- 335

In his book The Lapita Peoples Kirch (1997:217) says:

At the same time Kirch is aware that a word for areca nut is reconstructible for Proto Austronesian (p. 303, note 33).

Of course, from the fact that a term for areca nut is reconstructible for a proto-language it does not follow that the speakers of the language did chew betel (just as, for example, the existence of a word for ‘pot’ in a proto-language does not warrant the conclusion that the speakers of that language made pots). A richer set of reconstructions having to do with betel-chewing is necessary to be able to draw that kind of conclusion; and the richer the lexical field reconstructed, the sounder the conclusion. The purpose of this paper is to determine whether there is linguistic evidence of betel-chewing by speakers of Proto Oceanic. And, if one accepts the assumption that the Lapita people spoke Proto Oceanic, the conclusion reached for Proto Oceanic will have direct relevance to the question of whether the Lapita people chewed betel.

The findings of the present study provide an affirmative answer to the question posed in the title: speakers of Proto Oceanic did chew betel. The evidence presented here is primarily linguistic, but it is supported by archaeological evidence.

Besides betel-chewing there is another major use of a drug in the Oceanic region, namely kava-drinking. Even though the present study is concerned with betel-chewing, I will say a few words about kava in the concluding section, in particular concerning the temporal relation between betel-chewing and kava-drinking in the prehistory of the Oceanic area.

There are three main ingredients used in betel-chewing: areca “nut”, betel - 336 pepper and lime. The areca “nut” (often referred to as “betel nut”) is not, botanically, a nut: it is the endosperm of the seed of the Areca catechu palm (Purseglove 1972). Nevertheless, I will follow tradition and will speak of areca nuts. The betel pepper is a vine, Piper betle; normally, it is the catkin or the leaves that are chewed. The third ingredient lime, is usually made from coral or shell burned in fire and reduced to powder. I will use the term betel-chewing to refer to the masticatory use of these three ingredients together in a “quid”.

The genus Areca contains a number of other species besides catechu. While it is catechu nuts that are normally used in betel-chewing, the nuts of some other species of Areca may be used in times of shortage of catechu. Such nuts, however, are usually considered inferior, and are avoided if catechu nuts are available.

In addition to the three main ingredients—areca, betel and lime—others may be used. In the Oceanic area, tobacco is sometimes added to the betel quid. However, since I am concerned with reconstructing lexical items having to do with betel-chewing in Proto Oceanic, and since tobacco is a relatively recent introduction in the area (Marshall 1987), its use in betel-chewing will not be considered here. Areca nut may also be chewed by itself, without betel pepper and lime; I will return to this later. Areca and betel pepper are also chewed together, without lime. This latter practice seems to occur in places where chewing a complete betel quid is discouraged due to the influence of some Christian missions. In the absence of lime, the saliva does not turn bright red and the fact that a person has been chewing is not immediately obvious. For example, Karen Nero (pers. comm.) has informed me that on Palau people use areca and betel pepper without lime to hide chewing from the Protestant missionaries.

Chewing a betel quid produces abundant bright-red saliva, which is usually spat out at intervals, or sometimes swallowed. It also acts as a mild stimulant; and, depending on the potency of the areca nut, the person may experience a brief feeling of slight intoxication or dizziness. Chewing betel also acts as a hunger suppressant, and a person may chew a lot at a time of unavailability of food or when fasting. A betel quid is also commonly chewed after a meal, as a kind of “mouth wash”.

Various medicinal properties have been attributed to the areca nut by its traditional chewers.For example, Mosko (1985:90) says that the Mekeo people drink mashed areca boiled in water as medicine for certain illnesses, “particularly of the abdominal region”. And Powell (1976) says that in New Britain areca nuts are used to relieve toothache and other mouth infections. Areca nuts are also used as a vermifuge. Leaves of Piper betel have an antiseptic function. I witnessed use of masticated Piper betel as an antiseptic - 337 on Manam Island (Papua New Guinea) when it was applied to a cut on a person's head. Betel-pepper leaves may also be used to alleviate the pain of nettle sting by being rubbed on the affected area (Ann Chowning, pers. comm.). 1

Lime also has uses other than as an ingredient in a betel quid. In many places it is (or was) used for body decoration, to bleach people's hair, as infilling material in carvings and pottery, as pigment and in the production of certain dyes, and in ritual and magic.

On the negative side, chewing betel is said to cause cancer of the mouth, with either the betel pepper or the lime considered the culprit.

Many ethnographic and other studies comment on the important role that betel-chewing plays in social interaction among individuals, chewing betel together being a sign of sociality. So, for example, Thurston (1987:69) says that in northwestern New Britain “[o]ne of the first things people do after greeting is share betel or tobacco. Requesting betel and/or tobacco is a sign of trust, because its very use indicates that there is no suspicion of a spell having been cast on the product”. According to Firth (1957:116), on Tikopia “[w]ads of the half-masticated betel are passed freely from one person to another, particularly from older folk to younger, or from men to women, in the same way as…food”. And Oliver (1955:234) considers betel-chewing to be one of the signs of “affection” and “amity” among the Siwai (his “Siuai”) of southwest Bougainville. Although the Siwai speak a non-Austronesian language, what Oliver says about the social significance of betel-chewing there applies to Austronesian-speaking societies as well.

However, the relevance of the ingredients in a betel quid extends beyond chewing, into more formalised, ceremonial, ritual and economic contexts in numerous societies, and, of necessity, only a few examples are mentioned here. In traditional Dobu society, areca nuts played a role in marriage proposals: “Betel nut is first sent by the prospective groom or by his parents to the girl's parents. If this gift is accepted it is an indication of acceptance of the proposed marriage. If it is rejected the marriage is off” (Fortune 1963:189). According to Christine Dureau, on Simbo Island also areca nuts play a role in the initiation of marriage. The boy's family and the girl's family exchange areca nuts which they then chew together. Areca also plays a role in dealing with incestuous relations: “Senior representatives of the guilty couple gather with the two offenders. They ritually break a piece of shell money (nowadays they can tear a banknote of high denomination in half), stretch out a length of rope or string or twine and cut it in half, [and] take a single areca nut and cut it in half.” These acts symbolically represent undoing the kinship relation between the two offenders, which in turn undoes the incest (Christine Dureau, pers. comm.). In Mekeo, a chief's messengers - 338 bearing invitations to junior chiefs to attend a feast carried areca nuts with them (Hau'ofa 1981). Arosi people used to apply lime to their bodies to gain mena ‘spiritual power’ (Fox 1924). In some places, areca, betel and/or lime have also been important articles of trade. For example, Harding (1967) discusses trade in areca and lime by the Siassi in the Vitiaz Strait area. And Hau'ofa and Mosko have discussed the economic importance of areca and betel for the Mekeo. According to Hau'ofa (1981:18), “[b]etelnut 2 today provides [parts of] Mekeo with their main source of cash income.” Mosko (forthcoming) says that the inhabitants of two villages in north Mekeo “have been rather suddenly the beneficiaries of a huge surge of money and commodities from participation in maketsi, or ‘marketing’, the large scale production and sale of betel pepper,” which they supply to Port Moresby. This new wealth has contributed significantly to a large-scale transformation of traditional north Mekeo society.

Traditional accounts of the origin of areca or betel pepper are not easy to come by, but I do have a text from Toqabaqita (north Malaita) that does mention their origin. There was a man who had two sons. The man was getting old, and so he called his sons to tell them what to do when he died. He said, “When I have died, bury me at my house, and after eight days come to see my grave.” When the old man died, his sons did as he had told them. When they went to see his grave on the eighth day after burying him, they saw that a lot of little things had grown on top of the grave. They said, “Let's wait for the things to grow.” They waited, and the things grew. They looked at them, and they knew, “Oh, this is a coconut [tree]; this a betel pepper [vine], this a banana [tree], this an areca nut [tree]. They took those things and planted them around their house. When the things grew big, they tasted them. They looked at the coconut, “Oh, that's our father's head.” They looked at the areca and said, “Oh, that's our father's teeth.” They tasted the betel pepper, “Oh, that's his gall bladder.” And they saw the banana, “Oh, that's his penis.” Interestingly, the name of one of the two sons is Bualigia, which contains the word bua, a reflex of the Proto Oceanic word for ‘areca’ (see further below). In Toqabaqita, bua is an archaic word for areca; there are three other words that are used today: qota, qageru and raqaa. Ivens (1930) gives a similar story from Lau, also from north Malaita and linguistically very close to Toqabaqita. There, the man has a son by the name of 'Ota li gia (rather than Bualigia), qota being the word for areca in Lau (and one of the words for areca in Toqabaqita). 3

In a Mekeo myth, primitive men—who lived without, and had no knowledge of, women, gardens, fire, or pottery—learned to use lime when chewing betel from a culture hero. Acquisition of lime was part of a broader civilising process, whereby the men acquired knowledge of other things - 339 they lacked: gardens, fire, pots, water for cooking, and also women to marry (Mosko 1985). Similarly in Lau: when Ivens (1930:251-52) discusses the origin of areca, he says that “at some time or other in the past the areca nut proper [as opposed to wild areca] was introduced into North Mala by immigrant chiefs, and given by them to the aborigines, who are represented as living wild in the bush. The eating of the nuts reduced the aborigines to a state of civilization.” (Ivens assumes that it was Indonesian immigrants arriving at Malaita from Santa Isabel who introduced betel-chewing there.)

Lime used in betel-chewing is kept in special containers, traditionally usually made of gourds, coconut shells or sections of bamboo. Lime containers have tight-fitting lids or stoppers to protect the lime from humidity. Traditional lime containers often bore carved decoration. Today small lidded tins are frequently used. To pick a quantity of lime from a container, a special lime spatula may be used. Lime spatulas are made of various materials, such as wood, bamboo, bone or turtle shell. They may be elaborately carved and decorated (see, e.g., Beran 1988 for examples of lime spatulas from the Massim area of Papua New Guinea). Instead of a spatula, a betel-pepper catkin or a rolled-up betel-pepper leaf may be used, which then forms part of the quid.

In some areas, lime containers are symbolically associated with the vagina. In the Mekeo myth where the use of lime in betel-chewing is revealed to a bush man (see above), the ignorant bush man first mistakenly inserts a lime spatula into his benefactor's wife's vagina rather than into a lime container (Mosko 1985). (In the Mekeo myth there is also an association between a lime spatula and the penis: the bush man had first been taught by his benefactor how to copulate.) And in Kove (New Britain), the word eaua, in which the sequence au possibly reflects the Proto Oceanic word *qapuR ‘lime’, ‘lime container’ (see further below) and refers to both, can also be used to refer to the vagina, relating to a myth similar to the Mekeo one (Ann Chowning, pers. comm.).

Another set of implements sometimes used in betel-chewing is a special mortar and a pestle. These are used by people unable to chew (especially hard) areca nuts. The betel mixture is mashed in the mortar and then inserted into the person's mouth. Mortars too may be elaborately carved from wood and decorated, although a modest coconut shell may also do.

The ingredients and implements used in betel-chewing are normally carried by a person—male or female—in a handbag-like “basket”. There is no fully suitable English word for this kind of artefact; I will refer to it as personal basket (basket being the word used in Melanesian Pidgin). - 340 Depending on the type, a personal basket may be carried slung over one's shoulder or in one's hand. In some areas, woven netbags are used.

Traditionally, the practice of betel-chewing was restricted to only some parts of the Oceanic-speaking area. (Due to more recent, post-contact migrations, betel-chewing has extended to some areas where it was not practised traditionally.) It was widespread across what nowadays is Papua New Guinea and in the area that comprises the nation of the Solomon Islands, including the Polynesian outliers. It was also practised in parts of Micronesia: on Yap, Palau and in the Marianas. There is no evidence of betel-chewing in prehistoric times in other parts of Micronesia, in the New Caledonia area, in nearly all of what nowadays is Vanuatu, or further east (Fiji, Polynesia).

Betel is chewed in the Santa Cruz area (Speiser 1913, Davenport 1969), including the Polynesian outliers Taumako (Rivers 1914:I, Davenport 1968a) and Pileni (Even Hovdhaugen, pers. comm.), and as far east as Tikopia (Rivers 1914:I, Firth 1957) and Anuta (Rivers 1914:I, referring to information given to him by John Maresere). Rivers (1914:I) also mentions betel-chewing on Vanikoro in the southern area of Santa Cruz; and Davenport (1968b) gives a traditional text from Utupua, also in the southern part of Santa Cruz, where betel-chewing is mentioned. Betel-chewing is sometimes said to have been completely absent from the geographic area that comprises present-day Vanuatu; however, there is evidence of its being practised in the Torres Islands in extreme north Vanuatu. Marshall (1987) says that betel-chewing occurs in the Torres Islands; and according to Thorgeir Kolshus (pers. comm.), who has done anthropological fieldwork on Mota in the Banks Islands, “betel is…extensively used in [the] Torres.”

Figure 1 shows the “betel line”, which identifies the boundary of betel-chewing: betel was not(normally) chewed traditionally east of the line. Note that Australia is not part of the betel-chewing area. (In the westward direction, betel-chewing extends through Indonesia, Malaysia, the Philippines and Taiwan, and further west beyond the Austronesian-speaking area into Southeast Asia and the Indian subcontinent. Betel is also chewed on Madagascar and in some areas of the coast of eastern Africa.)

It is evident from the previous discussion that there is a fairly rich set of concepts associatedwith betel-chewing: areca nut, betel pepper, lime, the quid itself, lime container, lime spatula, mortar, pestle, personal basket, the activity of chewing betel itself, and the state of intoxication that sometimes results. Being able to reconstruct terms for the three main ingredients—

- 341

Figure 1: The Betel Line

areca, betel pepper and lime—is not, of itself, strong evidence that betel-chewing was practised in Proto Oceanic times. After all, as mentioned above, all three elements are used with functions other than in a betel quid. If, however, we were able to reconstruct terms for other aspects of the lexical field having to do with betel-chewing, especially when there is good reason to believe that they were (relatively) specific to betel-chewing, we would be much more confident that betel-chewing was indeed practised by speakers of Proto Oceanic. In this section, I present reconstructions of Proto Oceanic lexical items that have to do with betel-chewing, together with supporting evidence. To some degree, the list of reconstructions is a systematisation of results arrived at by other scholars. I have not managed to provide - 342 reconstructions for all the concepts mentioned above; nevertheless, the evidence is sufficient to conclude that betel-chewing was indeed an aspect of life of speakers of Proto Oceanic.

In reconstructing Proto Oceanic lexical items, I have used the subgrouping of Oceanic assumed in the Oceanic Lexicon Project (Department of Linguistics, Research School of Pacific Studies, Australian National University; see Ross 1998), as shown in Figure 2. There are three primary subgroups of Oceanic: Admiralties, Western Oceanic and Central/Eastern Oceanic; the positions of the St Matthias group and of Yapese within Oceanic are not certain. Betel-chewing is practised by speakers of languages in all five groups.

Family Tree. Oceanic, Western Oceanic, Admiralties, Central/Eastern Oceanic, St Matthias, Yapese, Figure 2: Subgrouping of Oceanic assumed in the present study.

Recently, Blust (1998) has proposed a different picture of Oceanic grouping. He suggests that the languages of the Admiralties form a primary subgroup in opposition to all the other Oceanic languages. He uses the term “Oceanic” to refer to the latter grouping, and the term “Broad Oceanic” to refer to the family that consists of the Admiralty group and the new “Oceanic” group; see Figure 3:

Family Tree. Broad Oceanic, Admiralties, Oceanic, Figure 3: Blust's (1998) subgrouping proposal concerning the languages of the Admiralties.

I will retain the more traditional subgrouping here, but even if Blust's revision were to be accepted, the major conclusions of the present study would not be materially affected. Most of the reconstructions would still be valid at the highest level (and it has to be kept in mind that the lexical information on the Admiralty languages that is publicly available is quite restricted), and, likewise, the conclusion concerning the fate of betel-chewing - 343 in Oceanic prehistory presented in the concluding section would not be contradicted.

In some cases, there is rich evidence for the given reconstruction, and only some of it is presented here. In other cases, the evidence is much more restricted and/or less straightforward—formally or semantically—and additional evidence would be very welcome. 4

There is a well established Proto Oceanic (POc) reconstruction for ‘areca’, and cognates are also found outside of Oceanic. In fact, a term for ‘areca’ is reconstructible for Proto Austronesian (PAn). (More than 35 years ago, Chowning (1963) reconstructed *bua ‘Areca catechu (areca nut)’ for what she referred to as “Proto-Melanesian”.)

Capell (1943:138) considered what he called “Indonesian” buwah II ‘areca nut’ to be a “specialization” of buwah I ‘fruit’. Capell's statement is echoed by Milke (1968). 7 More recently, Wolff (1994:515) has said that ‘fruit’ was the original meaning of the etymon and that the meaning ‘Areca catechu’ is a later development (although in the Appendix [pp. 528 and 538] he gives Areca cathecu [sic] as the meaning of what he reconstructs as PAn *buwaq). However, it is clear that the words for ‘areca’ and ‘fruit’ are different etyma: POc *buaq ‘areca’ and POc *pua ‘fruit’ (Ross 1988); for example: Nggela mbua ‘betel palm and betel nut’ and vua ‘fruit’, Tolai (Kuanua) buai ‘betel nut tree and nut’ and vuai ‘fruit or seed (of a plant or tree)’, Pileni phua ‘kernel of betel-nut’ and fua and hua (in free variation) ‘fruit’.

There is no evidence of betel-chewing by speakers of Proto Polynesian. However, betel is chewed on at least some of the Polynesian outliers, such as Taumako, Pileni, Tikopia, Rennell and Bellona. The ancestors of those Polynesians adopted the custom of betel-chewing, and in some cases the relevant terminology, after migrating out of Polynesia.

- 344

There is also a solid POc reconstruction for ‘lime’, and this etymon also can be traced back to PAn. The term referred not only to lime but metonymically also to lime containers.

In a number of languages, the same term is used both for lime and lime container; for examples see the Manam, Motu, Kwaio and Ulawa forms listed above. (The metonymic connection between lime and lime container is not restricted to languages with a reflex of POc *qapuR; for example, in Toqabaqita fena refers to both.) At present at least, there is no evidence for reconstructing a separate term for ‘lime container’, and so it can be assumed that POc *qapuR had both meanings, ‘lime’ and ‘lime container’.

While there are secure reconstructions for ‘areca’ and ‘lime’, there is, to my knowledge, no published reconstruction for the third main ingredient in a betel quid, betel pepper. In a personal communication, Meredith Osmond has informed me that she has reconstructed, as part of the Oceanic Lexicon Project, POc *pul ‘betel pepper vine’ on the basis of the following evidence: Loniu (Adm) pun ‘betel pepper plant, leaf and fruit’, Takia (WOc) ful ‘betel pepper vine’, and Lukep-Pono (WOc) ul ‘betel pepper vine’. However, this reconstruction needs to be amended. Data from Manam and Chamorro are evidence for reconstructing *pulu rather than *pul:

- 345

Since a cognate of POc *pulu is found in Chamorro, it is possible to reconstruct *pulu ‘Piper betle vine’ for Proto Malayo-Polynesian.

In many present-day languages one finds verbs that refer specifically to chewing betel. (In other languages, non-specific verbs are used, such as ‘eat’.) And it is possible to recontruct for POc a verb that was used to refer to chewing betel, although its meaning may have been broader than that:

- 346

Ross (1988) has reconstructed POc *jamu ‘chew (areca nut)’ on the basis of Mussau samu ‘bite’ (Blust 1984a gives samusamu(-a) ‘bite’), Manam-zamu(ku), Proto Markham *jaim ‘areca nut’, Tiang (WOc) sem, Proto Southeast Solomonic *damu, and Proto Central Pacific *jamu ‘scraps of food’. However, there are some problems with this reconstruction. First, the Manam form for ‘chew betel’ is zem, not zamu(ku). I am not aware of any form zamu(ku), and Böhm (1975) does not give it either. (Böhm also gives zem for ‘chew betel’.) Second, although the d in the Proto Southeast Solomonic form damu may reflect POc *j, it may also reflect other POc phonemes, in particular *dr. In fact, I will argue later that the Proto Southeast Solomonic form reflects a different etymon. If one disregards the Proto Southeast Solomonic form, it is only the Western Oceanic witnesses that have meanings having to do with areca or with betel chewing. If a form *jamu is indeed to be reconstructed for POc, there is no convincing evidence that it referred to betel chewing.

There is another verb reconstructible for Proto Oceanic that may have had some, probably only tangential, relation to betel-chewing:

- 347

Most likely, the meaning of *˜asi was not specific to betel-chewing. Rather, it may have referred more generally to chewing, crushing with teeth something that causes liquid to come out or saliva to form, the object chewed not to be swallowed. In Toqabaqita the meaning of *ngasi shifted through metonymy to ‘sugar cane’, sugar cane being the typical kind of object to which the action referred to as *ngasi was applied. And in Sengseng a word for ‘sugar cane’ is nas-˜in, which is a nominalisation of nas ‘chew sugar cane’. 14

A full betel quid consists of areca nut, betel pepper and lime, but sometimes people chew areca nut all by itself. As mentioned, various medicinal properties have been attributed to the areca nut, which is one of the reasons why it is sometimes chewed without betel pepper and lime. Ann Chowning has pointed out to me that there is another, more mundane, reason why some people chew areca by itself: they just like the taste. It is, in fact, possible to reconstruct a verb that referred to chewing areca nut by itself, even though that was not the exclusive meaning of the term:

- 348

Among the implements used in betel-chewing, a term is reconstructible for a lime spatula. However, before presenting the evidence, some introductory remarks are necessary. As mentioned above, in 1988 Ross (p. 78) reconstructed *jamu ‘chew (areca nut)’, partly on the basis of forms from Southeast Solomonic languages, which he interpreted as reflecting Proto Southeast Solomonic *damu ‘chew (areca nut)’. However, more recently he has proposed the reconstruction *d(r)amut ‘lime spatula’ for POc on the basis of the following forms: Mussau (StM) ai ra-ramuti ‘lime spatula’ (ai ‘stick’), Lou (Adm) rem ‘lime spatula’, Gedaged (WOc) dam ‘(bone or wood) lime spatula’, Nakanai (the alternative name “Lakalai” is used elsewhere in this paper) (WOc) damu ‘lime spatula’, Sa'a (CEOc) demu ‘chew betel nut’ and Arosi (CEOc) damu ‘chew betel nut’ (Meredith Osmond, pers. comm.). He also suggests that the final i in Mussau ai ra-ramuti indicates that *d(r)amuti was a transitive verb whose meaning may have been ‘add lime to a chew of betelnut’. I agree with Ross's suggestion that the Southeast Solomonic (Sa'a and Arosi) forms do not reflect POc - 349 *jamu ‘chew (areca nut)’ (see above); however, his reconstruction *d(r)amuti needs to be modified both in its formal and semantic aspects.

There is evidence for reconstructing a POc term for ‘lime spatula’ which was a nominalisation of a verb. And although the verb may have been used to refer to adding lime to a chew of betel (as Ross has suggested), its basic meaning was ‘lick’. In betel-chewing, the areca and pepper are normally masticated first, and the lime is added only later. The lime is normally placed in the mixture of areca and betel to avoid burning one's lips or mouth with the lime. When a spatula (rather than a betel catkin or rolled-up betel leaf) is used to deposit the lime into the betel mixture, the action is not unlike that of licking (even though the tongue is usually not directly used).

I will first present evidence for reconstructing the verb(s); the evidence includes a few forms that will subsequently be used to reconstruct the term for ‘lime spatula’. Although a number of languages indicate a stem-final consonant in the POC form, there is no agreement among the witnesses on the quality of the consonant. (The symbol “C” is used to represent the potential undetermined consonant.) Furthermore, a doublet needs to be reconstructed, distinguished by u vs. i alternation in the second syllable. (U-i “crossover” is attested in some other Austronesian etyma; see Blust 1970.) The witnesses that have the meaning ‘lick’ are listed first, followed by witnesses that have the meaning ‘chew betel’ or ‘lime spatula’.

- 350

The term for ‘lime spatula’ was a nominalisation of the verb ‘lick’ by means of the “instrumental” prefix i- (Pawley 1972:39, 84):

The connection between the verbal meaning ‘lick’ and the nominal meaning ‘lime spatula’ is shown by the Sengseng form hom ‘lick’ and hom-ŋin ‘lime spatula’, where the latter is a nominalisation of the former. In - 351 Sengseng, -ŋin nominalisations refers to entities involved in some way in an event designated by the source verb; for example, kes-ŋin ‘knife’, lit.: ‘cutter’; nas-ŋin ‘sugarcane’, lit.: ‘the chewed’ (Chowning 1985:185).

As mentioned above, the term *qapuR referred not only to lime but also to lime containers.

Although mortars for crushing areca nuts are common in betel-chewing areas, I have not been able to reconstruct a term for Proto Oceanic.

The ingredients used in betel-chewing and probably also at least some of the implements used were carried in personal baskets. A term for a basket is reconstructible for Proto Oceanic, and the evidence suggests that it was used to refer to personal baskets (and perhaps to some other kinds of basket as well).

- 352

Kirch (1997) says that the Lapita people, whom he identifies as speakers of Proto Oceanic, probably did not chew betel. The evidence presented in the present study points to a different conclusion. Proto Oceanic had terms for all three ingredients in a betel quid, as well as other terms specific to betel-chewing. There was a verb that could be used to refer to chewing a betel quid. And there was also a verb that referred to chewing areca by itself. A specific term for one implement used in betel-chewing is also reconstructible, that for a lime spatula. Another term referred to lime containers, as well as to lime. From the linguistic evidence it can be concluded that speakers of Proto Oceanic did indeed chew betel. And the fact that they chewed areca by itself suggests that they may have attributed some medicinal properties to areca.

The conclusion that speakers of Proto Oceanic chewed betel gives rise to other, broader questions. Investigations of those questions is beyond the scope of this study, and so they will only be noted here and will be left for future research. One question is this: did the people who spoke the precursor of Proto Oceanic bring the custom of betel-chewing with them as they moved into what became the homeland of Proto Oceanic, or did they acquire it only later through contact with the non-Austronesians who were already living in the area? There is evidence, some of it only circumstantial, that speakers of the immediate precursor of Oceanic did engage in betel-chewing.

First, although the origin of the areca palm is not clear, it is believed to have been domesticated somewhere in the island Southeast Asia area (Corner 1966, Purseglove 1972). (The history and distribution of the betel-pepper vine is even more obscure.) The fact that the areca palm was domesticated suggests a fairly intensive use which went beyond custom medicine. But, of course, one cannot be certain that it was speakers of an Austronesian rather than a non-Austronesian language who domesticated the areca palm. Second, there is archaeological evidence—in the form of betel-stained teeth—of betel-chewing in several parts of the Austronesian-speaking area: in Taiwan (the assumed homeland of Proto Austronesian) from perhaps as far back as 1500 B.C., in the southern Philippines from about 3000 B.C. (Bellwood 1997), and in the St Matthias Islands from as early as 1500-1200 B.C. (Kirch et al. 1989). 24. It is the latter finding that is of particular relevance to the present study, as the St Matthias Islands lie within the Bismarck Archipelago, the presumed homeland of Proto Oceanic. There is also archaeological evidence from the St Matthias Islands of the use of lime as infill material to decorate dentate-stamped Lapita pottery (Kirch 1997). According to Roger Green (pers. comm.), the fact that this decorative use of lime existed in Neolithic Southeast Asia but not in pre-Lapita times in the Bismarcks or in - 353 New Guinea (Spriggs 1997) suggests that lime was brought into the Bismarcks from the West, most likely by pre-Proto Oceanic migrants.

Clearly, some caution needs to be exercised in the evaluation of the archaeological evidence. First, one cannot be absolutely certain that the betel-stained teeth belonged to speakers of Austronesian languages. And second, the use of lime as decorative material is, of itself, no evidence of betel-chewing, because lime is commonly used—in a variety of functions— outside of the betel-chewing area. 25 At the same time, the linguistic and the archaeological evidence are fully compatible with each other. As shown above, a verb that referred to chewing betel can be reconstructed not only for Proto Oceanic but also for Proto Malayo-Polynesian, a distant ancestor of Proto Oceanic. The linguistic evidence, combined with the archaeological evidence and with the evidence from initial domestication of areca, leads to the conclusion that the people who spoke the immediate precursor of Proto Oceanic did chew betel, and that they did bring the custom with them. It is highly unlikely that the idea of chewing areca nuts, betel pepper and lime together in a quid originated more than once.

If so, how does one explain the restricted distribution of betel-chewing in the Oceanic area? The betel quid is only one of two major drugs used in the Oceanic area, kava being the other one. Many scholars, such as Rivers (1914), have commented on the nearly complementary distribution of betel-chewing and kava-drinking. The distribution is not perfectly complementary because there are areas where both are or have been have practised, such as parts of New Guinea and the Admiralties, and (parts of?) the Santa Cruz area; and areas where neither has apparently been practised, such as New Caledonia, New Zealand and Easter Island. (The betel line in Figure 1 identifies the boundary of the distribution of betel-chewing, not a boundary between the betel-chewing and the kava-drinking areas; see Map 1 in T. Crowley (1994) for kava-drinking areas to the west of the betel line.)

Pawley and Green (1973:32) have reconstructed for Proto Oceanic a term for kava, *kawa ‘Piper methysticum’ as part of a list of reconstructions having to do with “agriculture [and] arboriculture”, the implication being that speakers of Proto Oceanic consumed kava. Pawley and Green considered kava to be indigenous to the New Guinea-East Indonesia area. Similarly, Brunton (1989:81), in a social anthropological study of kava, posits “a link between kava and the speakers of Proto-Oceanic”. Brunton assumes that kava originated in the Bismarck Archipelago area—the presumed homeland of Proto Oceanic—and spread from there as “a consequence of the dispersion of the Oceanic-speaking peoples”. He attributes cultural and religious significance” to kava. According to him, kava-drinking used to be more widespread in Melanesia, and its disappearance in some areas “is simply - 354 one aspect of the much broader problem of the high degree of religious instability in Melanesia” (Brunton 1989:93).

However, a very different conclusion has been reached by Lebot on the basis of a detailed botanical study of the distribution of cultivars of kava (Lebot 1989, Lebot et al. 1992). Lebot (1989) argues that the evidence points to northern Vanuatu as the place of origin of Piper methysticum. And in Lebot et al. (1992), Brunton's suggestion that kava originated in the Bismarck Archipelago area is explicitly disputed. The botanical evidence shows that kava was first domesticated in northern Vanuatu and that it “was carried eastward into Fiji and Polynesia and westward into scattered areas of New Guinea and into two islands of central Micronesia” (p. 6). Lebot et al. (p. 26) quote from a personal communication from Roger Green (“{}” enclose Lebot et al.'s insertions):

If Piper methysticum, which is the domesticated variety of Piper wichmannii, originated in northern Vanuatu, it is highly doubtful that kava was drunk in Proto Oceanic times, although speakers of Proto Oceanic may have been familiar with the wild Piper wichmannii.

T. Crowley (1994) has reconsidered the linguistic evidence for reconstructing a term for kava in light of Lebot's botanical evidence that domesticated kava originated in northern Vanuatu. He says (p. 92) that “the only word for kava that can be assumed with any certainty to have any degree of antiquity in the Melanesian area is Clark's (n.d.) reconstruction of *maloku in the language ancestral to probably all of the languages spoken between Efate and the Torres Islands in Vanuatu, that is, Proto North-Central Vanuatu”. Words for ‘kava’ found in some kava-drinking areas north and west of Vanuatu, that appear to be cognates of the Proto Polynesian word *kawa are presumably later borrowings, adopted together with the custom of kava-drinking introduced from Polynesia in post-Proto Oceanic times.

We are then left with the following conclusions: speakers of Proto Oceanic chewed betel but did not drink kava. The custom of betel-chewing was carried - 355 by speakers of a descendant or descendants of Proto Oceanic as far east as the Torres Islands in northern Vanuatu. And it was somewhere in northern Vanuatu that kava was first domesticated. This is unlikely to be mere coincidence. It appears that in the area of northern Vanuatu betel-chewing was replaced by kava-drinking. The spread of betel-chewing came to an end there, and only kava-drinking was carried into areas settled subsequently. We may never know the reasons for why the custom of betel-chewing came to an end, but most likely that was due to abandonment rather than to loss. All three ingredients in the betel quid are eminently portable, and most likely were routinely carried on canoe voyages. In fact, lime is widespread east of the betel line. There are no reasons to assume that the areca palm and the betel vine would not have grown on islands east of the line. Most likely, there is a close connection between the end of betel-chewing and the beginning of kava-drinking: the custom of betel-chewing was abandoned when and because the custom of kava-drinking was adopted.

According to information from Thorgeir Kolshus (pers. comm.), the betel line runs between the Torres and the Banks Islands, two island groups that are very close to each other. The presence of kava-drinking as well as betel-chewing in the southern Santa Cruz-Tikopia area is consistent with its geographical proximity to the area where kava-drinking originated and betel-chewing stopped. Presumably, the custom of kava-drinking spread there from northern Vanuatu. Thorgeir Kolshus (pers. comm.) says that “in the Banks the [areca] palm is found but usually not utilised (if there are no Torres Islanders, that is). Kava is abundant in the Banks, but in [the] Torres they don't seem to have the right soil for it to thrive.” Even though kava-drinking managed to spread to Santa Cruz and Tikopia, for whatever reasons it did not spread to the Torres Islands, unsuitable conditions for its growing there possibly being a factor.

Crowley (1994) suggests that kava-drinking did not spread to southern Vanuatu in the early times after it developed in northern Vanuatu. Rather, the linguistic evidence suggests that it was introduced there at a later time through back migrations from Polynesia. Lynch (1996) has argued—on the basis of linguistic evidence—that Piper methysticum (kava) and its use were introduced into southern Vanuatu by speakers of a Polynesian language, specifically Futuna.

There is no evidence either of betel-chewing or of directly introduced kava-drinking in southern Vanuatu, and there is no evidence either of betel-chewing or of kava-drinking in New Caledonia. Of course, of itself the absence of both customs in the two areas is not much evidence. However, Geraghty (1988:147) has suggested that what he calls the “Southern Oceanic” subgroup within Oceanic, which contains the languages of New Caledonia - 356 and the non-Polynesian languages of the Loyalty Islands “is more closely related to Vanuatu, and especially Southern Vanuatu, than to any other language family”. The absence both of betel-chewing and of kava-drinking in southern Vanuatu and in the New Caledonia-Loyalty area is consonant with the linguistic evidence; together they suggest a shared settlement history.

Of the two drugs used in the Oceanic area, kava has been given considerably more prominence by researchers than betel-chewing. While there are general studies of kava (e.g., Brunton 1989, Lebot 1989, Lebot et al. 1992), there are no comparable syntheses of betel-chewing. The present paper is a small step in that direction. Paying closer attention to betel-chewing might lead to a better understanding of the prehistory of the Oceanic area within the betel line. If descendants of speakers of Proto Oceanic were responsible for the spread of the areca palm and possibly also of betel pepper, and if it were to be shown that there are different varieties of both species, then the distribution of the varieties might give us a window on the migration patterns.

On the basis of the nearly complementary distribution of kava-drinking and betel-chewing, Rivers (1914) assumed two prehistoric migrations into the island Pacific: one by “kava-people” and another, later one by “betel-people”. This assumption of two migrations finds no support in the linguistic evidence. In fact, if anything, the linguistic evidence contradicts it. The languages spoken by the presumed descendants of the “kava-people” and the languages spoken by the presumed descendants of the “betel-people” belong in one and the same language group: Oceanic. Furthermore, the Oceanic group also contains the languages of New Caledonia, where there is no evidence either of kava-drinking or of betel-chewing. The situation is even more complex in Micronesia. Two languages spoken by betel-chewers are not Oceanic (although they are Austronesian): Palauan and Chamorro. One the other hand, Yapese, also spoken by betel-chewers, is Oceanic. Two other Oceanic languages of Micronesia are spoken by kava-drinkers: Ponapean and, in earlier times, Kosraean. There is no evidence either of betel-chewing or of kava-drinking elsewhere in Micronesia. If Rivers were correct, one would expect to find some correlation between the distribution of kava-drinking, betel-chewing and the absence of both on the one hand, and linguistic groupings on the other; but there is none.

Brunton (1989) calls kava “the abandoned narcotic”. He suggests that it was used by speakers of Proto Oceanic, that it was subsequently carried in migrations through the Oceanic area, and that its use was more widespread than it is now. He attributes its demise in much of Melanesia to what he calls “religious instability”. The picture presented in this paper is quite different. Betel-chewing was practised by speakers of Proto Oceanic, and - 357 the custom was carried southeastward through migrations as far as the Torres Islands in northern Vanuatu. In northern Vanuatu it was replaced by kava-drinking. It is the betel quid, not kava, that is the abandoned drug.

In preparing this paper I profited greatly from discussions with Ann Chowning. I am grateful to Ann, to Christine Dureau and to Roger Green for valuable comments on an earlier draft. I also want to thank the following people for sharing with me unpublished data on various languages (the languages are identified in the Appendix): Robert Blust, Joel Bradshaw, Ann Chowning, Even Hovdhaugen, Tikila McFarlane and Meredith Osmond.

- 358

Languages from which data have been used in this paper, and sources of the data

- 359

• Bellwood, Peter, 1997. Prehistory of the Indo-Malaysian Archipelago. Rev. ed. Honolulu: University of Hawai'i Press.
• Beran, Harry, 1988. Betel-Chewing Equipment of East New Guinea. Aylesbury, UK: Shire Publications.
• Blust, Robert A. 1970. I and u in the Austronesian languages. University of Hawaii Working Papers in Linguistics, 2:113-45.
• —— 1983-84. Austronesian etymologies—II. Oceanic Linguistics, 22-23:29-149.
• —— 1984a. A Mussau vocabulary, with phonological notes. In Papers in New Guinea Linguistics No. 23. Pacific Linguistics A-69. Canberra: Australian National University, pp. 159-208.
• —— 1984b. Malaita-Micronesian: An Eastern Oceanic subgroup? Journal of the Polynesian Society, 93:99-140.
• —— 1998. A note on higher-order subgroups in Oceanic. Oceanic Linguistics, 37:182-88.
• —— Forthcoming. Austronesian Comparative Dictionary.
• Böhm, Karl, 1975. Das Leben einiger Inselvölker Neuguineas. St. Augustin: Anthropos-Institut.
• Brunton, Ron, 1989. The Abandoned Narcotic: Kava and Cultural Instability in Melanesia. Cambridge: Cambridge University Press.

• Capell, A., 1943. The Linguistic Position of South-Eastern Papua. Sydney: Australasian Medical Publishing Company.
• —— 1973. A New Fijian Dictionary. 4th ed. Suva, Fiji: Government Press.
• Chowning, Ann, 1963. Proto-Melanesian plant names. In Jacques Barrau (ed.), Plants and the Migrations of Pacific Peoples: A Symposium. Honolulu: Bernice P. Bishop Museum Press, pp. 39-44.
• —— 1985. Rapid lexical change and aberrant Melanesian languages: Sengseng and its neighbours. In Andrew Pawley and Lois Carrington (eds), Austronesian Linguistics at the 15th Pacific Science Congress. Pacific Linguistics C-88. Canberra: Australian National University, pp. 169-98.
• —— 1991. Proto Oceanic culture: The evidence from Melanesia. In Robert Blust (ed.), Currents in Pacific Linguistics: Papers on Austronesian Languages and Ethnolinguistics in Honour of George W. Grace. Pacific Linguistics C-117. Canberra: Australian National University, pp. 43-75.
• —— and Ward Goodenough, n.d. Bileki (Lakalai, West Nakanai) Vocabulary. Typescript.
• Churchward, C. Maxwell, 1959. Tongan Dictionary. London: Oxford University Press.
• Clark, Ross, n.d. Some Proto-North Central Vanuatu Reconstructions with Selected Supporting Evidence. MS.
• Codrington, R.H. and J. Palmer, 1896. A Dictionary of the Language of Mota, Sugarloaf Island, Banks' Islands. London: Society for Promoting Christian Knowledge.
• Corner, E.J.H., 1966. The Natural History of Palms. London: Weidenfeld and Nicolson.

- 360

• Crowley, Susan Smith, 1986. Tolo Dictionary. Pacific Linguistics C-91. Canberra: Australian National University.
• Crowley, Terry, 1992. A Dictionary of Paamese. Pacific Linguistics C-121. Canberra: Australian National University.
• —— 1994. Proto who drank kava? In A.K. Pawley and M.D. Ross (eds), Austronesian Terminologies: Continuity and Change. Pacific Linguistics C-127. Canberra: Australian National University, pp. 87-100.

• Davenport, William, 1968a. Social organization notes on the northern Santa Cruz Islands: The Duff Islands (Taumako). Baessler-Archiv, Neue Folge, 16:137-205.
• —— 1968b. Social organization notes on the southern Santa Cruz Islands: Utupua and Vanikoro. Baessler-Archiv, Neue Folge, 16:207-75.
• —— 1969. Social organization notes on the northern Santa Cruz Islands: The Main Reef Islands. Baessler-Archiv, Neue Folge, 17:151-243.

• Firth, Raymond, 1957. We, the Tikopia: A Sociological Study of Kinship in Primitive Polynesia. 2nd ed. London: George Allen & Unwin.
• Fortune, R.F., 1963. Sorcerers of Dobu: The Social Anthropology of the Dobu Islanders of the Western Pacific. Rev. ed. London: Routledge & Kegan Paul.
• Fox, Charles E., 1924. The Threshold of the pacific: An Account of the Social Organization, Magic and Religion of the People of San Cristoval in the Solomon Islands. London: Kegan Paul, Trench, Trubner & Co.
• —— 1955. A Dictionary of the Nggela Language (Florida, British Solomon Islands). Auckland: Unity Press.
• —— 1974. Lau Dictionary. Pacific Linguistics C-25. Canberra: Australian National University.
• —— 1978. Arosi Dictionary. Revised edition, prepared by Mary Craft. Pacific Linguistics C-57. Canberra: Australian National University.

• Geraghty, Paul, 1988. The reconstruction of Proto-Southern Oceanic. In Ray Harlow and Robin Hooper (eds), VICAL 1: Oceanic Languages: Papers from the Fifth International Conference on Austronesian Linguistics, Auckland, New Zealand, January 1988. Part I. Auckland: Linguistic Society of New Zealand, pp. 141-56.

• Hamel, Patricia J., 1994. A Grammar and Lexicon of Loniu, Papua New Guinea. Pacific Linguistics C-103. Canberra: Australian National University.
• Harding, Thomas G., 1967. Voyagers of the Vitiaz Strait: A Study of a New Guinea Trade System. Seattle and London: University of Washington Press.
• Hau'ofa, Epeli, 1981. Mekeo: Inequality and Ambivalence in a Village Society. Canberra: Australian National University Press.

• Ivens, Walter G., 1929. A Dictionary of the Language of Sa'a (Mala) and Ulawa South-East Solomon Islands. London: Oxford University Press; and Melbourne: Melbourne University Press.
• —— 1930. The Island Builders of the Pacific. London: Seeley, Service & Co.

• Keesing, R.M., 1975. Kwaio Dictionary. Pacific Linguistics C-35. Canberra: Australian National University.
• Kirch, Patrick V., 1997. The Lapita Peoples: Ancestors of the Oceanic World.

- 361

• Cambridge, Mass. and Oxford: Blackwell.
• —— and Paul H. Rosendahl, 1973. Archaeological investigation of Anuta. In D.E. Yen and Janet Gordon (eds), Anuta: A Polynesian Outlier in the Solomon Islands. Pacific Anthropological Records No. 21. Honolulu: Bernice P. Bishop Museum, pp. 25-108.
• —— Daris R. Swindler and Christy G. Turner II, 1989. Human skeletal and dental remains from Lapita sites (1600-500 B.C.) in the Mussau Islands, Melanesia. American Journal of Physical Anthropology, 79:63-76.
• A Kuanua Dictionary, 1964. Rabaul, Papua New Guinea: Methodist Mission Press.

• Lebot, Vincent, 1989. L'histoire du kava commence par sa découverte. Journal de la Société des Océanistes, 88-89(l-2):89-114.
• —— Mark Merlin and Lamont Lindstrom, 1992. Kava: The Pacific Drug. New Haven and London: Yale University Press.
• Lichtenberk, Frantisek, 1994. The raw and the cooked: Proto Oceanic terms for food preparation. In A.K. Pawley and M.D. Ross (eds), Austronesian Terminologies: Continuity and Change. Pacific Linguistics C-127. Canberra: Australian National University, pp. 267-88.
• —— and Meredith Osmond, 1998. Food preparation. In Malcolm Ross, Andrew Pawley and Meredith Osmond (eds), The Lexicon of Proto Oceanic. Volume 1. Pacific Linguistics C-152. Canberra: Australian National University, pp. 143-71. To appear.
• Lincoln, Peter C., 1977. Gitua-English Vocabulary (preliminary version). Typescript.
• Lister-Turner, R. and J.B. Clark, n.d. A Dictionary of the Motu Language of Papua. Second edition, edited by Percy Chatterton. Sydney: Government Printer.
• Lynch, John, 1996. Kava-drinking in southern Vanuatu: Melanesian drinkers, Polynesian roots. Journal of the Polynesian Society, 105:27-40.

• Mager, John F., 1952. Gedaged-English Dictionary. Columbus, Ohio: Board of Foreign Missions of the American Lutheran Church.
• Marshall, Mac, 1987. An overview of drugs in Oceania. In Lamont Lindstrom (ed.), Drugs in Western Pacific Societies: Relations of Substance. Association for Social Anthropology in Oceania Monograph No. 11. Lanham, MD: University Press of America, pp. 13-49.
• Milke, Wilhelm, 1968. Proto-Oceanic addenda. Oceanic Linguistics, 7:147-71.
• Money, Lucy, n.d. Babatana Wordlist. Typescript.
• Mosko, Mark S., 1985. Quadripartite Structures: Categories, Relations, and Homologies in Bush Mekeo Culture. Cambridge: Cambridge University Press.
• —— forthcoming. Magical money: Commoditisation and the linkage of maketsi (“market”) and kangakanga (“custom”) in contemporary North Mekeo. In J. Robbin and D. Akin (eds), State and Local Currencies in Melanesia. Pittsburgh: University of Pittsburgh Press.
• Muzzey, Marguerite, 1979. Hote-English Dictionary. Typescript.

• Newell, Leonard E. and Francis Bon'og Poligon, 1993. Batad Ifugao Dictionary with Ethnographic Notes. Manila: Linguistic Society of the Philippines.

• Oliver, Douglas L., 1955. A Solomon Island Society: Kinship and Leadership Among the Siuai of Bougainville. Cambridge, Mass.: Harvard University Press.

- 362

• osmond, Meredith and Malcom Ross, 1998. Household artefacts. In Malcolm Ross, Andrew Pawley and Meredith Osmond (eds), The Lexicon of Proto Oceanic. Volume 1. Pacific Linguistics C-152, pp. 67-114. Canberra: Australian National University. To appear.

• Pawley, Andrew. 1972. On the internal relationships of Eastern Oceanic languages. In R.C. Green and M. Kelly (eds), Studies in Oceanic Culture History. Volume 3. Pacific Anthropological Records No. 13. Honolulu: Bernice P. Bishop Museum, pp. 1-142.
• —— and Roger Green, 1973. Dating the dispersal of the Oceanic languages. Oceanic Linguistics, 12: 1-67.
• Powell, J.M., 1976. Ethnobotany. In K. Paijmans (ed.), New Guinea Vegetation. Amsterdam: Elsevier; and Canberra: Australian National University Press, pp. 106-83.
• Purseglove, J.W., 1972. Tropical Crops: Monocotyledons. 2 volumes. London: Longman.

• Riesenfeld, Alphonse, 1947. Who are the betel people? Internationales Archiv für Ethnographie 45:157-215.
• Rivers, W.H.R., 1914. The History of Melanesian Society. 2 volumes. Cambridge: At the University Press.
• Ross, M.D., 1988. Proto Oceanic and the Austronesian Languages of Western Melanesia. Pacific Linguistics C-98. Canberra: Australian National University.
• —— 1998. Proto-Oceanic adjectival categories and their morphosyntax. Oceanic Linguistics, 37:85-119.

• Speiser, Felix, 1913. Two Years with the Natives in the Western Pacific. London: Mills & Boon.
• Spriggs, Matthew, 1997. The Island Melanesians. Oxford and Cambridge, Mass.: Blackwell.
• Sullivan, Roger, 1998. Betel-Chewing and Schizophrenia. Seminar presented to the Department of Anthropology, University of Auckland, 7 May 1998.

• Thurston, William R., 1987. Processes of Change in the Languages of North-Western New Britain. Pacific Linguistics B-99. Canberra: Australian National University.
• —— 1996a. Amara: An Austronesian language of northwestern New Britain. In M.D. Ross (ed.), Studies in Languages of New Britain and New Ireland. I: Austronesian Languages of the North New Guinea Cluster in Northwestern New Britain. Pacific Linguistics C-135. Canberra: Australian National University, pp. 197-248.
• —— 1996b. The Bibling languages of northwestern New Britain. In M.D. Ross (ed.), Studies in Languages of New Britain and New Ireland. I: Austronesian Languages of the North New Guinea Cluster in Northwestern New Britain. Pacific Linguistics C-135. Canberra: Australian National University, pp. 249-392.

• Topping, Donald M., Pedro M. Ogo and Bernadita C. Dungca, 1975. Chamorro-English Dictionary. Honolulu: University Press of Hawaii.

• Waterhouse, J.H.L., 1949. A Roviana and English Dictionary. Revised and enlarged by L.M. Jones. Sydney: Epworth.
• Wolff, John U., 1994. The place of plant names in reconstructing Proto Austronesian.

- 363

• In A.K. Pawley and M.D. Ross (eds), Austronesian Terminologies: Continuity and Change. Pacific Linguistics C-127. Canberra: Australian National University, pp. 511-40.

• Yen, D.E., 1973. Agriculture in Anutan subsistence. In D.E. Yen and Janet Gordon (eds), Anuta: A Polynesian Outlier in the Solomon Islands. Pacific Anthropological Records No. 21. Honolulu: Bernice P. Bishop Museum, pp. 112-49.

• Zorc, R. David Paul, 1994. Austronesian culture history through reconstructed vocabulary (an overview). In A.K. Pawley and M.D. Ross (eds), Austronesian Terminologies: Continuity and Change. Pacific Linguistics C-127. Canberra: Australian National University, pp. 541-94.

- 364 Page is blank