Volume 109 2000 > Volume 109, No. 2 > The Samoan category matai ('chief'): A singularity in Polynesia? Historical and etymological comparative queries, by Serge Tcherkezoff, p 151-190
                                                                                             Previous | Next   

- 151

In contemporary Samoan discourse as well as in all writings by Samoans or Westerners, from the 1930s to the most recent books and theses, the word matai is presented as the specific Samoan word which has always been used to designate ‘chiefs’. There is even a tendency to say that this specifically Samoan notion embodies the whole culture. Samoans often explain to the visitor that “the Fa'aSamoa (the ‘Samoan custom’) is the Fa'amatai (the ‘matai-system’)”. 1 It is certainly true in many aspects for the 20th century, for Western and Eastern Samoan politics (Tcherkézoff 1997a, n.d.b), even if, in the State of (Western) Samoa, recent governmental decisions have begun to abolish some privileges of the matai at the level of national politics. 2 But the equivalence between Fa'aSamoa and Fa'amatai may have been less true for earlier periods.

Scholars, Samoan and Western, have also tended to search for an etymology of the word matai within the sole Samoan context. The common hypothesis relates the word to the base mata-. 3 This is also problematic. There are linguistic difficulties which contradict this option, from a Polynesian comparative view and from the consideration of the 19th century Samoan society.

A revision of the matter is proposed here and opens a new consideration on the historical transformations of ‘chiefs’ in Samoa. The matai is not simply the Samoan type of the “Polynesian chief”. There are two shortcomings in such a statement. One is the recurrent idea that the “Polynesian chief” is fundamentally of one type, in contrast for instance to the “Big-man” of the Papua New Guinea highlands. The second shortcoming is the omission of the transformations that affected the various kinds of Samoan leaders during the last two centuries.

In fact, the word matai cannot be decomposed, because it is a base in itself, from the beginning of the Polynesian languages. We know this because the word is present throughout Polynesia—contrary to what is usually thought—with a very stable meaning of being ‘the best’, ‘the best through personal skills’, a ‘master’ in a craft activity. From this comparative point - 152 of view, it is the Samoan contemporary meaning of ‘chief’ which appears unique, but not the word matai in itself. The word could be Proto-Polynesian and thus linguistic comparison between Polynesian languages is called for. But the meaning of ‘chief’ is only Samoan and this singularity in Polynesia calls for an historical analysis.

Matai as title-holder

Milner's Samoan-English dictionary (1966, from his enquiries in the mid-1950s) accurately reflects the contemporary situation. It has for the following entry: “matai: n.l Matai, i.e. titled head of a Samoan extended family (formally elected and honoured as such)…. 'o le ~ tufuga: Master carpenter. 2. Master.”

The meaning of “titled head” is quite clear in Samoan society. The role of a craft specialist tufuga is also well-defined. But the distinction between what would seem to be two different uses of the word matai is not significant in reality, at least today, because each “master carpenter” happens to be as well the “titled head” of his family.

Let us explain first the notion of “titled head”. The definition of the Samoan extended family ('āiga) is the group of people who co-operate by preserving the name of a founding ancestor and who respect all the ritual obligations associated to this name. Western literature called these names “titles”, because, in each extended family, this founding name is preserved and given to one member of the family at each generation. This person becomes then “the matai of the family” ('o le matai 'o le 'āiga). Western literature says often “title-holder” instead of “titled head”. For the name itself, Samoans today say suafa matai, literally ‘matai names’. English-speaking Samoans use the expression “matai titles” more often, as does the Western Samoan Constitution of 1962 in its English language version.

An important distinction to note is that such a “matai title” can be of two kinds: “an ali'i title” or “a tulāfale title” (as said in English by Samoans; in Samoan, the word for ‘name’ is used: suafa ali'i). An “ali'i title” implies that the founding ancestor was a chief who enjoyed the prerogatives restricted to the ali'i, such as the right to have his own “kava name” (used in formal meetings when the ceremonial drink made of kava roots is presented to all the matai) and several other specific rights, all denied to a tulāfale. 4 In Western writings, ali'i were usually referred to as “chiefs” or “high chiefs” and the tulāfale as “orators”.

The matai as a “titled head” or “title-holder” is the person chosen by the group to carry the name of the founding ancestor. During each generation, this name must be embodied in a living person who is chosen by the whole - 153 assembly of adults. Without this embodiment, there would be no-one to speak in the name of this ancestor and, particularly, to deal with the lands associated with this founding name. The reason is that the only landowners recognised in Samoan aganu'u 5 ‘custom’ are these founding ancestors. In each 'āiga, the ancestral land-owner must be represented (embodied and made visible) by a living descendant, in front of other groups, within the village and the district community. This descendant, ritually empowered by the founding name (through a ceremony after which he is called “matai”), has then the right to say: “this is my land”, “this is our land”. Thus, he is able to preserve the land from any claims raised by non-members of this given 'āiga. 6

This logic was in operation when the first Europeans arrived. The system of choosing the chief and the subsequent identification of this person with the founding ancestor are signalled in the first writings of the missionaries and travellers. But there is one difference between then and now. There is a strong possibility that, at that time, the people who were only referred to as “matai” were not titled heads or were not “titled” in the same way as now. Primarily they were heads of local families. These local families had a say only within their nu'u ‘village’, while other chiefs, designated as ali'i, were the political and ritual heads of extended kin networks and, sometimes, of large territories, known as itūmālō, consisting of numerous villages and resulting from a mālō ‘victory’ in war. It seems that the ali'i who had authority on a large scale was referred to as a ali'i pa'ia ‘sacred chief’. 7 This meant that his title (the founding name of his 'āiga) had come to be considered by a great number of other families as “sacred”: founded by the first gods, blessed by numerous victories which had revealed the mana ‘supernatural power’ attached to this name, etc. This chief was not the “king” of vast territories and probably did not have full authority (pule) on all the lands of the itūmālō. He was still, like other family heads, the trustee for the sole land of his 'āiga. But, unlike other chiefs, he was the sacred centre of vast alliances between his village and other villages.

Title-holder and Craft Specialist

In Milner's definitions of the word matai for contemporary Samoan, the notion of “titled head” is clear. But the distinction Milner makes between this first meaning and the second one—master carpenter—is not obvious in situations observed today. 8 A tufuga is a craftsman, a specialist. He can be a specialist in the art of building houses (and formerly also canoes), or in the art of tattooing (still practised), etc. 9 He usually works with helpers from his own 'āiga. In this sense, there is of course a real distinction between a “master” and his helpers. But nowadays, and also for the last few decades, - 154 a tufuga is always the “titled head” of his family, and the two meanings in Milner's entry cannot be differentiated. We shall see that, in a more distant past, the word could have designated specifically the craft specialists independently of the question of title-holding. As Milner worked in the 1950s and with elderly people, the distinction he encountered could be taken as representative of a certain tradition. 10

In the contemporary classification, the matai is the ulu o le 'āiga ‘head of the family’. He is bestowed with the family “title”: the family ancestral name (suafa matai). Here we are discussing the 'āiga as an ‘extended family’ grouping. It is not just a household or a group of households ('au'āiga, pui'āiga, although sometimes the word 'āiga is also used). It is a grouping of the various people who, wherever they live, acknowledge a link (and had maintained it through participation in group duties) with at least one of the various persons who had embodied at some time in the past the ancestral name of that family. Those persons, who had thus embodied the ancestral name, form the core of the genealogy of this title and are said today to be “the various matai of the past”. But it does not tell us what terminology was used in the past.

Why am I raising the possibility of a transformation? One observation is enough. We said that the “title” of a given such family can be an “ali'i title” or a “tulāfale title”, “chiefly title” or “orator title”. Thus, in the taxonomy, the category of “matai” is broader than the category of “ali'i”: matai=ali'i+tulāfale. But this conclusion creates a singularity. It is the only case, throughout Polynesia, in the present or in the past, where the ali'i as a group is encompassed in a broader class. This fact in itself is sufficient to allow us to think that the contemporary Samoan situation is a historical transformation. Let us look now at the literature.


Every contemporary official and informal presentation of the Fa'amatai, written or spoken, by Samoans or Westerners, conforms to the pattern we have summarised. This is also the case in the recent past, as one goes backwards checking through the literature, including (for Western Samoa) the New Zealand administrative reports, back to nearly the beginning of the Mandate.

One can consult a booklet compiled by the New Zealand Administration established in Western Samoa (the “Department of Native Affairs”) on “Samoan custom” (Grattan 1948), prepared in 1942 for the information of the U.S. personnel who were stationed in Western Samoa during the Second World War. The preface says explicitly that the booklet does not try to reconstruct any “tradition” but concentrates on “present day conditions”. What are those conditions?

- 155

The “unit of social life” is the 'āiga, described accurately as a widely extended group through blood, marriage and adoption connections. It is made up of people “…who acknowledge one person as the matai or head of that particular family. Such a matai is a titled person, either a chief (ali'i) or an orator (tulafale)” (Grattan 1948:10). Examples of high ali'i names are considered as “matai” names (p.11). The general distinction between titled (matai) and untitled (taule'ale'a) is presented as it would be today (p.12). And when the author mentions some ideological changes in the respect paid to the matai, he uses sentences like: “But whereas under the old order the matai commanded a special respect…” (p.149), which show that this political category of matai was certainly not regarded by the author as a new one. Grattan's work reflects the situation in 1940, and probably that of the 1930s too, because if there had been any major change in the preceding ten years, it would have been mentioned.

Twenty years before Grattan's work, in 1925 precisely, the New Zealand Administration of Western Samoa had newly established itself in the western part of the archipelago through a Mandate of the League of Nations and it had published a Handbook of Western Samoa (author unknown, Wellington, Government Printer), apparently written in 1924. 11 The chapter describing the “people” reads as follows:

In the Samoan social system the unit is the family—or, rather, an aggregation of families forming a clan. The political system is so nearly coincident that for the present purpose it may be treated as identical.

The clan, whose head is called a matai, owns all the land and parcels it out to its members as necessity arises. All produce of the soil is theoretically the property of the matai in trust for the community, but in modern days it is becoming increasingly common to allow the actual cultivator to retain for his own use the fruits of his labour.

The chieftainship is elective, though the son of a deceased leader often succeeds his father. The power of the chiefs is apparently great. They are treated with punctilious respect, and are hedged about with strict etiquette. In reality, however, the power is in the hands of a group of men called tulafale, or orators, whose function is that of Ministers of State…. Every important affair is in their hands, and they may even depose a chief…. (Handbook… 1925:42). 12

Clearly, what the Handbook calls a “clan” is the 'āiga. Yet, the Handbook seems to equate “chieftainship” and “clan's head” called “matai”. This last word seems then to be the general one. But, still the “chiefs” and the “tulāfale” are distinguished. Apparently, things are not as clear as they will appear to be later on when authors say directly that all chiefs are “matai of - 156 the ali'i or tulāfale type”. It could be a sign that we are approaching a transitional period where the notion of “chief” still referred primarily to the ali'i. Data from the turn of the century confirms this.

Peter Buck in American Samoa

Confirmation of the fact that, in the 1920s, we are at the end of a transition is given by the writings of Peter Buck—Te Rangi Hiroa, who undertook fieldwork on material culture in the whole archipelago in 1927 13 and published his Samoan Material Culture in 1930. On one hand, we find in Buck's text a general presentation which is similar to all subsequent literature:

… the usual Samoan custom of ceremonial feasting when a person is elected to the position of matai or chief (p.85).

The laulautasi is a presentation of cooked food to visitors by the chiefs (matai) of the village. The sua ta'i with pigs baked whole are given by high chiefs whose position of wealth enables them to give it. The laulautasi is a contribution in which all the lesser chiefs share (p.142).

In these examples, all chiefs are termed matai, whether “high chiefs” or “lesser chiefs”.

But, on the other hand, when Buck presents what he had been told about the ceremonial division of the cooked pig, 14 the matai-type of chiefs occupy a considerably more limited place, being superceded by other kinds of chiefs.

Figure 72.—Pig, ceremonial divisions: 1, ulu (head) to the au manga (young men who cook); 2, ivi muli ulu (neck) to the tulafale (talking chief); 3, o'o (back) to the ali'i (chiefs) of second grade; 4, alanga lima (shoulder) to the tulafale (talking chief); 5, tuala (loins) to the ali'i of the first grade; 6, itu mea tele (big side) to the ali'i of second grade; 7, itu pale asu to the family of chief; 8, muli to the women; 9, alanga vae (leg) to the matai (chief of lesser rank); 10, alo (abdominal wall) to the taupou (village maid) (p. 121).
- 157

In contradiction to the previous citations, Buck tells us here that matai are “chiefs of lesser rank” (9), distinguished from high ali'i (5: “ali'i of the first grade”) and also from other ali'i (3: “ali'i of second grade”). The previous citations are from pages where Buck summarises or evokes briefly the social structure as he understands it generally. Here he notes in detail what has been said to him. This piece of information is then highly relevant.

Another detail is worth noting. It is obscured by the order followed by Buck, an order obviously determined by Buck's geometric apprehension of the animal (see the image, numbers go simply from left to right, beginning with the head). We know that this is not the ceremonial order of presentation which normally gives precedence to the items 5 and 10. Also the order always opposes parts abundant with grease and meat, that are the lot of the tulāfale and lesser chiefs, and pieces where there is much less to eat, which are for high chiefs, i.e., ali'i and not tulāfale, and ali'i whose title is high in the hierarchy of founding names of the polity where the distribution is done. The parts without much meat are for that very reason the most mamalu ‘dignified’ parts and thus are fit for the ali'i, as Samoans explicitly say. 15

This opposition is well known, even today, and common to all villages. The chiefs who speak the most, the loudest, who eat the most, who receive big mugs of drink, etc. are always the lesser ones. Those who have the final say are all the contrary. In the ceremonial context, where people are under the guidance of the divine, it is not by expressions of force but by expressions of restraint, of silence—a sign of the intensity of the communication or tapua'i between this person and the divine—that one imposes his views on others. In this context, the fact that the share of the matai (9) belongs to the class of abundant pieces given to the tulāfale (4 and 2) is very significant: rear legs/front legs, in contrast to parts 3 and 5 (the back) that are for the ali'i. 16 At a secondary level of distinction, it could even be pertinent that the front leg goes to the tulāfale, while the rear one goes to the matai. There are other indications from the 19th century that seem to distinguish (i) high ali'i, (ii) lesser ali'i, (iii) tulāfale as orators linked to a high ali'i, and lastly (iv) the bulk of “heads of a family”. Unfortunately, in this literature, the word matai is not mentioned: one finds only the words or expressions “ali'i”, “tulāfale”, “chiefs”, “rulers”, “heads of a family”. 17 Here we are concerned only with the main distinction between the ali'i and the matai.

In another discussion on kava, Buck speaks of the necessity for all chiefs to have a stock of dried kava always at hand for constant calls made upon them by visiting friends and relatives. He adds that: “Those below the rank of matai (head of a small family group) were exempt from such calls” (p. 147), as if the “matai” level is the last one of a range of levels of chiefs. The order referred to is not specifically the rank of the title but the fact that the - 158 family headed by a matai is a “small” one. We will see in the last section that, indeed, throughout Polynesia, the difference between a multi-local extended network headed by an ali'i, ariki, etc. and a localised small group that has its headman could have been a recurrent distinction.

When describing the tattooing of an adolescent, Buck differentiates in the same way:

Preliminary procedure. For the children of ordinary chiefs (matai), the father consults the artist he desires and they arrange the date for commencing the operation. On the appointed day, the artist visits the village and commences work without any ceremonial.

If a chief's son, the father visits the expert and offers a fine mat to seal the bargain… (p.641).

The three aforementioned examples are the only ones that can be found in the voluminous work of Buck, probably because the book deals only with “material culture”, thus providing few opportunities to describe the role of the various chiefs. But these examples are enough to resolve the matter. Without any doubt, in the minds of the older informants of 1927, the ali'i were not a sub-class of the matai. Quite the contrary, matai were lesser chiefs compared to ali'i, at least compared to a certain number of ali'i. Because Buck worked specifically with older informants, 18 his enquiry of 1927 tells us something of the beginning of the 20th century. Obviously, these years represent a period of transition for the topic of our enquiry.

Erich Schultz in Western Samoa.

The early years of the 20th century were indeed a time of transition in the whole archipelago. In the western part, it saw the advent of a German colonising power that would last until the First World War (1899-1914). A short publication on “The most important principles of Samoan family law” was written by Erich Schultz (1911) who is referred to as the “Imperial Justice, Apia”. This text was translated into English from a German manuscript with the approval of the author. Schultz, who shared with the Governor Solf an intellectual interest—and an applied interest for colonial reasons—in Samoan history and social organisation, was Chief Justice until 1910, under the Governorship of Solf, and succeeded to Solf's office that year. The article then must have been written around 1909. As Chief Justice, Schultz also dealt with all cases involving lands and matai titles. He was especially interested in this topic, as he was the instigator of the institution created by German Administration in 1903: the Land and Titles Commission. This body, under his guidance, even if it included several senior Samoans, was supposed to rationalise principles of succession to titles and land, and settle disputes. 19

- 159

Through this institution, the Germans wished to put an end to internal warfare—they succeeded in that—and to control the chiefly system, since their plans were to transform Western Samoa into a huge copra producing farm. They stopped any proliferation of small independent non-German settlers and dealt with high chiefs concerning any questions of land that arose. 20

Schultz explains the Samoan “principles” as he sees them:

The Samoan race is divided like a clan into families, āiga, which again are split up into groups of branches…. At the head of every branch stands the matai, or head of the family. One of these is the chief, matai sili, of the whole clan. Every matai has a name—igoa, suafa—which is handed down from generation to generation….

1. The matais of the family branches are either subservient to the rule—pule—of the matai sili, or they have their own pule. This depends upon their origin, which varies in different families, and is generally to be traced back to the decree of the founder of the family, or some other ancestor….

2. The position of matai makes itself felt not only in the family but in the village. The family is the centre of social life; the village, that of political life amongst the Samoans. Both are so closely interwoven, however, that neither can be understood without the other. The village, nu'u (and) a'ai consists of parts…. The political organ of the village is the “village meeting” fono fa'ale-nu'u, which takes place on the village green, an open place, malae malae-fono. Matais only have part and voice in this assembly….

The matai is either ali'i, “chief”, or tulāfale, “speaker”. We suspect that the class of “speakers” has sprung from the servants or dependants of the “chiefs”. The other meaning of the word tulāfale is “houseroom”, and the circumstance that the word matai has only lately been applied to “chiefs” lead us perhaps to the conclusion that the tulāfale was the original, and at first the only apparent form for the head of a family (1911:43-46: emphasise are mine).

The rest of the text does not give any commentary or indication of what Schultz refers to by: “the word matai has only lately been applied to ‘chiefs’.” But when Schultz evokes this “circumstance”, he seems to present it as something quite well-known or obvious, since he mentions it only as being part of the reason for another discussion (his hypothetical history of relative power between ali'i and tulāfale).

Schultz wrote this text in 1909 (in German). He used for it his observations on the first decade of the century. The system had already taken or was taking the form we know today, since Schultz wrote: “The matai is either ali'i, ‘chief’, or tulāfale, ‘speaker’.” But this encompassing category of matai is a recent extension—at least the fact that the category of ali'i is - 160 encompassed within the notion of “matai”—since Schultz specifically informs his readers: “…has only lately been applied…”.

Of course, what Schultz writes is not only an “observation”; he was also one of the main actors in the establishment of the German administration. What he says on the matai could reflect how he and Solf wanted things to be. Their aim was to bring all the high chiefs and high orators to one and the same level, as a body of native representatives of the various districts, operating under the authority of the German administration. 21 Of course, Schultz could not have created in a ten-year period an entirely new notion of inclusive matai-ness; the movement probably started earlier, owing to the effects of the missionary presence (Tcherkézoff n.d.a). But Schultz certainly added impetus to it by labelling the old difference between the main title-holder of a large family (usually an ali'i) and lesser chiefs holding titles of subgroups (other ali'i and tulāfale) simply as a difference between “Matai Sili” and just “matai” (see the section 1. of his text quoted above). 22

No “Matai” at All?

One thing that is astounding is that in almost all the literature published by travellers and missionaries the existence of the matai is completely ignored. The word is not mentioned. There are many details given about “high chiefs”, “chiefs”, “rulers”, “chiefs and orators”, etc., but the Samoan words quoted are only ali'i (pa'ia) and tulāfale. There are two significant exceptions: the missionary dictionaries of the mid 19th century and Augustin Krämer's massive ethnography of the late 19th century. These two sources have in common a certain attention to detail and especially to Samoan terminology. For compilers of dictionaries, this attention is, of course, a requirement. Krämer's case is more unexpected. He is the only ethnographer who makes numerous notes on the supposed meaning of Samoan words, adding lexicons to his volumes and giving extensive narratives in the vernacular. His background as doctor, geologist, botanist and German museographer explains this. A third exception should be added to our list. William Thomas Pritchard's book mentions the word matai once but does not elaborate at all. Still it should not be discarded, because this young man, son of the famous missionary and British consul George Pritchard, had for several years in the 1850s extensive interaction with Samoan friends and families.

The general absence of the word matai (except when detailed terminology is required because of the specialisation of the work) tells us that the “chiefs” with whom Europeans interacted were the ali'i and their tulāfale. Apparently, when these Europeans participated in a fono (formal meeting), and when - 161 they recorded myths and legends, they never heard the word matai; or, if they did, they interpreted it as a negligible notion, not worth mentioning. Even in the Marist dictionary, the way the word matai is dealt with, gives the same impression (see below). Such a constant omission cannot in each instance be the result of the inability of the observer to understand what is being said and shown to him. It must then be assumed that, from the beginning of detailed literature (1830) up to the end of the century, local classification could not have been the same as today: in those years, clearly there was not any encompassing class of “matai” of which the “ali'i” would be a sub-division.

Let us now look at the three sources which are exceptions to this silence concerning matai.

Missionary Dictionaries

We have the London Missionary Society dictionaries (i.e., the successive editions of “Pratt's dictionary”). One can follow the historical transformations of a number of entries, from the first edition (1862) through the following editions: the second edition in 1876, the third in 1893, the fourth and last in 1911. But for the entries “matai” and “ali'i”, there is no change:

  • —MATAI, s. the head of a family. 23
  • —ALI'I, s. chief

There is nothing specified about matai as being a specialist tufuga. If one looks at this latter entry, one finds:

  • —TUFUGA, s. I. a carpenter, II. a tatoo-marker.

And it is only the entry ALI'I that is given the meaning ‘chief’, without examples to clarify the issue. 24

The history of Pratt's first edition (with long delays in finalising the list and publication) tells us that the words were collected and analysed mainly at the end of the 1840s. Thus, it is quite significant for us that the distinction ali'i ‘chief’/ matai ‘head of family’ is already present in the 1862 edition. Hence, we can say that this distinction existed in Samoa in the 1840s, and that this distinction existed before any major European influence; 1845 is only 20 years after the beginning of intensive contacts with European ships, 25 15 years after the first contact with missionaries, and nine years after the establishment of the first permanent LMS missions.

The other missionary dictionary available for the 19th century was compiled by the Marists (Violette 1879; data are from 1845-54, thus strictly - 162 contemporaneous with Pratt's data). The Samoan-French section has of course the expected entry: “ALII: seigneur, maître, noble; O alii les nobles; Monsieur (Sir)”, but there is no “matai” entry. As, generally speaking, Violette's dictionary is quite good and detailed, we may assume that the word matai was uttered rarely enough (at least in front of Europeans) to be disregarded by this lexicographer. Violette did not purely ignore the existence of the word, because it can be found in the French-Samoan section. But it is in a kind of secondary level presentation, which is another sign that Violette did not give much importance to this word:

  • CHEF: 'O le alii (leader, chief).
  • CHEF-LIEU: O le tumua, laumua (the chief town). Chef de famille, 'O le matai, 'O le alii matua (the head of a family).

As French speakers know, the expression “chef de famille” means only “family head” and does not imply anything of the social status or rank of the person, whereas “chef” is always understood as a political/ceremonial role. Oddly enough, the entry “chef de famille: matai” is presented within an entry beginning with something entirely different; probably because Violette considered that, in French language, “chef-lieu” (the main town of a district) and “chef de famille” are both metaphoric uses of the word “chef”.

We can conclude from these various quotations that early dictionaries make a clear distinction between ali'i and matai. They indicate the word ali'i constantly as prominent: ‘chief’; and they either do not mention the word matai (Violette in the Samoan-French section) or mention it only as ‘head of a family’/‘chef de famille’.

Krämer's Data: Matai as Heads of Families at Village Level Only 26

Krämer stayed in Samoa for almost a year during his first round trip to the Pacific (1893-1895), driven by his interest in the marine species of the Pacific. He came back to Samoa for 11 months during 1897-1899. Like Buck, he carried out his work in the spirit of salvage ethnology, even if he arrived there 20 years earlier. Krämer makes it clear in his introduction that Hawaii left him with the impression that the traditional way of life had practically disappeared, and this is the reason why he wishes to map out and describe the way of life in Samoa before it was too late. Hence, Krämer also worked with older informants, chiefs and orators. His data can then be considered as representative of the last third of the century.

Predictably, Krämer includes some pages on the generalities of social organisation. Here, one finds the distinctions between ali'i pa'ia, other kinds of ali'i and, of course, the tulāfale. In these passages, the notion of “matai” - 163 does not appear, which is again a sign of its secondary role in the whole setting, even in those very last years of the century when Krämer gathered his information.

Krämer presents the notion of “ali'i pa'ia”, refers to those chiefs as “TitelHäuptlinge”, cites as examples Tuiatua, Tuia'ana and others, and describes the various taboos surrounding them. Following from this, he mentions the “Tulafale”: some are directly attached to an ali'i pa'ia name, some are independent. Their own title is transmitted from one to another within their family. Groupings of orators, called in specific districts the “Nine Houses”, the “Six Houses”, etc., are in charge of choosing and bestowing a given ali'i pa'ia name and are thus very powerful (Krämer 1902:10-11). Even in his description of the kava ceremony held by the participants in a village fono (a Samoan narrative 1902:20-22), the word “matai” is not mentioned. The only people referred to are the ali'i, the tulāfale and the 'aumaga (untitled men who sit outside and carry out the collective work as decided by the fono). The narrative stresses the fact that the ali'i have a personal kava name and a personal cup from which to drink the kava, while the tulāfale do not.

Let us come now to the matai as head of family in Krämer's analyses. In his first volume, Krämer is concerned mainly with genealogies and ceremonial sets of phrases used for each polity (villages, districts). He lists them all and gives a set of Samoan narratives. This explains why only ali'i and tulāfale are recorded. But there are two passages when matai are mentioned, within a discussion on another topic (see below). In his second volume, Krämer presents the “ethnography” and his text is thus more detailed; it is less concerned with the ceremonial side and focuses mainly on “daily life”. Here the persona of the matai becomes quite visible:

The differences between chiefs and orators (ali'i and tulafale), 27 their rights and duties were dwelled on extensively in the first volume …. But of how to distinguish in daily life a chief from a nonchief 28 so far little has been disclosed. A well known expression is the following: “On Samoa, the difficulty is not to know who is a chief, but to know who is not”. 29 This is quite correct to a certain extent, as the families of chiefs were quite numerous and large, and since the members thereof considered associating with the white people their privilege, in former days visitors to Samoa encountered very few common people. Today another essential element is added in the fact that the mission has broken the power of the chiefs so that today any smart alec can dare to pose as an ali'i. However for the informed there exist even now enough signs by which to recognise the chief of a village or of a family, and by that title I should like to identify in the following lines solely the alderman, the matai, apart from title chiefs and of course of kings, 30 all - 164 of whom were sufficiently discussed in the first volume, f.i. p. 10 31 (1995:108-109).

Krämer distinguishes the matai from “title chiefs” and from “kings”:(l) the matai, “Häuptling”, within the village setting: either “the chief of a family” or even “the chief of a village” (the chief of the main family within the local hierarchy), who is usually an ali'i (also described as “Häuptling” by Krämer in contrasting him with the orator); (2a) “title chiefs, TitelHäuptlinge”, which, given the context, we must understand as ali'i titles whose importance is defined as being at a broader level than the village (see Krämer's usage of these words in his vol. I); and (2b) “kings”: historical and legendary tupu, and the recent Tama ‘āiga titles (see Tcherkézoff n.d.a) who had a national role and entertained rivalries at a national level during the period 1870-1900, when they dealt with foreign powers and with local settlers’ factions. This ternary distinction is made by Krämer according to what he sees (“for the informed there exist even now enough signs…”). Hence it can be dated from the end of the 1890s.

It is then clear that, in the last decade of the century, the notion of “matai” as a “chief of a family” was there at the village level. But it was still within a transformative process. If this were not the case, Krämer would not have said that only the well-informed can discriminate as he does. He would not have considered it necessary to mention here that the power of contemporary ali'i is “broken” compared to what it is said that their power was in earlier times. Neither would he have added that today any one “can dare to pose as an ali'i”.

In volume I, Krämer mentions the word “matai” twice with the meaning of head of family, but he does it very briefly while discussing other topics. Presenting the notion of “Familie”, he stresses that, as in systems used by the nobility, the name of a Samoan family is transmitted to only one member. But in Samoa, Krämer adds, it is not necessarily the first-born who carries the name. The choice is made taking into consideration who has served and helped the group. Thus,

it can happen that a head of the family (matai) 32 gives his name, and along with it, the authority over the family to any member of the family and even to a non-kin adopted child (1902:31, my translation).

In another section, Krämer describes marriage. He writes: “If a young chief [junger Häuptling]” wants to marry a certain young lady “and if the orators of his village agree with it because her family is well-known and possesses many fine mats”, preparations begin. Intermediaries go to the young lady's - 165 village to see for themselves how great the dignity of the family is. Initial gifts of food are made. “If the chief of the lady's village [Dorfhäuptling] accepts these gifts”, the marriage then proceeds. If the gifts are not accepted, but the lady still gives signs of a possible acceptance, more gifts are brought, and this time “goes the prospective bridegroom chief or chief's son” [der freiende Häuptling oder Häuptlingssohn]. The family of the prospective bridegroom tries to use all its influence on their village to gather more gifts. The question is then, Krämer says:

to know whether the whole village becomes interested because the genealogy of the bride is so high, and to know whether the bridegroom chief 33 has enough dignity, if he is or will be a matai, 34 so that the whole village will follow him (1902:35, my translation)

The text is difficult to interpret. Krämer's vocabulary is often inconsistent, in that he varies his German words even when only one Samoan notion is involved (see for instance his various translations for “matai”: Älteste, Familienhäupt[ling/er], Familienoberhäuptling). Apparently, Krämer means that the bridegroom is of a chiefly (ali'i) family, whether the chief himself or one of his sons, and that, if he is a son, his status is enhanced if he is already designated as the future head of his family. But we do not know if Krämer introduces here the word “matai” from his own initiative, as an ethnographic touch for his meaning of “head of the family”, or if, really, he has in mind cases where any great ali'i (and not only a head of a small local family, see below) is also called a matai in his capacity as being the head of his own family. If this is so, as Krämer implies also in the first passage we have quoted, it must be a recent use, because of Schultz's remark (quoted above) on the “circumstance that the word matai has only lately been applied to ‘chiefs’” and because of the absence of this notion in the 19th century literature before this.

In the lexicon of Samoan words given at the end of volume I, there is an entry “metai”, which proves in itself that the notion is clearly part of Krämer's view of the Samoan society. It reads:

matai: the head of family, who carries the name of the family and who rules over his family (see the entry āiga) (1902:479, my translation). 35

The entry “āiga” to which Krämer refers his reader is the following:

'āiga: the family, the relatives. The long ā contrasts with 'aiga, 'aina, to eat. The āiga is the main factor for gaining power in Samoan society, because everything rests on greatness and respect pertaining to the family. Each - 166 family has a matai, who carries the name of the family, while other members retain only their proper names…. If a family is very large, it subdivides into sub-families which are called ituāiga or puiāiga and which also have their own matai. But all these matai are then all under one matai sili. The word āiga applies also to certain villages who have kinship links with a high chief [Grosshäuptling], in totality or in part. This is how people talk about the āiga of Malietoa, of Tuiaana, etc.; āiga is also the name for Manono (āiga o le tai) (1902:476, my translation). 36

It is important to take notice of the following two points. We can see how Krämer views the system: “matai sili”/“matai” (as leader of subgroup), in a manner similar to that described by Schultz not long after. We also see how Krämer implies that the matai, whether the head of a “subfamily” or of a whole family (and then considered “matai sili”), is of a different order than the “high chief” to whom a whole village is his “family” (and even a set of villages, as it could be added from other observations).

The Matai in the Samoan Narratives collected by Krämer

Until now, the various remarks we have considered from Krämer and which dealt with the existence of the matai are his own comments. We must now look and see whether there is any Samoan narrative which includes the mention of this word. There are only two instances of this.

The first one is in volume I, in a discussion of the administration of the village. In Krämer's presentation of the whole status system, and in the Samoan narratives added to illustrate that topic (such as the distribution of the kava, etc.), only ali'i and tulāfale are present. But when Krämer adds a Samoan narrative to illustrate the “village administration”, matai crops up. This very fact seems to show that, at this time, the notion of “matai” was operating only or mainly at the village level.

Krämer begins by telling his reader that in the beginning of volume I he has described fonos (assemblies) of districts. Now he wants to speak of a village in itself. Then he gives a Samoan narrative. The title of the narrative is “The authority of the village ('o le pule a le nu'u)”. It begins: “If the village gathers to hold its formal meeting ('Afai 'ua potopoto le nu'u e fai sa latou fono)….” The text then describes the “dignity” (mamalu) of the occasion, how the “aumaga” (the untitled men, also called taulele'a: the group of men who are not heads of families) sits outside the house where the meeting is going to take place; it mentions only that speeches are made about regulation of life in the village, that then the kava is drunk, and then food is brought.

One chief (ali'i) says: “Bring my two pigs and my fifty taros!” An orator (tulāfale) also says: “Bring my two pigs and hundred taros!” …Another one - 167 says: 37 “the decision of our village is made …”. Another: “Your speech is right, your advice is right.” The untitled men (taulele'a) prepare the food; the group of orators (faipule) continues to have speeches in [this] last part of the fono. 38 Their speeches are….

The text records all kinds of decisions about planting taro, coconut trees, etc. and then goes on:

Then the untitled men (taule'ale'a) bring the food. The one in charge of the authority says: 39 “take fifty taros and one pig and bring them to the pastor … and share among people outside and inside!” When the meeting is over, the speech 40 is made from inside the house for those outside: “Listen you the untitled men (aumaga), our meeting is over, look carefully to [what had been said]”. Then all the words which had been [said] there [i.e., in the house of the meeting] are carried away. Then [the house] is emptied, the matai goes, he carries [the words] to his family (aiga) 41 (Krämer 1902:40-41, my translation from the Samoan).

It is notable that during the meeting people are designated as ali'i and tulāfale. But once the meeting is over, the people who carry the orders to their respective family groups become known as matai. But, again, we do not know if everyone at the meeting, including the main ali'i, can be termed matai when he is out of the meeting and acting as the head of his family, or if this applies only to the orators. In his German translation of this Samoan narrative, Krämer translates “matai” by “Familienhäupter”.

The other Samoan narrative is found in volume II, in the section describing the gift of fine mats from a chief's family to the builder who constructs for this family a fale tele, the largest Samoan type of house and the most difficult to build:

The following are then the words of the carpenter (tufuga) to the chief (ali'i): “It will be good if tomorrow you prepare the feast for our work.” And so the chief says: “it is well.” Then the chief goes right away and counsels with his family. 42 And the chief speaks as follows: “Tomorrow we will gather fine mats, beginning in the morning and until the afternoon. Then our family is to assemble in my house 43 in order to count our mats.” When that was carried out that way the chief (ali'i) says to his whole family: 44 “Now spread your fine mats!” So one of the matai says: 45“Good, our gift is ten pieces of bark cloth and five mats, which we will bring into your house.” 46 And likewise another matai of another family says: 47 “We have here four fine mats and four pieces of bark cloth.” And so they collect the fine mats of the chief (ali'i) and the branches of his family, about fifty, and two hundred pieces of bark cloth. 48 So the chief says: “Let us go and take them to the house where - 168 the carpenter (tufuga) is.” So the carpenter says: “Welcome your highness!” 49 And the chief answers: “You deserve our special esteem, artisan of the king” 50 (Krämer 1995:276). 51

This piece of information makes things clear: the matai are heads of “families” 'āiga (which can be specified as being “au 'āiga”, see the end of the excerpt), but these families are subgroups of one extended family (also called 'āiga) headed by an ali'i.

Of course, as is usually the case with the narratives collected by Krämer, these examples are taken from a high family. We must not exclude the possibility that some chiefs of subgroups are themselves of the ali'i type, and certainly some must be of the tulāfale type. Their “subgroups” can be “families” at the village level, if the main “chief” ali'i mentioned here has a title of regional importance. But when they are considered in relation to this chief of the whole extended family, when they act as leaders of their own subgroup, they are designated each of them as “matai”; while the chief of this whole group of subfamilies, to whom the various matai bring their share, is only designated by the narrator as “ali'i”. So the event recorded replicates with regard to the position of matai that in the first Samoan narrative. This position cannot in any way be equated with the matai position as we know it for the 20th century.

Should we consider that this situation is also different from the 1830s to 1870s when the notion of matai as “head of family” existed indeed (since it is given as such in missionary dictionaries), but never came up in the missionaries' and travellers' descriptions of actual family gatherings and interaction with Europeans? Of course, these early descriptions are limited to interaction visible to Europeans, and do not include such narratives as those recorded by Krämer where in depth details of any interaction are mentioned, as here between a chief and the leaders of subgroups. So it could be that the situation given by this narrative is the same as at the beginning of the 19th century. But the transition is nevertheless obvious, for two reasons. First, this situation is not what it is going to be shortly hereafter. Second, the distance between the sacrosanct chiefs and other chiefs has already been reduced, as Krämer noted (above)—“today any smart alec can dare to pose as an ali'i”, etc.

Matai-tufuga in W.T. Pritchard's Book

Polynesian Reminiscences was written by William T. Pritchard, who lived in Samoa for ten years, from 1848 to 1858. Because of his long stay during a time when he was not involved in official duties (he became consul only - 169 in his last year there), he had time to develop a deep interest in local customs. He wrote in depth on the political and social aspects of “the chiefs”, but unfortunately he uses only this English word and no Samoan words. From the names quoted for some chiefs in specific events, it is obvious that Pritchard refers only to the ali'i. He also evokes the presence of “speakers” in descriptions of specific events (1968:92). On page 135, tulāfale appears as “councillors” but also “head of family”:

Systematic wooing and the formal nuptials were managed as follows, in the case of chiefs: —The Tulafale, or heads of families, who, as such, are the chiefs' councillors, met in solemn conclave, and selected a bride of suitable rank.

A few pages later, he describes how “a girl is admitted into the status of womanhood” through a simple distribution of property, without any ritual manipulation of her body (p.142), and then he goes on to speak about boys and tattooing. Here, for the first and only time in his book, the word matai appears:

In singular contrast is the custom under which youths are still initiated into manhood by the operation of tatooing…. A young chief is usually tatooed at about the age of eighteen, and when the time for the operation is come all the lads of his tribe, perhaps twenty in number, from the age of fourteen and upwards, prepare to join him. Tatooing is a regular and honourable profession, and the operator ranks as a matai—a master or professor. When application is made for the services of a matai, the application is always accompanied by a present of fine mats…. The party of operators consists of the matai, and five or six assistants… (Pritchard 1968:143).

The implicit classification is thus: ali'i—higher, tulāfale—lower as heads of families and chiefs' councillors, and, out of this status system, the matai as a person defined by his “profession” and by his qualities of being a “master or professor”.

Matai-tufuga in Krämer's Book

The well-developed organisation of public life is further demonstrated by the existence of a well-regulated pattern of crafts and trades…. It must once more be kept in mind that such crafts and trades were chiefly the property of certain families who anxiously watched over their rights. The head of family, the matai [Ein Ältester, matai] 52 was at the same time the master, tufuga [der Meister, tufuga], who had a host of helpers 'au… under him whom taught the trade. Payment for services rendered was in the form of fine mats, bark cloth, food, etc., depending on performances. The principal - 170 workmasters [Werkmeister] are the following: Men's Work: 1. 'o le tufuga fau fate, the house builder…, 10. 'o le tufuga tā tatau, the tattooer…, 22. 'o le tufuga fili'afa, the rope maker… (1995:97, German words from the original edition 1903:90).

Krämer's remarks here are the only ones, of all the literature, which attempt to clarify the duality of the matai as “head of family” and the matai as “principal workmaster”. The “master” of a given craft is called tufuga followed by the word of the given craft. His helpers form the 'au + the word of the given craft ('au- means ‘the group of’ and can apply to any group of persons, institutionalised or not). The master is the head of his family, and he is called matai because this is the word for “head of family”.

But another passage on these specialists is a little less clear:

Just as the man who does the incision 53 was called tufuga tefe so the tattooer is called tufuga tātatau. 54 The trade normally stayed hereditarily in the family as seen in Vol. I p. 152, and the builders of houses and tattooers enjoyed among the people special recognition as agaiotupu “the King's workmen”. A matai, master, simply called tufuga or autū o tufuga, normally has several, up to six, helpers, 'autufuga (Krämer 1995:72).

This passage taken alone leaves open the interpretation that “matai” is an (ceremonial?) equivalent of “tufuga”, the latter being the “simple” word (“simply called tufuga”). Some of the lexical comparisons of Polynesian languages would corroborate it, as we shall see. They plead in favour of matai being the person “most skilled” in an activity, hence “master”. In the same direction, we have the indications in Pratt's dictionaries: words like mataisiva “the best dancer”, matai'oloa “the best good”, and last but not least: mataisau “a respectable word for tufuga” (see below).

Still, the first quotation and what we have seen in narratives by Samoans, as well as Krämer's own analysis, all these data give without any doubt the meaning of “head of family”. The question must remain open. Did that latter meaning evolve, and when, from a first meaning exclusively linked to the craft masters' activity—as Pritchard's description as well as the existence of several compound words would suggest? Or did the general meaning of being the ‘first’ hence ‘the head of one's group’ appear as such and apply itself to heads of family as well as to craft masters? Might the semblance of a developmental sequence be only that Krämer's book is much more detailed than all preceding ones, which were content to describe high chiefs' existence and did not bother to deal with local heads of families?

- 171
Matai-tufuga in Missionary Dictionaries

Why do we not find the meaning of matai as ‘(master) craftsman’ in the missionary dictionaries? We know it existed in Samoa, from data present in Pritchard and Krämer (above). My only guess is that the situation in 1850-1890 was what it is today. A matai as a craft master was also head of his family, as Krämer says for the period he is observing. The missionaries' dictionaries missed this point, which is corrected only in Milner's contemporary dictionary. Obviously, they had understood the word matai as meaning first of all “head of family”, and hence, in front of the word tufuga used for all craft specialists. They did not realise that in fact, the linguistic situation was syncretic and evolving. The matai as ‘(craft) master’, designated just as “matai” as well as “matai tufuga”, belonged to the old Polynesian meaning of the word matai ‘skilled, the best’ (see below). While the matai as ‘head of a family’ was the specific Samoan application of the meaning of this word.


Pritchard's and Krämer's writings seem to be the only ones mentioning the role of the matai. Both indicate the same role of craft “master”, although observations were made nearly 50 years apart, but Krämer indicates also the “head of family” meaning and gives us detailed explanations. These two authors have in common a certain interest in details which are not found elsewhere. For instance, if one wants to learn details and words of the marriage ceremony, especially for the rarely described moment of ritual defloration of the bride, one can find them only in the texts of these two authors. Also, marriage and the marital status of daughters of families of medium or low rank are described only by them. Hence, we should not be too surprised to find only Pritchard and Krämer as witnesses of the existence of the people who were called matai in the 19th century.

Krämer writing nearly half a century after Pritchard gives us crucial indications of recent changes concerning the multiplication of family units (this is in his main text; in the lexicon, the “family” is presented as an original and unchanging social unit). For the time defined by his generalisations—which is often too vague, since he incorporates earlier writings and tries to show us how eternal-traditional “Samoa” was—“matai” is clearly the name of the function of family head, much like today, but with a fundamental distinction. Branches of an extended family are each headed by a matai, but Krämer speaks of the function of carrying the family name only for the matai of the “family” and not for the matai of the “sub-family” or “ituāiga”. We may then suppose that the notion of “matai” did not include necessarily what is called today a “title” and which is today the main feature of every matai's status.

- 172

This would explain why all master tufuga are also “matai”: the tufuga who leads the group engaged in building or tattooing—a group which is always a family group, the family of the tufuga—is, of course, the head of this family. This would also explain why all the rest of the 19th century literature, apart from the dictionaries, and apart from the brief mention of craft master by Prichard, ignores this notion—or, more precisely, does not bother describing it. Europeans dealt with high chiefs (main family in the village or in the district) and with their orators, and not with headmen of subgroups. Krämer was the first whose occupation was not to deal with local authorities for commercial, evangelical or political purposes, but to interview for hours and even days chiefs and orators (see his Preface, vol. I) about the rules and obligations of status positions and life-cycle events.

But even if every matai was already a title-holder in those years, i.e., a person carrying a transmitted specific name of his subgroup's founder, that title was significant only within the fono of that village. Indeed, the main transformation between the 19th and the 20th centuries, at least for what became in 1900 Western Samoa (“German Samoa” then “Western Samoa” in 1920), was the following: from (i) a system where each village nu'u was dependant on regional hierarchies at least at the district itūmālō level, the word implying that it is always a result of war alliances and victories even if the shift of allegiances was frequent and easy, to (ii) a system where a central government became established, where internal wars were declared illegal and effectively ended. This centrality had the major effect of placing all the villages more or less at the same level, as comparable units in relation to the central power. Consequently, the difference between local matai and supra-local ali'i and tulāfale became less and less perceptible.

This transformation occurred mainly between the 1870s and 1910s. In 1873-75, the first central Constitutions and central Parliament were established, through complex dealings with the settler community, for whom this was a first step towards an attempt at controlling the Samoan communities. In 1903, this was achieved by German colonisation (started in 1899) through the establishment of a central Land and Titles Court. This is obviously the period that should be looked at, especially through German archives, to see when and how the word matai became actually predominant in the official discourse. 55

The Word Ma(a)tai

The word matai is certainly not unique to Samoa. Pollex (1999), the computer file of Proto Polynesian compiled by Bruce Biggs and Ross Clark, - 173 has *maatai with reflexes listed for Rennell-Bellona, Samoa, Tahiti, Tikopia, Tokelau and Tonga (entry dated 1994):

*0 7/9/94.
REN Matai/sau.:Expert craftsman, tattooer.
SAM Matai.:Titled head of extended Samoan family; master.
TAH Maatai.:Skilful, knowing.
TIK Matai/tangata.:Leading man (Fth).
TOK Matai.:Male head of a cognatic descent group; master, headman, boss.
TON Maatai.:To be expert.

Actually the number of languages listed is even larger: there is also East Uvean, East Futunan and probably Marquesan. We shall look at this in more detail, going back to each dictionary. 56

For Tonga, Churchward's dictionary (1959) has: “mātai, a.v.i., to be expert, to be very clever or skilful (i, at).” Baker's dictionary (1897) has: “matai: s. a clever one; the best one; a single one…. matai: a. clever.”

For Tikopia, Firth's dictionary (1985) has:


1: leading element (pcw mata 1; 57 matai tangata leading man; te matai o tangata the best of men; matai tusi o te nuanga, fakamailonga te renga e sisi initial inscribing of the turmeric extraction, sign of the turmeric being hung up; matai toki leader of the adze, i.e. mark made with charred stick on topstrake and hull of canoe […] in song cycle, applied esp. to initial leading song. 58

For Rennell-Bellona, Elbert's dictionary (1975) has: “matai/sau: expert craftsman, tattooer (PPN matai).”

For Tokelau, the authors 59 of the Tokelau Dictionary (1986) have: “matai, n. [Sam matai]. 1. Recognized male head of a cognatic descent group. 2. Master, headman, boss. v. Be a matai.” Two examples are given for this verbal use. The first one only says “he is the matai of our kāiga”, but the second one refers to a modern job: “E i ei te vaitaimi nae matai ai ia i te gāluega, there was a time when he was the headman on the job.” The authors add that “This word is borrowed from Samoan but has somewhat different connotations in Tokelau”, without being more precise. The “borrowing from Samoan” refers to the matai as head of a family, since ethnohistorical research on Tokelau shows clearly that in the past, as preserved in oral literature, - 174 there were only aliki as chiefs (Huntsman and Hooper 1996). The “different connotation” obviously refers to the second definition which looks different from the contemporary meaning of the Samoan word matai ‘chief’. But when one considers that, in fact, one of the early Samoan meanings could have been ‘the best, master’ in any context (see below), besides the specific meaning of ‘head of family’, this second example of “master, boss” seems to fit perfectly. Thus, if this word is a borrowing, it is not a very recent one. And if it is indeed borrowed, it reinforces the hypothesis that, some time ago in Samoa, the Samoan word matai meant ‘master, etc’ In contemporary Samoan, the meaning of ‘boss’ in modern contexts could not be assumed by the word matai since the latter word had become specialised as “chief as title holder”. It is then the word pule (‘authority’, and ‘person in charge of the—’) which was and is used. But in Tokelauan apparently, where no evolution of this sort has occurred, the word had room to adapt its old (and presumably Proto Polynesian) meaning of ‘best, first’ to modern contexts, like being the ‘boss’ of a job. 60

For Uvea-Futuna, Rensch's dictionaries (1984, 1986) have for both languages: “mātai: maître-ouvrier; cf. tufuga.” Bataillon's dictionary for Uvea (1932; data of mid 19th century) has: “mātai (T. de Futuna), s. voir tufuga.

All the former examples are from Western Polynesia. For the confirmation of a Proto Polynesian meaning, it is useful to have a Samoan-Tongan correspondance, but we cannot exclude a borrowing from one to the other. That is why it is also important to find our word in at least one language outside Western Polynesia.

For Tahiti, the missionary dictionary (Davies 1851; data are from the 1830s and mainly from Orsmond) has: “Mataī: skilfulness, dexterity; skilful, knowing; skilfully.” 61 Jaussen's dictionary (1969, data of mid 19th century) has: “matai: dextérité, adroit.” 62

Thus, the meaning expert and/or leader, being the best and/or the first seems to be the shared meaning, which would then go back to some early stages of Polynesian languages. 63 When Tongan, Samoan (or another western language of the non-Tongic branch) and a language of Central-Eastern Polynesian are in accordance, we know that the base has a good chance of being Proto Polynesian or at least to go back to before the colonisation of Eastern Polynesia, more than 1000 years ago.

Samoan Language Again: Compound Words with Matai-

With this meaning in mind, one immediately notices the presence of some compound words based on matai- in the Samoan language. Their meaning makes it evident that the base has indeed the meaning of ‘the best, the first’ - 175 and has nothing to do with any mata-. In Milner's dictionary for contemporary Samoan, we find:

  • mātaisau, polite word for craftsman, carpenter
  • mataitōga, the most valuable fine mat (out of a collection or number of such) [in the example quoted by Milner, it is given to the senior orator; tōga=‘fine mat’]
  • mataitū, verb, head, direct, govern (the performance of a collective work) [=‘to stand firmly’].

For the 19th century, in the four editions of the missionaries' dictionary (Pratt), the list is longer:

  • mātaisau and mataitōga [given with a similar meaning in all editions]
  • matai-tu, the chief carpenter.

But there is also:

  • mataisiva, the best dancer in a night dance [appears in the 2nd edition and afterwards; siva=‘dance’]
  • mataitagata, a good-looking man [ibid.; tagata=‘man’]
  • mataitamaita'i, the mistress of a family [ibid.; tamaita'i=‘lady’, a respectful word for addressing or referring to any female]
  • matai' oloa, the best article amongst a lot of goods [all editions; 64 'oloa=goods, valuables].

This list of compound words based on matai- clearly share a common root meaning: ‘being the best/the first’. We can now be sure that matai is a root in itself. Thus, it has nothing to do with mata and the numerous compound words based on mata- 65 which are of the type (Milner 1966:135): mātagōfie ‘beautiful’ [“pleasant to be looked at”], mātagā ‘ugly’, mataamau ‘be generous’, matamuli ‘be shy’, mataala ‘be watchful, alert’, matai'a ‘have a good eye for fish’ [i'a ‘fish’], etc.—and, semantically close to our topic (but not etymologically), mata'āiga ‘be always ready to help one's family’.

“A Leader for Samoa!” Se Matai for Samoa!

Besides the compound words, we have very solid proof that, in 19th-century Samoan, the word matai even used alone could mean “leader” in a general way, apart from its (old?) applied meaning of ‘(craft) master’ and its (new but certainly pre-contact?) meaning of ‘family head’.

Tupua Tamasese Titimaea, before dying (1891), addressed to his people his will: his own son Lealofi should be bestowed with these names. He also - 176 said to his political body, Tūmua—the group of influential orators of the two eastern and western districts of Upolu—that, for the time being, they should follow the government established by the other side, Malietoa Laupepa, but that, if Malietoa did not respect the Tamasese side enough, they should take the power and nominate Lealofi. It was a time of rivalries for the “kingship” (called as such by the Europeans who had been trying since 1870 to have a central stable power established). Shortly after, in 1898, Malietoa Laupepa died. A high orator addressed some of the “chiefs and tulafale” of the Tamasese side who had entered in Malietoa's government, and asked them to recognise now Lealofi, using the following words: Fili mai ou Tumua se matai mo Samoa (Choose your Tumua a Leader for Samoa). 66

The event took place only 20 years after the very first attempts to create a centralised government in the Samoan archipelago. These attempts did not follow the traditional idea that an ali'i who gathers enough of the main regional titles is considered as the tupu (usually translated as “king”; literally the “emergence”, here the emergence from a general agreement, obtained by peace or war). For the first time, the idea was that the various factions should choose one person who, beyond the people attached to his own family title, could also be viewed by other large families as a leader of a unified country. 67 My hypothesis is that this is the reason why the call of this orator has been formulated with the word “matai”: se matai mo Samoa.

Conclusive Hypothesis on Samoan History

The only conclusion of the linguistic comparative enquiry then is that, at some point in its evolution, Samoan language had applied to all family leaders the old word matai which meant ‘the best in his specialised activity’. As the ethnographic literature of the 19th century dealing with Samoa shows, the matai were then the heads of those family groupings (called au 'āiga but also just 'āiga) that were parts of an extended network ('āiga Sa + name), based on kin and on land affiliations and headed by an ali'i. At that point, the position of a matai was still clearly differentiated from the position of the ali'i, at least those ali'i for whom a whole regional network of villages was in each case their “family” 'āiga.

Some questions are left unanswered. Of course, this does not preclude the possibility that even the “high chiefs” were called matai in relation to their own local 'āiga. But then, we must acknowledge two different Samoan concepts of “chief”, one as a “leader” matai of his group because he has been chosen to be the embodiment of the ancestral name of that group, another as a “high chief” ali'i (pa'ia) in the sense that the name carried by - 177 him as a title is recognised way beyond his local 'āiga and is thus considered, by many, as “sacred”. All ali'i would then have been also matai, but not all matai would have been able to be ali'i.

The question also remains open whether all the matai who were not ali'i were then tulāfale or if the tulāfale were only some family leaders specifically linked to an ali'i title, while other matai were just… matai, just leaders of their own families. But then, we must assume that these latter matai were not “titled”, since all information available converges to make us think that the founding names transmitted as titles were and still are either “ali'i” or “tulāfale” type of names, without any third category of titles.

The evolution continued, largely through historical conflicts and dealings with the European settlers and, later, colonisers and administrators. It produced a kind of general levelling of the old hierarchies in relation to a new centralised power and to the idea of a Parliament where representatives of various villages could sit. Gradually, it became less pertinent that a representative was an ali'i or a tulāfale. This levelling brought the notion of “matai” to the most general plane of consideration. All high chiefs, other dignitaries (such as high orators) and ordinary heads of sub-level family groups became merged into this new and unique category of “being a matai”. From the 1930s, the system became one where all chiefs and orators were then viewed first in their capacity of family leaders matai; only a secondary distinction still contrasted those matai whose “title” is an “ali'i title” and those whose title is not (and is then a “tulāfale title”).

From a hierarchy ali'i >tulāfale> matai, the system went into a single class of “matai” subdivided in two parts: matai with an ali'i-title and matai with a tulāfale-title. This contemporary situation seems to make of Samoa an exception and an aberrant case within Polynesia. The linguistic comparison and the historicised reading of the ethnographic literature attempted here, joined to a set of hypothesis on the transformations of Samoan politics (Tcherkézoff n.d.a, n.d.b), provides a possibility of coherent comparison.

Apparently, the transformation occurred between 1870 and 1910. Its possibility was prepared by a first levelling owing to missionary influence, but it took place in relation to: (i) the new system of centralised politics and (ii) the fact that this new system, induced for a great part by Europeans who were preparing—and soon after administering—colonisation, was built on the new idea of “political representation” of basic groups by recognised or elected “leaders”. This idea blurred the old distinctions between ali'i and non-ali'i and made way for a new encompassing notion which was not dependant on the ali'i/tulāfale opposition and which was even able to include this distinction as a simple subdivision.

- 178
Conclusive Hypothesis on Ancestral Polynesian Society

The category of matai seems then to have had quite a different application than the category of ali'i (and cognate words) designating the “chiefs”. A Polynesian ali'i/ariki, etc. was not a chief because he was “the best” in some context. He held the first status in his group (positions taken in the meeting house, respect paid to him, etc.) because he held a position defined as the locus where any individual made to “sit” there (nofo) becomes the embodiment of the founding ancestor and of his powers. 68 While a matai, if we generalise from the various Polynesian lexicons listed above, was (and is in several languages) the best in his activity, like a master in a craft activity. This is the reason why, in Samoan, the word designated the master tufuga, and it is also the reason why, in compound words, this base brought the meaning of ‘best’ within a set of things or people.

Such a conclusion, reached through a consideration of the Samoan context once placed within a Polynesian linguistic comparison, fits within a further step in a comparison provided by Kirch and Green (forthcoming), through their hypothesis about the two levels of leadership in Ancestral Polynesian systems. After making a wide comparison of Polynesian social and linguistic systems, the authors come to the conclusion that the leaders of the *kainanga type of social group (in short, a clan and/or an area of authority including land and people) were always the *qariki (Samoan reflex: ali'i). But there was also a type of ‘elder’, probably leader of the *kaainga type of social group (basic kin-and/or-residential group and its premises and land): 69 a more “informal” type of leadership, “men who served as leaders of family-based kin groupings, and thus held a standing in the community at large”, “one who oversaw the use and allocation of the *kaainga's land and resources”. These “elders” are contrasted to the “more formal and hereditary category of ariki, who occupied a higher-level of chiefly rank”. The hypothesis of the existence of these local leaders is based on the systematic appearance, besides reflexes of *qariki, of another linguistic category to designate leadership. 70

Moreover, what seems to me quite revealing is the great variety of words used for this latter category, in contrast to the singularity of *qariki. Kirch and Green (forthcoming) note: several cases of reflexes of *fatu (like the patu of Niue), but also the tuupele of Pukapuka, the pule kainga of Futuna, the matai of Samoa, the ra'atira of the Society Islands, the rangatira of New Zealand Maori. This contrast might also be correlated to the hypothesis suggested by the Samoan context, namely that the type of authority of the more localised leaders was of a material type, distinguished thus from pure sacredness (pa'ia). In material achievement, the possibilities are various, hence the variety of words for these “leaders”. In Samoa, it might have - 179 taken the form of a craft expertise. On the contrary, the ali'i/ariki/ eiki were everywhere, as we know, the embodiment among humans of the great founding gods—and the gods do not work. 71

Thus, this distinction is of a specific kind, and may be quite specific to Polynesia. 72 It is neither the distinction between the “Polynesian” chief who fills a powerful position pre-defined versus the “Melanesian” self-made leader (“Big-man”) who attracts followers around him (Sahlins 1968), 73 since the matai does not create the family grouping of which he becomes the head. Nor is it the ascribed/achieved distinction, since the Samoan high ali'i (regional authority) was not chosen by the ruling orators solely on predefined kinship or rank rules (the rule of primogeniture or of the first collateral line was not necessarily followed in Samoa, and various personal qualities, which had to be displayed before hand by the incumbent, were discussed, as is still the case today for all matai selections).

The distinction of the levels of leaders is completely different. First of all, it is a hierarchy of the territorial extension of authority (regional/localised), as illustrated by the Samoan case and by the other cases mentioned by Green and Kirch (forthcoming). Secondly, in my interpretation, it appears to be the contrast between a sacred chief who only “sits” nofo, in order to be illuminated and wrapped around by the ancestral or godly power mana, and a leader who “works”, “does” things and “makes” things happen. This latter kind of contrast is widely known from Western Polynesia in various contexts: kinship (sister/brother), 74 gift-giving (sacred cloth/food), 75 assymetrical system of greetings, 76 the whole opposition between kinship relationships and sexual commerce. 77 We can add to this list the relationship between the Samoan ali'i and the orator tulāfale. 78 This brings us back to the crucial question of the equivalence or distinction between matai and tulāfale in early 19th century Samoa. But, on that question, all available data seems unable to deliver a clear answer.

The notion of the “Polynesian chief”, usually apprehended as simple and unitary because of the contrast, in the context of all Oceania, with leaders of the Big-man type in some Melanesian societies, appears to be a much denser notion. The ali'i/matai distinction in old Samoa is only one of the many oppositions which makes this notion more complex.


This paper has benefited from the discussions in Canberra and in Auckland in March 1999 following lectures I gave on the topic. All my thanks to Judith Huntsman, and to Margaret Jolly and Darrell Tryon, for their invitations. As is evident in the last section, Roger Green presented me with the possibility of extending my comparisons, and I am very grateful to him and Patrick Kirch for giving me some pages of what was then their manuscript on “Ancestral Polynesian Society”. Thanks - 180 also to Grace Meagher (EHESS, Marseilles) for Englishing my first draft and to Judith Huntsman for the long and attentive work required in the final editing.

- 181 Page of endnotes

- 182 Page of endnotes

- 183 Page of endnotes

- 184 Page of endnotes

- 185 Page of endnotes

- 186 Page of endnotes

- 187
  • Baker, Shirley W., 1897. An English and Tongan Vocabulary, also a Tongan and English Vocabulary with a list of idiomatic phrases; and a Tongan Grammar. Auckland:[no publ.].
  • Bataillon, Mgr, 1932. Langue d'Uvea (Wallis). Paris: Libr. orientaliste Paul Geuthner.
  • Brown, George, 1910. Melanesians and Polynesians: Their Life-Histories Described and Compared. London: Macmillan.
  • Buck, Peter, (Te Rangi Hiroa), 1930. Samoan Material Culture. Honolulu: B.P.Bishop Museum Bulletin, 75.
  • Calvert, James (ed.), 1872. A Fijian and English, and an English and Fijian Dictionary (David Hazelwood, 1850). London: Calvert.
  • Chowning, Ann, 1996. POc *mata: How many words, how many meanings? In J. Lynch and P. Fa'afo (eds), Oceanic Studies: Proceedings of the First International Conference on Oceanic Linguistics. Canberra, Australian National University, Research School of Pacific and Asian Studies (“Pacific Linguistics Series C”, 133): pp.47-60.
  • Churchward, Maxwell, 1959. Dictionary Tongan-English, English-Tongan. Nukualofa: Government Printing.
  • [Davies, John] 1851. A Tahitian and English Dictionary. Tahiti, LMS Press (reprint: New York: AMS Press, 1978).
  • Dordillon, René Ildefonse Mgr, 1931. Grammaire et dictionnaire de la langue des îles Marquises. Paris: Institut d'ethnologie.
  • Douaire-Marsaudon, Françoise, 1998. Les premiers fruits. Parenté, identité sexuelle et pouvoirs en Polynésie occidentale. Paris: Ed. du Centre national de la recherche scientifique et de la Maison des sciences de l'homme (“Les chemins de l'ethnologie”).
  • Elbert, Samuel H., 1975. Dictionary of the Language of Rennel and Bellona. Copenhagen: The National Museum of Denmark.
  • Firth, Raymond, 1957 [1936]. We the Tikopia. London: Allen & Unwin.
  • ——1985. Tikopia-English Dictionary. Auckland/Oxford: Auckland University Press/Oxford University Press.
  • Grattan, F.J.H., 1985 [1948]. An Introduction to Samoan Custom. Auckland: R. McMillan (reprint; lst ed. 1948).
  • Green, Roger, 1989. Commentary [no title] to N. Thomas, The force of ethnology: Origins and significance of the Melanesia/Polynesia division. Current Anthropology, 30: 35-36.
- 188
  • Handbook of Western Samoa, 1925. Wellington: Government Printer.
  • Huntsman, Judith & Antony Hooper, 1975. Male and female in Tokelau culture. Journal of the Polynesian Society, 84:415-30.
  • ——1996. Tokelau: A Historical Ethnography. Auckland: Auckland University Press.
  • Jaussen, Tepano, 1969. Grammaire et Dictionnaire de la langue tahitienne. Paris: Musēe de l'Homme, Société des Océanistes (“Publication”, 22) (5th ed., revised by Mgr. Mazé and R.P. H. Coppenrath).
  • Kaeppler, Adrienne L., 1999. Kie hingoa: Mats of power, rank, prestige and history. Journal of the Polynesian Society, 108 (2): 168-232.
  • Kirch, Patrick V. and Roger C. Green, forthcoming. Hawaiki: Ancestral Polynesia, An Essay in Historical Anthropology. Cambridge: Cambridge University Press.
  • Krämer, A. 1902. Die Samoa-Inseln, v.I. Stuttgart: Schweizerbart.
  • ——1995 [1903]. The Samoa Islands: An Outline of a Monograph with Particular Consideration of German Samoa, v.II. Honolulu: University of Hawaii Press. (Original edition: Die Samoa-Inseln. Eintwurf einer Monographie mit besonderer Berücksichtigung Deutch-Samoas, v.II. Stuttgart, E. Schweizerbart).
  • Lutali, A.P. and W.J. Stewart, 1974. Legal aspects of the matai and land tenure system in the territory of American Samoa. Samoan Pacific Law Journal, 2:110-144.
  • Marck, Jeff, 1999. Polynesian Language and Culture History. Unpublished PhD thesis. Canberra: ANU, Research School of Pacific and Asian Studies (Department of Linguistics).
  • Meleisea, Malama, 1987. The Making of Modem Samoa. Traditional Authority and Colonial Administration in the Modern History of Western Samoa. Suva:University of the South Pacific, Institute of Pacific Studies.
  • ——1992. Change and Adaptations in Western Samoa. Christchurch: University of Canterbury, Macmillan Brown Centre for Pacific Studies.
  • ——1995. “To whom gods and men crowded”: chieftainship and hierarchy in Ancient Samoa. In Judith Huntsman (ed.), Tonga and Samoa. Images of Gender and Polity. Christchurch: University of Canterbury Press (Macmillan Brown Centre for Pacific Studies): pp.19-34.
  • Milner, George, 1966. Samoan Dictionary: Samoan-English, English-Samoan. London: Oxford University Press.
  • Neyret, J., MS. 1935. Fijian-English Dictionary. Typed MS., London: SOAS Library.
  • Pawley, Andrew 1985. Proto-Oceanic terms for ‘person’: A problem in semantic reconstruction. In V.Z. Acson and R.L. Leed (eds), For Gordon H. Fairbanks. Honolulu: University of Hawai'i Press, pp.92-105.
  • Pollex, 1999. Proto-Polynesian Lexicon. Computer file (regularly updated and expanded) compiled by Bruce Biggs and Ross Clark.
  • Pratt, George, 1862. Samoan Dictionary: English and Samoan, and Samoan and English, with a short Grammar of the Samoan dialect. “Samoa: Published at the L.M.S. Press”.
  • ——1876. Grammar and Dictionary of the Samoan Language (“Second Edition, edited by Rev. S.J. Whitmee”). London: Trūbner & Co.
- 189
  • ——1893. A Grammar and Dictionary of the Samoan Language, with English and Samoan vocabulary (“Third and Revised Edition”). London: The Religious Tract Society for the London Missionary Society.
  • ——1911. Pratt's Grammar and Dictionary of the Samoan Language (“Fourth Edition”, “Preface” by J.E. Newell). Apia: Malua Printing Press [reprinted 1960, 1977].
  • Pritchard, William T, 1968. Polynesian Reminiscences: Or, Life in the South Pacific Island. London: Dawsons (1st ed. 1866).
  • Rensch, Karl H., 1984. Tikisionali o Fakauvea-Fakafalani. Dictionnaire Wallisien-Français. Canberra: Australian National University (“Pacific Linguistic Series C”, 86).
  • ——1986. Tikisionali o Fakafutuna-Fakafalani. Dictionnaire Futunien-Français. Canberra: Australian National University (“Pacific Linguistic Series C”, 90).
  • Sahlins, Marshall, 1968. Poor man, rich man, big man, chief: Political types in Melanesia and Polynesia. In A.P. Vayda (ed.), Peoples and Cultures of the Pacific, New York: The Natural History Press: 157-176 (1st ed. in Comparative Studies in Society and History, 1963, 5 (3): 285-303).
  • ——1989. Commentary [no title] to N. Thomas, The force of ethnology: Origins and significance of the Melanesia / Polynesia division. Current Anthropology, 30(l):36-37.
  • Schoeffel, Penelope, 1979. Daughters of Sina: A Study of Gender, Status and Power in Samoa. Unpublished PhD thesis. Canberra: Australian National University.
  • ——1995. The Samoan concept of feagaiga and its transformation. In J. Huntsman (ed.), Tonga and Samoa: Images of Gender and Polity. Christchurch: University of Canterbury Press (Macmillan Brown Centre for Pacific Studies), pp.85-106.
  • Schultz, Erich, 1911. The most important principles of Samoan family law, and the laws of inheritance. Journal of the Polynesian Society, 20:43-53.
  • Tcherkézoff, S., 1993. The illusion of dualism in Samoa: “brothers-and-sisters” are not “men-and-women”. In Th. Del Valle (ed.), Gendered Anthropology. London, Routledge & Kegan Paul, pp.54-87.
  • ——1995. L'autocar à Samoa ou la hiérarchie au quotidien. Gradhiva, 18: 47-56.
  • ——1997a. Culture, nation, société: changements secondaires et bouleversements possibles au Samoa Occidental. Vers un modèle pour l'étude des dynamiques culturelles. In S. Tcherkézoff et F. Douaire-Marsaudon (eds), Le Pacifique-sud aujourd'hui: identités et transformations culturelles, Paris, CNRS Editions (“Ethnologie”), pp.309-373.
  • ——1997b. Le mana, le fait “total” et l'“esprit” dans la chose donnée. Marcel Mauss, les “cadeaux à Samoa” et la méthode comparative en Polynésie. Anthropologie etSociétés (numéro spécial “Comparaisons régionales”, ss dir. de Xavier Blaisel et Jean-Claude Muller), 21 (2-3): 193-223.
  • ——1998a. Is aristocracy good for democracy? A contemporary debate in Western Samoa. In J. Wassmann (ed.), Pacific Answers to Western Hegemony: Cultural Practices of Identity Construction. Oxford: Berg (“Explorations in Anthropology Series”), pp.417-434.
- 190
  • ——1998b. Identités en mutation aux Samoa: la terre et le développement. In D. Tryon and P.de Deccker (eds), Identités en mutation dans le Pacifique à l'aube du troisième millénaire. Bordeaux University Press (Université Michel Montaigne-Bordeaux 3, Centre de recherche sur les espaces tropicaux CRET, Sērie “Iles et Archipels”, 26), pp.111-157.
  • ——1999. Qu'est-ce qu'un acte sexuel au Samoa Occidental (Polynésie)? In Ph. Descola, J. Hamel et P. Lemonnier (eds), La production du social. Autour de Maurice Godelier. Paris, Fayard, pp.369-387.
  • ——2000. Le mythe occidental de la sexualité polynésienne: M. Mead, D. Freeman et “Samoa”. Paris, Presses Universitaires de France (“Ethnologie”).
  • ——n.d.a. “Are the Samoan chiefs matai ‘out of time’?” Tradition and democracy: contemporary ambiguities and historical transformations of the concept ‘matai’. In E. Huffer and A.So'o (eds), Governance in Samoa. Canberra/Suva: Australian National University (National Centre for Development Studies, “Asia-Pacific Serie”)/University of the South Pacific (Institute of Pacific Studies).
  • ——n.d.b. La Polynésie et l'Occident. Du malentendu au dialogue anthropologique (économie, politique, sexualité).
  • ——n.d.c. L'invention de la ‘Polynésie’ et de la ‘Mélanésie’. Du présent au passé: le racisme derrière la géographie.
  • ——n.d.d. Multi-culturalism and the building of a national identity: “Samoans” and “Europeans” in Samoa (1899-1962). In P. deDeckker & M. Godelier (eds), Les Identités du Pacifique—Pacific Identities. [Procedings of the Symposium held in Noumea, July 1999], Nouméa, University of New Caledonia and French Delegation to the Community of the Pacific Countries.
  • Tokelau Dictionary, 1986. Apia: Office of Tokelau Affairs.
  • Vaai, Alaelua Asiata Vaalepa Saleimoa, 1996. The Rule of Law and the Faamatai: Legal Pluralism in Western Samoa. Unpublished PhD thesis, Canberra: Australian National University, Research School of Pacific and Asian Studies (Department of Pacific and Asian History).
  • Violette, Louis-Thēdore, 1879. Dictionnaire Samoa-Français-Anglais et Français-Samoa-Anglais, précédé d'une grammaire de la langue Samoa. Paris: Maisonneuve.
  • Wilson, William H., 1982. Proto-Polynesian Possessive Marking. Canberra, Australian National University, Research School of Pacific and Asian Studies (“Pacific Linguistics Series B”, 85).
1   I heard it often, in the 1980s and early 1990s, in two kinds of discussions with various people from Western Samoa, concerning the fact that each village is governed by its council of matai and that only matai can be candidates for Parliamentary elections.
In this essay, matai, as a Samoan lexeme, will be italicised, except in the following usages: (i) as it is discussed as a sociological concept by Samoans, i.e., the notion of “matai”, (ii) as used by Samoans in English discourse or text, i.e., “our matai titles are…”, and (iii) in citing the word as it is used by ethnographers, i.e., “Schultz used ‘matai’ to mean…”.
2   From 1990 in Western Samoa, every adult can vote to elect Parliamentary candidates, while only matai had the right to vote in previous decades (see Tcherkézoff 1998a).
3   George Milner, the author of the modern Samoan-English dictionary (1966), indicated at the end of the entry “matai”: “pcw mafa”. Lutali and Stewart (1974) proposed that matai “may have developed from a combination of the Samoan mafa (eye) and ‘i ai’ (to or toward) so that matai might be derived from the concept of looking towards or up to another” (quoted in Vaai 1996:33). Malama Meleisea writes (1987:7, 1992:15) that the word “comes from ‘matai ai’ which has the connotation of ‘being set apart’ or ‘consecrated’”. I heard this same proposition from a Samoan student in a university seminar (Auckland 1999).
4   See Tcherkézoff (1993, n.d.b: Ch. 6). The question of the ali' i/tulāfale distinction in Samoan history interacts with the history of the category of “matai” but the data are confusing. One hypothesis is that all the matai were tulāfale at the time when “matai” was distinguished from “ali'i”. Another, which seems more probable, is that matai were all households heads but only some of them were titled as “tulāfale”.
5   But also in the modern constitutional law (see Tcherkézoff 1998b).
6   The possessive marker used in those cases is always of the “-o- class” in contrast with the “-a- class” (for instance lo'u in contrast to la'u): a linguistic indication that the locutor is not at the origin of the possession of the things referred to— “my land, my house, my parents, my body, etc.” are -o- class, “my clothes, etc.” are -a- class (see Wilson 1982).
7   Malama Meleisea (1987, 1995) has asserted that the supreme chiefs of that time were only the ali'i pa' ia. I am pursuing a complementary enquiry, on the linguistic origin as well as on the historical emergence and transformations of the notion of “matai”.
8   My enquiries started in 1981.
9   In the 19th century, the list of various craft masters designated as “matai” was much longer (see Krämer 1995:97 where some twenty different ones are listed).
10   To these meanings given by Milner, we can add another one for a word matai, recorded by Peter Buck in 1927 during his study of the material culture. About fishing gears and lines, Buck writes: “The fine line which is continuous with the snood is 9 or 10 feet long and is called the matai afo, which really means the head or leader of the fishing line. This is joined to the afo (line) which consists of five-ply sennit braid about 22 feet long” (Buck 1930:513). This meaning is also found in Tikopian and in Marquesan (see below). Firth, like Buck, considers that it is the same word than matai as ‘leader’, but of course we cannot have any certainty. At least, it shows us that Buck, when enquiring in Samoa, had in mind a meaning of the word matai which was primarily ‘leader’—and we will see the importance of that notion.
11   Because it mentions the retirement of Colonel Tate in 1923 (p. 16).
12   The text continues and describes the different administrative functions established by the new Administration (in continuity with the previous one, the German colonial administration): pulenu'u or “village headmen”, magistrates, plantation inspectors, Land and Title Commissioners, and “at the other end,… the faipule and fautua: the former powerful officers who are members of the Native Council and act in the capacity of District Governors; the latter, who represent the old clans of Tupua and Malietoa, holding nominal office, but who do not act in any official capacity.” (p.45); on the history of the Native Council and of the Fautua, see Tcherkézoff n.d.d.
13   Six months all together, two in Western Samoa, mainly in Savaii, two in Tutuila, two in the Manu'a group.
14   His informant on that was Fepulea'i Ripley of Tutuila.
15   For further information concerning this Samoan logic of reversals between ali'i and tulāfale and also between sisters and brothers, see Tcherkézoff 1993.
16   The fact that abdominal wall is given to the taupou (i.e., women of the family who are there as daughters and sisters: teine o le nu'u) and the rear muli is given to women (i.e., to spouses) is linked to other symbolisms: the whiteness and softness of the abdominal wall corresponds to the dignified and preserved state of the teine; the opposition front/back distinguishes teine and spouses, at least when spouses, supposed to be always from a different village than the husband's village, are living in the husband's family—see the word faletua for the wife of a matai whose title is an ali'i title.
17   A review of this literature is being prepared.
18   As it was the aim in those Bernice P. Bishop Museum expeditions of 1920-1940, where research was defined as “salvage ethnology”: to get the most “traditional” data.
19   For instances in directing the procedure of title splitting between several incumbents for a succession to the same name, see Meleisea 1992:40.
20   See Meleisea 1987, Meleisea and Schoeffel 1987, Tcherkézoff n.d.b: Chs 2-3. This institution had an interesting future. With the New Zealand Administration, the Land and Titles Commission became openly “a permanent tribunal to arbitrate in matters of custom” (Meleisea 1992:40), a court for settling all disputes relating to titles. With independence in 1962, the Land and Titles Commission began to play a key role in politics through the national registration of matai names, particularly for the electoral system of contemporary Samoa.
21   See Meleisea 1987, 1992; Tcherkézoff n.d.a.
22   The word “sili” means ‘main, first, superior’; it can be applied in any context, even in a comparison between objects or ideas. Thus, for Schultz, a chief, even a high ali'i, is—or should be—nothing more than a matai heading other matai within one large “clan”. Solf's creation of a “Sili” position for Mata 'afa follows the same idea of levelling (Meleisea 1992: 30-33, Tcherkézoff n.d.a). The difference is that Solf called it “Ali'i Sili”, probably because, as the extension of “matai” to high chiefs was only recent, he felt it was still necessary to keep “Ali'i” for this new administrative title because of the rank of the incumbent (Mata 'afa).
23   The only difference is that the 1862 edition has: “a head of a family”, which is in all successive editions “the head of a family”.
24   The only one is “'O le ali'i o lo mātou nu 'u”, without translation (the meaning is: “the ali'i of our village”). On one hand, it indicates what we know very well: a chief's title is always rooted within a social community called nu'u (a village as a polity, defined as a sacred circle of extended families, each with their own title). The “village council” is made up of the titled heads of these various families. The village territory is made up of the lands (at least the founding sites) of these families. But this example could imply also that, often, there was only one ali'i in the village, or that most villages had, among the ali'i residing there, one ali'i who was the ali'i of the village, the main authority of the village.
25   Before the 1820s, contacts had been only at sea (Roggeveen in 1722, Bougainville in 1768), except for the short and tragic landing of La Pērouse in 1787 and the brief passage of Edwards (1791). The next arrival of an exploring expedition was Kotzebue's very brief visit (1825), contemporaneous with the first visits by whalers. Dumont d'Urville and Wilkes arrived only at the end of the 1830s, a period when missionaries were already settled (first contact by John Williams in 1830, permanent missionary settlement from 1836).
26   Krämer's book Die Samoa-Inseln (2 volumes) was published in 1902-3 in German and in 1995 an English translation by Th. Verhaaren was published, dating from 1979. The work includes the author's analyses and numerous Samoan narratives, presented in Samoan with a translation; I shall cite from the original edition of volume I, because my citations will be mainly from Samoan narratives, and volume II from the English translation. Where it is relevant I will insert the actual German words used by Krämer. All citations from vol. I are my translation (from German when it is Krämer's text; from Samoan when it is Samoan narratives, and disregarding Krämer's German translation which often contains errors). For citations from vol. II, the same procedure will be followed for Samoan narratives—for Krämer's text, I shall use the English published translation, with relevant German words inserted.
27   [zwischen Häuptlingen und Sprechern (ali'i and tulafale)]
28   [einem Häuptling von einem Nichthäuptling]
29   Krämer indicates in a note that he found this sentence in Turner.
30   “…signs by which to recognise the chief [Häuptling] of a village [des Dorfes] or of a family [oder einer Familie], and by that title [in den folgenden Zeilen] I should like to identify in the following lines solely the alderman, the matai, [nür den Ältesten, den matai] apart from title chiefs [TitelHäuptlinge] and of course of kings [Königen]….”
31   See Krämer 1995[1903]:108-109 (English translation of Verhaaren); German words added by me from the original edition, Krämer 1902:101.
32   [Familienoberhaupt (matai)]
33   [der freiende Häuptling]
34   [ob er matai ist oder sein]
35   [das Familienhaupt, welches den Namen der Familie trägt und über die Familie herrscht (siehe āiga)]. Citations from volume I, here and below, are all taken from the original German edition: page numbers refer to the original publication, translations are mine.
36   ['āiga “die Familie”, die Verwandtschaft. Das ā lang im Gegensatz zu 'aiga, 'aina Essen. Die āiga ist ein grosser Machtfaktor im samoanischen Staat, da von Grösse und Ansehen der Familie alles abhängt. Jeder Familie steht ein matai vor, der den Namen der Familie trägt, während die übrigen Mitglieder nur sozusagen Vornamen behalten, resp. andere Namen wie beim englishen Adel. Ist eine Familie seh gross, so zerfällt sie in Unterfamilien, dit ituāiga oder puiāiga heissen und auch ihre eigenen matai haben, die dann aber alle unter einem matai sili stehen. Āiga nennen sich auch gewisse Dörfer, die durch Verwandtschaft verbunden zu einem Grosshäuptling ganz oder geteilt zu stehen pflegen. So spricht man von den āiga des Malietoa (Stuebel p. 98), des Tuiaana p. 149 u. s. w.; āiga wird speciell auch Manono genannt (āiga o le tai p. 160).]
37   The status is not specified; he must be an orator because his speech is called by the next speaker “your fetalaiga”.
38   'a e nonofo le 'aufaipule i le toe tauati le tuāfono. The word 'aufaipule refers apparently to orators. We know that today the ceremonial expression 'o ali'i ma faipule is used to designate all the matai of a village. It seems to me that, today as in the 19th century, if ali'i are only called and designated as “ali'i”, the orators are called differently according to the context. In their capacity as “speaker” they are fetalaiga (the word applies to speech and, as a polite term, to the person speaking; base tala “tell”); in their capacity as adviser, support and sometimes deciding collective body for succession to regional titles, they are tulāfale said also tula (tula-a-fale: they are metaphorically the “base, support” of the house of the ali'i; tula: base for the mast in the old double canoes, base where sits the tamed pigeon, traditional pet of the high ali'i in the 19th century); in their capacity as looking to the execution of decisions made by the village fono they are faipule ‘to make, carry, the authority’; chiefs ali'i are sacred, they are defined by what they are, orators by what they “make”.
39   Ona fai atu lea 'o le pule. Apparently 'O le pule is the person directing this meeting of the faipule. See previous note.
40   40. Lauga: usually the word refers to the ali'i speaches, but it can apply to any formal speech.
41   The full text of these last sentences is; 'Ua 'uma la latou taumafataga, ona fai atu ai lea 'o le lauga a falenei i fafō: Fa' afogafoga maia 'o'oe le'aumaga 'o i'u nei'o le aofia, 'o le a tou silasilamai i ai. Ona laulau atu ai lea 'o 'upu 'uma na 'ua i luga. Ona tu'ua lea, 'ua alu le matai, 'ua laulau i lona aiga. My translation differs notably from Krämer's one.
42   'ua filifili ma lo latou āiga.
43   i lo' u fale 'o lo tatou āiga.
44   lo latou āiga uma.
45   Ona fai atu lea 'o le isi matai.
46   i lou maota. This is the special word for the house of an ali'i.
47   Ona fa' apea fo' i lea 'o le isi matai o le isi āiga.
48   …'o 'ie tōga a le ali' i ma lona 'auāiga….
49   Afio maia lau afioga!: specific address to ali'i.
50   'O susuga lava oe le agaiotupu.
51   In the original (Krämer 1902) as in the English translation (Krämer 1995), the Samoan text and its translation in German/English are of course presented separately.
52   “Head of a family” is proposed by the translator of Krämer in English; in another place, he has “Alderman” (see below). Krämer uses here only Āltester but in other passages (see below) has indeed “Familienhaupt” or “Familienoberhaupt”.
53   The incision of the prepuce for the circumcision—required for all boys.
54   The word : ‘to hit’—hitting the tattoo design with a point which had been dipped in the pigment.
55   The main problem is that the various consular despatches and legislative texts kept by consular offices—which can be easily consulted nowadays—use only European languages and mention only the word “chiefs”. The research would ideally concentrate on the few Samoan written papers which could be found (Gurr's material, etc., see note 66 below).
56   I checked nearly all dictionaries available; of course I could not check all ethnographies available.
57   As Milner and Meleisea did for the Samoan, Firth raises the question of the link with the base “mata”.
58   The entry continues: “2: plaited cord of 4-ply, used to attach hook to main line as a ‘leader’ … it breaks if hook caught in bottom and allows main line to be recovered [pcw matai 1].” In this second meaning, we find what Buck has noted for Samoan (see notes above).
59   Anonymous collective, see the introductions (many Tokelauans contributed; Ropati Simona was the official lexicographer; Robin Hooper actively participated; Judith Huntsman and Antony Hooper were closely associated with the project, etc.).
60   In the fascinating ethno-history of Tokelau, gathered and analysed by Judith Huntsman and Antony Hooper (1996), one can see that the whole history of the chiefly system is only made of aliki and faipule, not of “matai” (see Chs 5 and 7; see also Ch. 2, notes 5-6).
61   Of course, it also has numerous words based on mata- referring to eyesight and appearance, and it has another list of words based on another “matai-” which is a doublet of matangi/matani, ‘wind, air’.
62   In the Marquesas, we find the meaning already mentioned for Tikopia by Firth and for Samoa by Buck (see notes above): the cord attaching the hook to the main line. Dordillon's dictionary (1931) has: “matai: ficelle qui attache l'appât à la ligne.”
63   Leader as “expert”, as the “best”, hence the “first”, and not leader as the avant-guarde, as the front row of an armed group for instance; this latter type of leader or of leading action is usually denoted with reflexes of the Proto Polynesian base *taki in many Polynesian languages, including Samoan.
64   The only difference is that the 1862 edition spells the entry “mataioloa”.
65   At least it has nothing to do with mata- within the specific Samoan linguistic history. Whether there is, at earlier stages (Proto-Polynesian or Proto-Oceanic), a link or not between the two bases *mata- and *ma(a)tai is a different question that is not relevant to the topic discussed here. The most recent and complete synthesis on *mata at the Proto-Oceanic level (Chowning 1996) does open some possible semantic links (see pp.50,58,59).
66   The Samoan sentence and its translation is quoted by R.P. Gilson in his typed notes where he summarised in the 1950s the papers of Edwin Gurr, Judge in Samoa in the end of the 19th century (PMB 1003, reel 5 of Gilson papers, p. 123 about File 18 of Gurr's papers, section entitled by Gilson “Notes on Samoan Customs in reference to Names and Titles, MSS partly typed partly hand, probably Gurr's though Carruthers may have been involved”). For the historical events of this period, see Davidson 1967:61-67.
67   As we know, one year after the event, the power was neither in the hands of Lealofi for the Tupua Tamasese side, nor in the hands of the young Tanumafili for the Malietoa side, but in the hands of Papālagi powers (the German Kaiser for the Western islands and the U.S. Navy for the Eastern islands).
68   The question of the ali'i chiefs gaining status through victory at war is not different, since this kind of victory reveals that the victor had the sacred power or mana.
69   The brief formulations for the definition of the *kainanga and *kaainga are mine and are approximate. The most complete comparison of meanings in Polynesian languages of the two sets of reflexes derived from these two PPn roots is in Marck (1999: 241-247). Etymology does not help because these two roots are somehow doublets, even if the first is held to go beyond the level of Polynesian languages while the second is only Polynesian (Pawley 1985). Another difficulty is that a number of reflexes of *kainanga became restricted in meaning: they designate only the people who are the subjects of the high chief.
70   Kirch and Green, forthcoming, Ch. 8, section “PPN *Fatu, Leader of the *Kaainga” and the table “8.9. Cognate reflexes of PPN *fatu, Leader of the *kaainga” I quote with the authors' permission. Roger Green very kindly gave me a copy of these five pages from his and Patrick Kirch's manuscript: Hawaiki: Ancestral Polynesians. An Essay in Historical Anthropology.
71   This dualism of large territory leader (*kainanga: *qariki) vs. localised kin-groups' leader (*kaainga: patu, matai, etc.) is related in complex ways to the sacred/secular dualism that is more often used to characterise Polynesian systems (e.g., Tui and reflexes of *qariki vs. sau, hau; and also “strangerking” who becomes chief vs. chiefs of the land who become “orators” of the chief, such as, in Samoa, ali'i vs. orators tulāfale). Its elucidation may provide a new basis for the general study of diarchic configurations in historical Polynesia.
72   If we consider, like Pawley (see notes above), that the second level of group and leader (*kaainga) is a Polynesian addition to the first level already present in Oceanic languages (*kainanga). It could be that, for this case, “Polynesia” should be understood as “including Fiji” (see Tcherkézoff n.d.c for discussion of the history of Euro-centric boundaries in the Pacific). On one hand, we note that our word matai is quite present in Fijian: 19th century data have matai as ‘handicraftsman’, ‘a mechanic’ [examples of stonemason, blacksmith], ‘native carpenter’, matai sau (idem), matai ni ta waqa ‘a boat builder’, ai matai ‘the first of; first fruits…’ and also, as adjective, matai ‘clever’ (Calvert 1872, Neyret 1935). On the other hand, this matai as craftmaster does not seem to have evolved into any kind of leader of a regular local group.
73   “Polynesian” and “Melanesian” are between quotations marks because Sahlins had made it clear that, although Polynesia provided him with examples of the first type (“positional or office power”) and Melanesia with examples of the second type (“personal dominance”), his point was to oppose two political forms and neither two culture areas nor the too often invoked opposition between ascription and achievement (Sahlins 1989). Melanesia, at least, provides examples of the two kinds and many intermediaries. The major problem of course is that Polynesia is a relatively homogeneous culture area while Melanesia is not (Green 1989, Tcherkézoff n.d.c).
74   See Huntsman and Hooper 1975, Schoeffel 1979, Tcherkézoff 1993, Douaire-Marsauson 1998.
75   Sacred cloths versus cooked food and tools, hooks, canoes, etc.; see Firth 1936, Tcherkézoff 1997b, Douaire-Marsaudon 1998. In the Tongan case, the sacred cloths when given are designated as koloa ‘valuables, treasures’ while the other kind of gifts can be described as ngaue ‘(products of a) work’ (Kaeppler 1999: 170), even if the two notions are not of the same level and do not form a complementary opposition at any ceremonial level (Douaire-Marsaudon 1998: 134). Those sacred cloths given as “koloa” are “a fine art that creates valuables”; as such they are defined on a superior level and cannot be compared (either in a complementary or contrasted way) to “objects made by tufunga ‘craftsmen’” (Kaeppler, ibid.). We have seen, at least for the Samoan case, the proximity of the notions of tufuga and matai. The ali'i wrapped in a sacred fine mat is at a different level than the tufuga and—formerly—the matai.
76   Malo le galuega—Malo le tapuai! (see Tcherkézoff 1995, n.d.b: Ch. 5). Galue ‘work’ is the Samoan cognate of the Tongan ngaue.
77   At one level (sister-brother relationship), the sister encompasses the brother in her immovable sacredness. At another other level (male-female as sexual partners, already engaged or considered as possible partners), the male—only him—is said to “make” the act and is said to be able to “force” sometimes the female partner, because he is the “strong” side. But this “strong/weak” opposition is absolutely removed from the brother-sister level of relations, as long as the sister behaves as a “sister” and not as a “(sexual) female” (see Tcherkézoff 1993, 1999, 2000, n.d.b: Ch.7).
78   See Tcherkézoff 1993 and n.d.b: Ch.6 for the working of the two-levels relationship between the two (see also our commentary above on the division of the cooked pig in Buck's data).