Published Native Texts

In printed native texts, apart from misprints which are liable to occur in any language, there are two items which require attention. One is the alphabetical symbol for the ng sound and the other is the glottal closure.

In the printing equipment with which the early missionaries were provided, the fonts contained all the letters of their own alphabets. As the Polynesian alphabets consisted of fourteen letters at the most, a number of the English letters were not needed. Among these was the letter g. In choosing the form of representation for the ng sound, some missionaries thought that it would render the work of printing easier if a single letter were used instead of the combination of n and g. As the letter n was included in the native alphabet, it was decided that the spare letter g should be used to represent ng. This was adopted in western Polynesia (Samoa, Tonga, etc.) and in French Oceania. In New Zealand, the compilers of the alphabet preferred to include the two letters ng in the alphabet as a double consonant. The New Zealand double letter was adopted in the Cook Islands. There is no trouble with the letter g if the reader knows that it represents the Polynesian ng sound, but if he does not know, he makes the mistake of giving it the g sound of his own alphabet. Thus the unknowing pronounce the important Naval Station of Pago in American Samoa literally as Pagopago or Paygopaygo instead of Pangopango.

The glottal closure, as the name implies, is a closing of the space between the vocal cords which, with a restrained emission of breath, replaces certain consonant sounds in speech. It is present in Polynesian speech and creates one of the dialectical differences in the language. In the islands in which it occurs, the glottal closure always replaces particular consonants, as follows:

h:	Cook Islands, Mangareva, New Zealand (west coast, North Island)
ng:	Society and Austral Islands
k:	Society and Austral Islands, Hawaii, Samoa
r:	Marquesas

A curious usage occurs in Uvea and Futuna where the long vowel a is pronounced as two short vowels with the glottal closure between them.

At the early stage in the study of speech, the missionaries did not consider the glottal closure worthy of a symbol, and consequently the consonants represented by the closure were omitted from the local alphabets. More recent linguists decided that the glottal closure should be represented by a symbol and selected the inverted comma, termed the hamza, which should be placed page 38superior to the position which the original consonant would have occupied in the written word. In actual speech, there is no difficulty in recognizing the glottal closure; but in the written word, a good deal of confusion is caused by the absence of a symbol representing it. Thus, totally different words may be written in the same way. A good example is furnished by the Cook Island word ua, which represents four different words identical with the Maori words, ua (rain), hua (fruit), uha (female), and huha (thigh). If the hamza is used to represent the glottal h, the words are readily distinguished as ua, 'ua, u'a, and 'u'a.

The necessity for the hamza was not recognized until comparatively recent times, hence it does not occur in early texts. Even the Hawaiian dictionary compiled by Andrews and revised by Parker does not use the hamza, though it was published as late as 1922. On the other hand, the hamza was correctly used in Pratt's Samoan dictionary published in 1911. In many of the Bishop Museum publications, insufficient attention was paid to the inclusion of the hamza in native texts. It was included in some words and omitted in others, and even the same words had the hamza in some lines and lacked it in others. Thus, the presence of the hamza in the native text does not mean that it has been inserted correctly throughout. The native texts have consequently suffered in their value as source material for students in Polynesian linguistics. The blame lies partly with the authors, for no editor could be expected to be an authority on the Polynesian language though some inconsistencies are obvious. As a result of experience, the Museum refuses to accept any native text from an area where glottals occur unless the hamza has been correctly included in the manuscripts submitted for publication. The errors of omission in the past were perhaps unavoidable at the time, but we now have no excuse for continuing them into the future.

Another process which should be understood is the shift which has taken place from one consonant to another in different dialects. The most widely spread are those between v and w and between r and l. The w occurs in New Zealand and Hawaii and the v in the rest of the Polynesian area. However, there is a possibility that the missionary committee responsible for the Hawaiian alphabet may have made a mistake in choosing w; in preference to v. The Hawaiians of today are using v in their speech instead of w; but its use, after a lapse of about a century, cannot be taken as proof that they are restoring the original consonant. The limited distribution of w may be taken as evidence that v more nearly represents the original sound in the language. In the r and l shift, r occurs in central, east, and south Polynesia and l in north and west Polynesia. However, the missionary committee in Hawaii had a close vote in deciding upon l instead of r and, again, they may have made a mistake. If so, r, from distributional evidence, might be regarded as being nearer to the original sound. It is, perhaps, possible that the shifts in the above two pairs were made page 39originally from intermediate sounds. Where foreigners found no difficulty in deciding upon an alphabetical symbol, we may assume that the shift had been complete, but where doubt existed to the extent of voting between two symbols, as in Hawaii, the shift was probably not completed. Thus, some of the committee heard v and r and others heard w and l in the same set of words. The acceptance of the alphabet by the Hawaiians also led, in time, to their acceptance of the sounds given to the symbols by their missionary teachers. In other words, the teachers completed the shifts.

Several more localized shifts have taken place. The most interesting is that from t to k which occurred in Hawaii. The shift was taking place in some of the islands when the missionary committee was considering the alphabet, but it had not affected Kauai. It is evident that some Hawaiians were using t (Tamehameha) and others were using k (Kamehameha). The committee put it to the vote and speeded up evolution by completing the shift to k. The Hawaiians had already replaced the original k with the glottal closure, and the committee helped them restore it to their speech, but in place of t. Thus, the Hawaiian dialect retains a glottal closure replacement for k and a k consonant as well ('umeke for kumete, bowl).

It is interesting to note that a similar shift from t to k has been taking place in Samoa in recent years, though it has not yet been universally accepted. Here again, k had been replaced by the glottal closure. There was no evidence of the shift when Pratt's Samoan dictionary was published in 1911. In 1927, the use of k instead of t was universal in American Samoa, both in speech and in writing. The London Missionary Society College for Samoan pastors in Upolu opposed the shift in their teaching, but in Tutuila the retention of t was regarded as pedantic, even by the orators.

In Ontong Java, the glottal replacement of k and the shift from t to k have taken place, whereas the Polynesian dialects in the nearest islands retain both the original k and t. Thus the shift from t to k has appeared in widely separated areas as independent developments. A number of other localized shifts occur which would form an interesting study for an expert in linguistics.