga. Hangarei (“Hangary”) is one of four small islands separated from Mangarongaro by shallow channels. The other page 6 islands are Matapurarua, Tuahua, and Manono. Etukaha, which Smith considered the correct form of Lamont's “Etuchacha,” is the small island of Atutahi in the south. Thus Smith's list of 15 islands diminishes to 8 islands and parts of two others. On the other hand, the following small islands that once were inhabited have been omitted: in the south, Vaiari and Atiati; in the east, Tepetepe, Kavea, Vaselu, and Takuha; and in the north, Niu-te-kainga. Of these seven islands, all except two have maraes. In addition, there are two large islands in the north, one containing the district of Rukutia and the other that of Nahe. The names of 33 small islets not fit for permanent occupation were given and marked on the map in their approximate positions, but there are probably others. One small island is situated well within the lagoon, and it is curious that its name of Motu-taiko should be the same as the name of the one island situated in Lake Taupo in New Zealand. The larger islands had no single name but were distinguished by the divisional names.

The islands are composed of coral and sand and nowhere rise more than 50 feet above sea level. In official records the total area is given as 40,000 acres.

Three passages through the reef admit small vessels into the lagoon. The northeast passage named Takuha is situated between the islands of Te Rae and Takuha. The northwest passage named Sekelangi lies between Te Kasi and Tokelau. The west passage between Motu-unga and Omoka is the largest and is about 40 yards wide and 21 feet deep. It admits vessels of fair size, such as H. M. S. Veronica, which can anchor within the lagoon. Trading schooners can tie up against the wharf at Omoka and those trading in the Cook Islands usually go to Omoka to lie up for the hurricane season. The lagoon is very deep but contains upgrowths of coral which have to be avoided in navigation. The natives have names for the various shallows and coral upgrowths. The reef on the lagoon side of the islands extends a varying distance from the shore and the deeper channels influence the location of landing places on the islands. The large pearl shell —responsible for the eventual growth of trade—and the small pipi shell are both abundant. The small Tridacna (pasua) abounds and forms an important food supply.

Contact with Western Culture

Lieutenant Watts of H. M. transport Lady Penrhyn was the first European to discover the Tongarevan atoll, in August, 1788, while he was on his way to China from Tahiti.

Kotzebue, the Russian navigator, visited the atoll in the Rurik on April 30, 1815, during his first voyage to the South Seas (1815–1818). He states page 7 (14, vol. 3, p. 217): “As soon as we had approached the shore, innumerable boats surrounded us, and a peaceable people offered to trade with us.” The people, however, had nothing to trade except coconuts, some utensils which they had taken by chance with them, and their weapons. “They first hesitated to barter their arms, and would not part with them, except for long nails or scarlet worsted girdles.” Kotzebue had no trouble with the islanders, who passed around the vessel “with much affability and confidence” but would not accept invitations to go on board. In addition to Kotzebue's account of his voyage, Choris (3), the artist who accompanied him, published an illustrated description of the voyage which contains an illustration of a Tongarevan canoe and two types of spears.

Endicott (5), sailing from Maui, Hawaii, in the Glide on a trading expedition to Fiji, sighted Penrhyn on November 6, 1830. He described the natives, who “came off in great numbers … perfectly savage and fierce, hallowing and shaking their spears.” According to his account the captain, in trying to persuade one of them to come on board, was slightly wounded in the neck by the spear of another native. The aggressor probably thought that the captain was trying to capture his relative. The captain immediately ordered his men to fire, and seven or eight natives were killed. Endicott (5, p. 30) sums up the position as follows: “We immediately filled our sails and stood on our course leaving the natives to bewail the visit of civilized people to their uncivilized shores.” He evidently saw no incongruity in his statement.

The Porpoise, one of the ships of the American Exploring Expedition under Wilkes, touched at Tongareva on February 15, 1841, while Wilkes was in Hawaii. The atoll was found to be 30 miles west of the position given by Captain Cash (31, vol. 4, pp. 296–298). A large number of canoes came out to the vessel and many of the people were allowed on board. A little unpleasantness occurred because of pilfering, but it blew over without any casualties. A certain amount of trade went on. The natives were found to be absolutely fair and honest in handing up their own goods for exchange after receiving the desired articles from the ship. Unlike the Glide, the Porpoise left without causing the natives to bewail the visit of a civilized people.

The brig Chatham, after trading in the Society Islands, Marquesas Islands, and the Cook Islands was wrecked on Mangarongaro in 1853, while on her voyage back to San Francisco. The whole crew was hospitably received by the inhabitants. The captain, doctor, and a few others sailed away in a boat made on the island, abandoning Lamont, the trader who chartered the Chatham, and some other members of the crew. Lamont, who subsequently got away on a whaling ship and wrote up his experiences, page 8 reached a position of great influence on the atoll. His book (15) is one of the best narratives of first-hand contact with a group of Polynesian people before they were influenced by western culture. As he described many customs that have disappeared entirely, his observations will frequently be referred to in this work.

In the year following the wreck of the Chatham, Christianity was introduced by native pastors from Rarotonga, the headquarters of the London Missionary Society in the Cook Islands. The people were collected together in four villages to facilitate the teaching of the new religion, which they readily accepted.

In 1864 Tongareva was almost depopulated by Peruvian slavers. Sterndale, quoted by Smith (23), states that in 1864 at least 1000 men, women, and children were taken to South America. The present population attributes to the influence of native pastors the chief responsibility for this exodus. The four villages were strongly desirous of building large churches. The slavers held out promise of good pay and a safe return, and pastors and people decided to go abroad to earn enough money to erect buildings worthy of the worship of God. However, the thousand people who left their island homes all died as slaves in exile, the victims of hypocritical and inhuman types of white men.

Through missionary influence the contact of Tongareva was with headquarters at Rarotonga. The London Missionary Society taught largely through native pastors and instituted a system of laws which applied to temporal as well as to spiritual matters. Many of the customs that had served their purpose in the old culture were displaced by the culture associated with the new religion. Also, the power of the chiefs was curtailed and much of it transferred to the missionaries.

As trade developed a loose British Protectorate was maintained over the Cook Islands, in spite of French attempts to link up the islands with the French Administration in the Society Islands. A British Consul was appointed at Rarotonga in 1888 and paid for by New Zealand. The ariki (high chiefs) of the Cook Islands formed a Federal Council at about the same time to administer the group. In 1900 the council of ariki petitioned the Governor of New Zealand to annex the Cook Islands as part of the British Empire and recommended that Penrhyn, Manihiki, Rakahanga, Palmerston, Pukapuka, and, if possible, Niue be included in the federation with New Zealand.

In 1901 the Government of Great Britain issued an Imperial Order in Council extending the boundaries of New Zealand to include the islands mentioned above. On June 10, 1901, a Proclamation was made at Auckland by Lord Ranfurley, then Governor of New Zealand, in the presence page 9 of H. R. H., the Duke of York, who later succeeded to the throne as King George V. Since that date Tongareva has been part of New Zealand.


Because so many of the islands were settled and because divisions of the large islands put war parties in the field it may be inferred that the atoll once supported a much larger population that it does at present. The estimate that 1000 were taken away by the Peruvian slavers in 1864 makes it likely that the pre-European population was about 2000. A contrast is revealed by the figures of the last few censuses.

Year Males Females Total
1906 420
1911 335
1916 326
1921 376
1926 201 189 390

The last census does not include five Europeans. The figures show a decrease down to 1916, after which there is a definite increase.

The Cook and Other Islands Report of the New Zealand Government for the year ending March 31, 1929, states that for the year the births were 22 (males, 17; females, 5) and the deaths 4 (males, 1; females, 3). There was thus an increase in population of 18 for the year.

The early navigators remarked on the well set up figures of the men. Kotzebue (14, p. 217) stated that they were strong and well made, and that the elderly people were corpulent and large.

Time permitted of the anthropometrical measuring of but 21 fullblooded males, details of which will be published later. A few remarks, however, are offered here.

The skin color is darker, probably due to extra exposure in fishing and diving for pearl shells and pipi pearls. The hair is black in color and straight or in low waves. Beards are not now worn but Lamont (15) remarked that they were abundant, and some were curled and with auburn ends. He also remarked on auburn hair among some of the women. The men have little or no hair upon the body. The eyes are medium-brown to dark-brown with no trace of the Mongoloid fold at the inner angle. The foreheads are well formed and the glabella is more prominent than in the people of the Cook Islands.

Early writers remarked on the tall stature of the men, but the small series averaged 170 cm., which is shorter than in most branches of Polynesian stock.

page 10

The maximum head length averaged 192 mm.; the head width, 158 mm.; the minimal frontal, 104 mm.; and the cephalic index, 82.3. The faces of many were very high and broad, the average face height being 130 mm.; bizygomatic width, 149.5 mm.; bigoniac, 116.6 mm., and the face index, 86.

The nose matches the face with the average height of 62 mm.; width, 43.2 mm.; and nasal index, 69. The high, broad faces, broad noses, and prominent glabellae make the countenance strong and massive in expression.

The women are much more slender than the men, and the expression is softer and milder. As in other branches of their race, their hands are small and beautifully formed.

Both sexes have all the hospitality of their race and are ever ready to make presents of coconuts and fish. They have so few material goods with which to express their feelings that the giving of pipi pearls, which they can secure by diving, has become their method of expressing friendship to the visitor who sojourns among them. They are quick to express their opinions and sometimes a village argument takes place with so much noise that a stranger imagines that blood will flow. Having expressed themselves, sometimes physically, the tumult subsides and no bad feeling is retained. The people are honest and outspoken, and kind, under an apparently austere demeanor.


The Tongarevan language is a distinct dialect of the Polynesian tongue. Reading and writing have been taught by native pastors, who have been trained by the London Missionary Society in Rarotonga. The Bible has been translated into the Rarotongan dialect, and the alphabet taught is that compiled for Rarotonga. This alphabet is quite inadequate for the recording of the Tongarevan speech, for not only has the h sound been omitted from Rarotongan, but it has no l, s, or w, all of which are present in Tongarevan speech. The vowels consist of the usual Polynesian five, a, e, i, o, and u. The Rarotongan consonants used consist of seven, k, r, m, n, p, v, and ng. The early missionaries who drafted the alphabets for the various Polynesian dialects had extreme difficulty in distinguishing between the sounds of l and r and of v and w. These and other doubtful sounds were settled for the usage of the Hawaiian alphabet by a committee which, as recorded by Spaulding (25, pp. 32, 33), took the votes of nine missionaries. Though the English language contained all the letters, the missionaries have standardized the Polynesian dialects by restricting them to one sound of a doubtful pair, though both sounds were evidently in use. Some of the difficulty arose because subdialectical differences occurred in different localities of the same group of islands and standardization was felt to be necessary for page 11 printing the Bible and other religious literature. Thus in Hawaii the original k sound had been dropped and was represented in speech by the glottal closure. Later, the k sound began to displace the original t sound, but the process was by no means complete. The committee settled the matter by adopting k and discarding t and thus arbitrarily completing the transition. The difficulty with the l and r and also with v and w is that probably the original sounds were intermediate.

In time the tendency was to go in one or the other direction. Thus, in New Zealand the l-r sound became definitely r and the v-w sound became w, so that the dialect became standardized in the sounds by the Maoris themselves. In Samoa, the l-r sound became definitely l and the v-w sound became v, so that the dialect also standardized itself in these two sounds. In the Cook Islands, however, and probably other areas such as Hawaii the process of the sharper definition of the uncertain sounds went in both directions. With regard to the assumed l-r sound, in some words the tongue reached the palate with more pressure and produced the distinct l sound, and in others the tongue was kept away from the palate and produced the r sound. Similarly, in the v-w sound in some words the lips contracted in and produced the w sound, whereas in others the closure of the buccal orifice was produced by the upper teeth touching the lower lip so that a v sound resulted.

The proper method of exact study would have been to watch and inquire as to the movements of the mouth and tongue in order to supplement the sense of hearing. Each word should then have been recorded phonetically and an alphabet compiled to meet the needs of the dialect. As it was, the compilers of the alphabets, with few exceptions, did not adequately represent the sounds of the Polynesian dialects. Table 1 gives the alphabets and sounds of some dialects for comparison with those of Tongareva.

If the s, which I doubt, l-r, v-w, and f-wh are treated as original consonant sounds, the total number of original consonant sounds is eleven. Of these Tongareva has maintained ten, but the f-wh sound is represented by h. There are thus no dropped sounds or glottal closures in the dialect. Furthermore, by duplicating the original l-r and v-w sounds into distinct l, r, v, and w, the dialect has added two more consonants and thus totals twelve consonant sounds. In the number of consonants it is only equalled by Manihiki. Tongarevan is more nearly allied with Rarotongan in consonant sounds, however, for, though Tongareva has the s, both are without the f-wh sound, and there is thus but one positive difference. Manihiki, by not having the s and having the wh sound, has two points of difference. In Tahiti the absence of the ng increases the difference from Tongareva. The presence page 12 of the s in Samoa is offset by the presence of the f and the dropping of the k and the standardization of l-r and v-w into l and v, which is opposite to the tendency in Tongareva. Tongarevan has also the distinct sibilant sound of ch which will be dealt with under the letter t.

Table 1. Polynesian Consonants
Original sound Tongareva Manihiki Rarotonga Society Islands New Zealand Hawaii Samoa
1. H h h h H H H (sorf)
2. K K K K ' K ' '
3. L l l L
R R R r
4. M M M M M M M M
5. N N N N N N N N
6. P P P P P P P P
7. T T T T T T (K) T
8. V V V V V v
W w w w W
9. Ng Ng Ng Ng ' Ng (N) Ng
10. F Wh (as F)
(h) wh (h) F (h) F
Wh Wh
11. S(?) S (h) (h) (h) (h) (h) S
11 12 12 10 8 11 9 9

Capital letter = Present in alphabet.

Small letter = Present in speech but not in alphabet.

( ) = Sound absent, but represented by letter in brackets.

Hamzah' = Sound dropped, but represented by glottal closure.

Rarotongan h, probably more correctly represented by the hamzah, is inserted for comparative purposes.

The h sound in Tongarevan is distinct. In a few words Tongareva follows the same usage as Rarotonga and New Zealand. In the following tables the h is inserted in Rarotongan words, though it is not included in the alphabet. The Samoan h is replaced by s or f.

Tongareva Rarotonga New Zealand Samoa Meaning
Tahuhu Tahuhu Tahuhu ('au'au) ridgepole
Aroha Aroha Aroha Alofa love
Hakiri ? Hakiri Sa'ili (to seek) to throw away

This h also appears in hanga, which forms the noun ending to verbs as surahanga (the turning over) from sura (to turn).

In most words, however, the h represents the wh of New Zealand and the f of Tahiti and Samoa. In this, it follows the Rarotongan usage.

page 13
Tongareva Rarotonga New Zealand Tahiti Samoa Meaning
Hare Hare Whare Fare Fale House
Hara Hara Whara (Astelia) Fara Fala Pandanus
Haka- Haka- Whaka- Fa'a Causative Prefix
Hari Hari Whiri Firi Fili To braind
Hariki ? Whariki Fari'i To place under

The Tongarevan k resembles that of Manihiki, the Cook Islands, and New Zealand and distinguishes the dialect from Tahitian, Hawaiian, and Samoan, in which the k has been dropped and is represented by the glottal stop (').

The distinct l is present in such words as talanga (story), talolo (messenger between lovers), and alelo (tongue).

The presence of the r was noted in such words as waru (to scrape), roro (coconut cream) and Rangi-saruru (an ancestor). Lamont (15) uses the r in several native words and never the l. It is probable that some words are in the intermediate stage, but on the whole the r is more common than the l.

The m, n, and p require little comment, except that n generally replaces ng in na for nga, the plural of the definite article, “the.”

The s, except in a few words, takes the place of the pure h of the Cook Islands and New Zealand and in this usage resembles Samoa. The following are examples.

Tongareva Rarotonga New Zealand Samoa Meaning
Saruru Haruru Haruru Salulu (to blow) to sound
Sape Hape Hape Sape mistake, club foot
Sere Here Here Sele to tie, to snare
Vavasi Vavahi Wawahi to break
Songi Hongi Hongi Songi to press noses
Sosore Hohore Hohore to peel off
Kaso Kaho Kaho 'aso part of house frame
Asu Ahu Ahu Asu to heap up, to bail out

Lamont (15) seems to have heard and sh sound, for he writes sarasara (to wave the hands) as “sharashara” and Hakasusa as “Haka Shusha.”

The t sound has undergone a curious modification in the sibilant sound of tch. The tip of the tongue is not so far forward against the incisor teeth as it is when producing the usual Polynesian t sound. The sibilant sound is more marked before the vowels i and e. Thus the line of the song, “Titia mai to titi maire” (Gird on your sweet-scented kilt) sounds like “Tchitchia mai to tchitchi maire.” The phrase “Taku ate” (my liver), used as a cry of alarm or anger, is sounded as “Taku atchě.” In the phrase quoted the t before e becomes sibilant, but the t before a is sounded as an ordinary t. Before o the t sound is unchanged, as in to (a spear), pro- page 14 nounced to. Before u there seems to be a tendency to use the sibilant, for, though in all the island names commencing with “Motu” the t seemed to be distinctly sounded, Mr. Wilson, the resident government Agent, held that Motukohiti was pronounced “Motchukohitchy.” Lamont, on the other hand, who wrote the sibilant t as ch wrote the same word as “Mutagohichy.”

The sibilant t does not occur in Samoa, but in Niue the t before i and e is sounded as s, which shows a similar tendency in sibilant change. Williams (32, p. 418) quotes the method in Tonga of variously representing the pronunciation of t when it is followed by i, j, or s. The nearest approach to the Tongarevan t appears to be in the Chatham Islands where the t undergoes a change in sound before all the vowels except o. As quoted by Williams (32, p. 418), Shand represented the sound by writing tch, tchi, tche or tc for t. The peculiar pronunciation also occurred with k before a and with h before a or o. Williams remarks that in producing the sound “the tongue appears to be somewhat arched into the palate and the letter uttered with a slight emission of breath which not infrequently produces the effect of a suppressed I, or sometimes E, sound before the proper vowel of the word.” Handy (8, p. 9) points out that in the Marquesan dialect, “In combination with different vowels there is a variation from a simple h sound to a distinct German ch, through sch to s, with a w sound often included.” The resemblance of the Marquesan treatment of h to that of the Chatham Islands further supports the contention of Williams (32, p. 420) that of the affinities he discovered between the Chatham Islands dialect and other Polynesian dialects that with the Marquesan dialect figures most frequently.

The v sound is present in such words as velo (canoe stern) and mataki-vikivi (to turn aside); the w sound in such words as wananga (teaching), waka (canoe), wahine (woman). In spelling native words Lamont uses oa as in oaka (canoe) and niu oara (grated coconut). The word for canoe must be either waka or vaka. If oaka, waka, and vaka are carefully pronounced, and the position of the mouth, lips, and teeth noted for each, it will be evident that Lamont's oa was meant to represent wa and could not represent va. The same applies to niu oara which was meant for niu waru, the oa again representing the w sound. In recording Polynesian words for the first time I have often failed to distinguish between the sounds oa and wa and have had to repeat the word slowly to get the correct sound. Lamont uses other words such as “oahine” (wahine), showing that before the advent of the Rarotongan alphabet the sound that he heard was w. Lamont did not use the letter v in any native word. This does not signify that the v was not present in the dialect, but only it was not present in the few words that Lamont used. Thus, the tendency with the v-w sound was more toward the w than the v, just as the l-r sound was more inclined toward the r.

page 15

The ng sound is present as in the Cook Islands, New Zealand, and Samoa as contrasted with Tahiti, where it is represented by the glottal stop, and with Hawaii, where it is represented by n.

The affinities in dialect given above are based entirely on the sounds. My impression from the vocabulary acquired during my brief stay was that there was more affinity with Rarotonga and New Zealand than with Samoa. The compilation of a complete vocabulary for the Tongarevan dialect may disclose other closer affinities. With a knowledge of the Maori and Rarotongan dialects I had no difficulty in speaking and understanding Tongarevan, whereas I experienced much more trouble with Samoan and Tahitian.

1 Numbers in parentheses refer to literature Cited, pages 221 and 222.