The practice of magic in Oceanic societies is based on the belief that ultimate power resides in the ghosts of the ancestors, and that there are people in the community who can summon this power under certain conditions to achieve particular ends.
Magic traditionally imbued every aspect of the lives of Oceanic speakers. It was the tool by which they sought to enhance outcomes in their gardening practices, their fishing, their state of health, their fertility, their efforts in love and war, and power over their enemies. It provided an explanation for otherwise inexplicable events, and served as a form of social control, enforcing prohibitions on certain behaviours. But like all tools, it could be misused by those with their own motives. It offered a means of reacting, attempting to influence, by counter-action or retribution, and was used for both good and evil purposes. Sorcery, sometimes used loosely as a term for any magic, more commonly refers to that performed to inflict harm on another. In its most extreme form it is magic directed towards the intended death of a victim (black magic). Where possible, black magic is treated here separately in §8.4.
Most approaches made to ancestors, as well as giving thanks and perhaps showing respect and obeisance, are to seek their good will for future endeavours. Approaches are invariably made at the start of fresh initiatives such as the breaking of new ground in a garden, the opening of a new men’s house, the construction and launching of a trading canoe, preparation for a fight, and preparation for hunting and fishing expeditions. Rituals on such occasions usually incorporate a sequence of specific spells. For example, in preparation for a fight, spells may be cast to inspire the warriors with courage, to enable them to creep upon the enemy unobserved, to cause the weapons to inflict lethal wounds, and to make the enemy sluggish and weak. In gardening, magic may be called on to preserve the fences, drive away plant diseases and insects, bring sun or rain, and make the yams increase in size (Hogbin 1964:86). The ritual is usually so closely identified with the occasion that to neglect it is to court disaster. In western Oceanic communities such rituals are typically led by a headman or expert magician. Spells also exist for individuals to seek good fortune in personal matters, such as attracting a person of the opposite sex, help in childbirth or success in fishing. At times the request is accompanied by offerings, usually harvest produce or fish.
Misadventure of all kinds, including illness, accidents such as falling from a tree, failure of rainfall or even volcanic activity can be interpreted as due to displeasure of the ancestors, and must be counteracted by magic, often accompanied by some form of payment. An afflicted person may seek to identify a past act suspected of causing offence, which can then be atoned. Commonly, misfortune is believed caused by deliberate human action against another, i.e. sorcery. In this case the victim may have a good idea of someone he or his close kin have offended, or know of those who hold a grudge against him, and he may have a good idea of those with the ability to practise sorcery and identify the likely sorcerer. Those with the skill to cause harm usually have the power to undo their magic.
A major problem for those of us engaged in historical reconstruction is paucity of detailed terminology in magical matters. Many early ethnographers were missionaries who could be expected to discourage any magical beliefs as incompatible with Christianity, and did little to record what they saw of magic practice. What was recorded was often local practice reinterpreted to accord with Christian ritual and belief. Also relevant was a natural reticence when it came to discussing secret and mysterious processes with strangers. If a secret spell is shared among strangers, its efficacy is believed lost. It may take years for an outsider living within a community to be allowed to know of such matters. Raymond Firth who spent a year in Tikopia in 1928–9, learnt much later that the chiefs had given orders that he was to be told nothing about their gods and ritual practices (1957:8). As well, communities took care to keep this knowledge to themselves lest it be used by neighbouring communities against them. So although there is a substantial degree of commonality in what is known about the general practice of magic in Melanesia, collected terms tend to be highly restricted in distribution and very generalised in their listed meaning. Consequently very few POc lexical reconstructions have been possible.
As an example of the kind of difficulties involved, one of the rare reconstructions possible is the following. It evidently related to some aspect of magic, but without further evidence we cannot know if the reference is generic or specific, or whether the implied actor is human or non-human.
| POc | *masi | ‘magic; perform magic’ | |
| NNG | Mengen | masi-a | [V] ‘perform (a magic ritual)’ |
| NNG | Mengen | masi-ŋ | [N] ‘the act of performing a magic ritual; medicine’ |
| NNG | Mengen | mas-mas | [V] ‘work magic, supernaturally perform’ |
| NNG | Mengen | mas-masia | [V] ‘perform magic on’ |
| NCV | Paamese | masi-ŋe | [N] ‘love magic’ |
| NCV | Big Nambas | masi-n | [N] ‘love magic’ |
| PNCV | *kai-masi | ‘sorcerer’ (POc *(k,kʷ)ai ‘person belonging to a category’; vol.5:48–50) (Clark 2009) | |
| NCV | Paamese | ei-masi | ‘evil spirit’ |
| NCV | Nguna | (na)kai-masi | ‘sorcerer’ |
The practice of magic is considered in §8.2, its practitioners in §8.3, and sorcery in §8.4.
Magic practice consists primarily in the recitation of certain spells and performance of associated ritual. Depending on the kind of magic, it may also involve inclusion of some substance to transmit the power from the magician to the object to be influenced, and may be accompanied by offerings.
The reciting of a spell effectively invokes the ancestral spirits. It seems that although anyone may talk to the ancestors, most major areas of concern involve established spells typically ‘owned’ by a particular person or clan. The wording of a spell is usually of great importance, having been passed down by previous generations, perhaps originating in a single named ancestor.
Malinowski (1948:54) writes of the Trobrianders that “the most important element in magic is the spell. … In an analysis of any act of witchcraft it will always be found that the ritual centres round the utterance of the spell. The formula is always the core of the magical performance.” He echoes Powdermaker (1933:298) who writes that, in Lesu (New Ireland), the power of the magic lies in the muttered spell. The rite is of minor significance, and sometimes does not exist, but the spoken words are essential. In Longgu “the verbal formula of the spell is always rhythmical to facilitate learning and subsequent recall, and, as absolute accuracy in repetition is considered essential, it often abounds in archaisms” (Hogbin 1964:86). Similarly, in To’aba’ita, Hogbin (1970a:119) writes that “by far the most important part of the magic is the spell, and the same word, akaloa, is in fact used for both. All spells are supposed to have been revealed in dreams to the ancestors by forefathers still more remote, and their effectiveness is supposed to depend on the accuracy with which they are repeated.” In Samoa, as described by Moyle (1988:73) “the possibility of success required exact recitation of the text and an actual performance of the ritual acts.”
Blackwood (1935:479), however, writes that in northern Bougainville the deed is more important than the words. “While the deed is often sufficient without the word, the word is never found alone, but always in conjunction with some act … Further, no great stress appears to be laid upon the exact reproduction of the form of words.” Firth (1967:203, 206) writes that, although the essence of the magic is the formula, “there is no belief in Tikopia that the form of words is so exact that a slip in the recital will invalidate their effect or bring misfortune upon the reciter … There is a great deal of individual variation in the words recited.”
Although the main focus of the spell will be the desired outcome, persons’ names may be included. In Mekeo, for instance, a spell includes a sequence of names beginning with the name of the originator and ending with the name of the person who taught the spell to the adept (Stephen 1987a:59). In Longgu a spell will contain a list of former owners, the list being proof of the spell’s potency for the present owner; otherwise they would never have continued to pass it down through the generations (Hogbin 1964:87).
The structure is repetitious. Longer spells may be in rhythmical language, being easier to remember, often with figures of speech and far-fetched analogies. Hogbin (1970b:177) offers as an example a spell from Wogeo to make a canoe sail swiftly which describes it as travelling faster than the flight of an eagle. The manner of utterance may also be of importance. Spells are typically uttered inaudibly to prevent anyone else learning them. Some may be sung.
Although it is possible to locate various terms meaning ‘to invoke supernatural intervention’ or similar, a POc reconstruction has proved elusive. Codrington (1891:145) wrote “It is certainly very difficult, if not impossible, to find in any Melanesian language a word which directly translates the word prayer, so closely does the notion of efficacy cling to the form employed.” A number of phonologically similar low-level reconstructions are possible, but it has proved impossible to combine the data into a single cognate set.
| Proto Malaita-Makira | *yaru | [VI] ‘invoke a spirit, make imprecations, put a spell on someone over something’ | |
| Proto Malaita-Makira | *yaruʔ-i | [VT] ‘invoke a spirit, make imprecations, put a spell on someone over something’ (Lichtenberk 1994b: 30) | |
| Proto Malaita-Makira | *aruʔ-a | [N] ‘magic, spell’ | |
| SES | Longgu | aru | [N] ‘generic term for magic, spell’ (Hogbin 1964) |
| SES | Longgu | aru-aru | [V] ‘whisper, murmur, usually behind one’s hands; talk to s.t. like a lime or leaf in your hands as a special protection against spirits so that no one hears what you say’ |
| SES | To’aba’ita | aruʔ-a | [N] ‘kind of malevolent magic, sorcery’; ‘man or woman who practises this kind of sorcery’ |
| SES | Lau | aru | [V] ‘practise black magic, harm by magic’ |
| SES | Lau | aru-a | ‘black magic, sorcery; familiar spirit’ |
| SES | Lau | aru-aru | ‘practise black magic; poison’ |
| SES | Lau | aru-i | [VT] ‘use magic on’ |
| SES | Kwaio | (wane) aluʔa | [N] ‘a sorcerer, one who keeps another’s leavings for malicious purposes’ (wane ‘man’) |
| SES | Kwaio | alu- | ‘prefix for various magic practices’ (e.g. alu-aluʔa ‘magic for raising pigs’) |
| SES | Kwaio | aluʔ-i | [VT] ‘talk to adalo in divination; utter a spell in magic; place a magic spell over’ |
| SES | ’Are’are | aruʔ-i | [VT] ‘to invoke the spirit, to make imprecations accompanied by incantations for recovery from sickness, over food on feasts etc.’ (also aro-a) |
| SES | ’Are’are | aruʔ-a | [V] ‘invoke a spirit over hena (lime) or betelnut for cure of a sick person’ |
| SES | Sa’a | seru-i | ‘use a magical spell over a person or object’ |
| SES | Arosi | aru | ‘to charm, put a spell’ |
| SES | Arosi | aruʔ-i | [VT] ‘to charm, put a spell on s.o.’ |
| SES | Arosi | hai-aru | [N] ‘a charm or spell’ |
| SES | Owa | aru-aru | ‘do magic, do traditional magic practices’ |
| SES | Owa | aru-aru-fa | [N] ‘magic’ |
At first glance, the items listed below form an EOc cognate set. Certainly their glosses justify one. But PNCV *[ta]taro reflects a putative PEOc *taro or, less probably, *taRo. PPn *[talo]talo, on the other hand, would reflect a PEOc *talo. And ’Are’are aro-a might reflect either *taro or *talo, so we cannot tell whether it is cognate with PNCV *[ta]taro or with PPn *[talo]talo. Without making an ad hoc assumption, we cannot reconstruct a PEOc term from these data.
| SES | ’Are’are | aro-a | [VT] ‘to pray over, invoke the spirits over’ |
| PNCV | *[ta]taro | ‘pray, wish for’ (Clark 2009) | |
| NCV | Mota | tataro | ‘to pray; invocation made to the dead’ |
| NCV | Raga | tataro | ‘pray, prayer’ |
| NCV | Mwotlap | tataro | ‘to invoke intercession’ |
| NCV | Nguna | (na)taro | ‘intercession, prayer of request’ |
| PPn | *[talo]talo | [v] ‘to invoke supernatural intervention; pray’; [N] ‘spell, incantation’ (POLLEX) | |
| Pn | Tongan | talo-talo | ‘to cast lots or employ divination’ |
| Pn | Pukapukan | talo-talo | ‘pray, invoke, recant’ |
| Pn | Samoan | ta-talo | ‘pray’ |
| Pn | Samoan | talo-talo | ‘incantation, prayer’ |
| Pn | Tikopia | taro, taro-taro | ‘recite traditional ritual formulae, incl. magical formulae and rituals over kava’; ‘utter Christian prayer’ |
| Pn | Tokelauan | talo | ‘a signal or request or help sent by waving a canoe paddle or a green coconut frond’ |
| Pn | Tahitian | taro-taro | ‘short prayer to the gods’ |
| Pn | Hawaiian | kalo-kalo | ‘prayer’ |
Rituals are intended to incorporate some aspect of the focussed object. The degree of ritual action required, however, varies with context. In Tikopia, where fishing is concerned, it is sufficient for a man to utter appropriate words while the casting of a line provides the necessary action (Firth 1967:201). When the focus is on, for example, success in gardening, the accompanying ritual may assume the greater significance. Even when similar rituals are described in the ethnographic literature, there is little linguistically that can be reconstructed. An exception is with magic associated with treatment of pain or disease. The spraying of some masticated substance such as ginger mixed with saliva from the mouth on to an affected body part is evidently a very old and widely practised ritual treatment right across the Austronesian world. Two similar reconstructions, POc *puRuk ‘to spray spittle etc. from the mouth for magical purposes’ and POc *puRas ‘spray water from the mouth’ are included in vol.5:362–3. Another practice that has been reported in places as far part as Motu (Seligman 1910:167), Bwaidoga (Jenness & Ballantyne 1920:139–141), Kove (Chowning 1989:224), Kwaio (Keesing 1982:118), Gela and Fiji (both Codrington 1891:198) is the manipulation of a body part to isolate something believed to cause the problem so that it may be seized or spat out. POc *samo(s) ‘stroke, massage’ may include this activity as part of its wider meaning (vol.5:363).
Another ritual may be undertaken when a person becomes seriously ill. The illness is thought due to the patient’s soul becoming separated from his body, apparently due to sorcery. At such times a curer will enter into induced sleep, thus permitting his own soul to travel to locate the soul of the unconscious person and return it to the land of the living. Specialised terms for such magical or induced sleep include:
| NNG | Manam | dimate | ‘induced sleep to enable soul of curer to travel to Land of the Dead to retrieve soul of ill person’(Wedgwood 1934-35: 297) |
| MM | Nehan | uelhohou | ‘fever-curing ritual where curer searches in sleep for lost soul of patient’ (uel- RECIP, hohou ‘sleep’)(Glennon and Glennon 2005) |
| SES | Ulawa | maʔahu isuli | ‘magical sleep to trace anyone or find out the cause of an illness’ (lit. ‘sleep and find out’)(Ivens 1927: 345) |
Further examples of this belief are included in chapter 7, but no reconstructions have been possible.
Many spells involve the use of particular substances relevant to the desired effect. Items chosen are thought to mimic, by their properties, the desired result. Love magic may include adorning oneself with the leaves of a sweet-smelling shrub such as Euodia hortensis (PCP *usi) to capture a woman’s affections (Ivens 1927:281, 336). The leaves and bark of particular plants are important in calling on ancestor spirits, while roots are chewed for magical purposes. Ginger (POc *laqia ‘ginger, Zingiber officinale) (vol.3:414) is considered a powerful substance used for both magical and medicinal purposes. Chewing ginger as part of a spell serves to make the words ‘hot’, that is, to augment their potency (Hogbin 1970b:180). It is also a common ingredient in the ritual treatment of disease, generally by the agent spitting the masticated substance on to the afflicted body part. Wild ginger (Zingiber zerumbet) also known as red ginger, is associated with magic and sorcery in Marovo and Maringe, and chewed by the magician in Tangga preparing for war (Bell 1935a:261). It was also wrapped in dried banana leaves and burnt in pots by Motu on their ocean-going canoes to help them go fast and well while on hiri trading voyages (Gwilliam 1982:52).
Other plants associated with magical practices include turmeric, Curcuma longa (POc *yaŋo, vol.3:412), used at Morovo and Kwara’ae, and various leaves including those of Dracaena augustifolia used in Arosi and Sa’a (Ivens 1927:290) and Cordyline fruticosa used in Buka (Blackwood 1935), Marovo (Hviding 2005:118), To’aba’ita (Hogbin 1970a:106), Kwaio (Keesing 1982:189) and Tikopia (Firth 1967:182). POc *jiRi (vol.3:418) refers to both Dracaena augustifolia and Cordyline fruticosa ). Fortune (1963:114–5) lists Cordyline terminalis, an alternative name for C. Fruticosa, as of ceremonial importance over a wide area, from the Admiralty Islands, through Milne Bay, New Britain, the Solomons, Vanuatu, Fiji and Polynesia. Skins of areca nuts, Areca catechu (POc *buaq, vol.3:393) are used in black magic in Sa’a and elsewhere (Ivens 1927:246).
Lime (POc *qapu(R) ‘lime, burnt coral or limestone’, vol.2:64) is also used for both magical and medicinal purposes. Blackwood (1935:477) mentions the many uses of lime (iav) in Petats. Not only is it added to the mixtures used for a number of medicinal and magical purposes, it is also efficacious alone, e.g. rubbed on a man who has been struck by spirits, or a person suffering from a sprain or broken bone. Ivens (1927:195) lists the ceremonial uses of lime (sahu) in Sa’a as (1) in black magic; (2) to induce magic sleep; (3) to restore those who were possessed; (4) in exorcism; (5) for protection of the malaohu [separated for initiation] boys, and of the fighting man against adverse ghosts; and (6) to protect houses from ghostly attack. It may be rubbed or painted on the skin or item to be protected.
In addition, some magical procedures require objects such as physical relics of the dead. In Lukep an ancestral relic [bar] like bone or hair is used to communicate with the spirit world (D’Jernes). A sorcerer in Mekeo will keep with him the relics (bones, teeth, hair) (faŋa ofuŋa, lit. ‘body dirt’) of his patrilineal ancestors, with whom he is in constant communication through dreams, signs, divination and his nightly invocation of them to assist him and watch over the members of his lineage (Stephen 1987a:57). Blackwood (1935:474) describes small bundles called ēto carried as a general protection by adults in Buka which contain relics of a dead relative or of an enemy who has been eaten. In Sa’a, relics of the dead [maŋite] including the skull, hair, fingernail and bone, are kept in a relic case in the corner of a dwelling house where offerings are made (Ivens 1927:178). The magical equipment of a sorcerer may also include oddly shaped stones, pieces of bark or shell, and dried reptile or insect parts (Stephen 1987a:60, Ivens 1927:292), particularly if they show some physical similarity to a body part on which magic is to be performed. (See also Seligman 1910:178ff).
Although offerings are generally made in atonement for perceived transgressions, they sometimes accompany a request to the ancestors for good fortune in a future endeavour, and are also made as a form of thanksgiving following harvest or successful fishing. There is considerable variation from place to place in the degree to which an offering is developed into formal ritual. In Manus, where people consider that the main role of ancestral ghosts is to maintain a moral code by punishing offenders with illness or other misfortune, people pay only to atone for sin, never to promote good fortune from the spirits or to avoid any future misfortune. “Payment to wipe off sin is just; payment to keep a ghost from malice, if it were done, would be simply bribery or tribute” (Fortune 1935:54).
Although offerings to ancestors appear to be a widespread part of ritual in the southeast Solomons and Polynesia, evidence for their existence in western Oceanic communities is very limited. A rare western Oceanic example comes from Malinowski (1935:467). He notes that although the Trobriand belief in spirits and the part they play was vague and shadowy, the ancestral spirits were acknowledged during the breaking of new ground in gardening. While the villagers offer a quantity of special food, usually fish, to the garden magician as ceremonial payment, a small portion is exposed to the ancestral spirits, sacrificially and with an invocation.
Offerings form a significant part of magic ritual in the southeast Solomons both to ensure good fortune and to atone for perceived transgressions. Hogbin (1964:78) describes sacrifices made at Longgu, the most important of which were those preceding dispatch of a fleet of trading canoes, when pigs were offered. Pigs were also sacrificed after an earthquake to prevent further destruction. Ivens (1927:179) describes offering in Sa’a known as uraʔiŋe that were made to ghosts to ensure their goodwill. Offerings of porpoise teeth were tied to the bows of war canoes and on the bows and spears used in divination. They were also put into the relic case found in each dwelling house that already contained the skull or jawbone or tooth of the departed householder and to which further contributions could be made in the event of illness. If desired, a pig might also be burnt in sacrifice on behalf of the sick (p.180). Sacrifice was also made at the launching of a new bonito canoe (p.250). In Kwaio, pigs, known as fōta ‘offerings’ are consecrated to ancestors before being sacrificed in atonement (Keesing 1982:69).
| POc | *uraki | ‘make an offering to the gods’ | |
| PT | Kilivila | ula-ula | [N] ‘an offering made to a ghost [baloma] as payment for magic (often fish)’ (Malinowski 1948: 182) 1 |
| SES | Sa’a | uraʔi | [VI] ‘make an offering to a ghost’ |
| SES | Sa’a | uraʔi-ŋe | ‘offering made to a ghost; relics include porpoise teeth etc. worn by priest round neck when going to battle; if to a ghost shark, thrown into sea’ (Ivens 1927: 179) |
| SES | Arosi | uraʔi | [N] ‘sacrifices at Birubiru rock, of money etc.’ |
| SES | Bauro | uragi | ‘make an offering’ |
| SES | Owa | uraage | ‘make an offering to spirits by s.o. wishing to die’ |
Bauan Fijian distinguishes between offerings made for atonement, i soro, and those made in thanksgiving, i madrali.
In Polynesia Williamson writes (1937:121):
Special ceremonial occasions such as births, marriages and deaths were accompanied by offerings to the gods. After fishing it was frequently the custom to offer share of the catch to the gods, and other important activities such a house-building, the launching of large canoes, and warfare were likewise occasions for the making of sacrifices.
In the following cognate set, the Tongan, Niuean and Samoan reflexes are examples of the specialised vocabulary required for food given to the chiefs who were seen as descended from the ancestral gods.
| PPn | *taumafa | ‘ceremonial food; offering to the gods’ | |
| Pn | Tongan | taumafa | ‘food, drink, smoke (regal)’ |
| Pn | Niuean | taumafa | ‘eat, used to chiefs only’ |
| Pn | Rennellese | taumaha | ‘dedicate food or hail the gods or ancestors’ |
| Pn | Samoan | taaumafa | ‘eat (polite)’ |
| Pn | Tikopia | taumafa | ‘portion of food allocated to a person in a distribution; traditional offering of food to gods or ancestors’ |
| Pn | West Futunan | taumafa | ‘offering’ |
| Pn | Māori | taumaha | ‘spell recited when food offered to gods’ |
| Pn | Hawaiian | kaumaha | ‘offering, sacrifice’ |
A PEOc reconstruction is tentatively proposed for making offerings in atonement. While the Sa’a term refers to ancestral atonement, reflexes in other languages may now be more general terms for compensation.
| PEOc | *soso | ‘to expiate, compensate’ | |
| SES | Longgu | toto | ‘pay compensation’ |
| SES | Sa’a | toto | [VT] ‘propitiate a ghost with sacrifice; pay a fine’ |
| SES | Sa’a | toto akalo | ‘a sacrifice burnt whole or killed and thrown away, pig, dog, cuscus, to remove ceremonial defilement’ |
| SES | ’Are’are | haʔa-toto(a) | ‘propitiate’ (haʔa CAUSATIVE) |
| SES | Kwaio | toto | ‘compensate, pay a fine’ |
| SES | Lau | toto | ‘pay a fine’ |
| SES | Arosi | toto | ‘to pay a fine, give money to be reconciled’ |
| SES | Owa | toto-mara | ‘pay compensation to’ (mara ‘?’) |
| Fij | Bauan | soso-ya | [VT] ‘to give in exchange, replace; atone, expiate’ |
| Fij | Bauan | i-sosoi | [N] ‘thing given in exchange; reparation, expiation’ |
The giving of thanks for the harvest is widely noted as a way of maintaining good relations with the ancestors, and ensuring their support in future endeavours. That mention of first fruits is rare in western Oceanic may be an accidental by-product of our ethnographic sampling, or may be because first fruits are typically not accorded significant ritual there. In Bwaidoga, at harvest, each man would simply place one of his largest yams in the back of his hut to rot, in order to pay the spirits (Jenness & Ballantyne 1920:126). In Kilivila where every stage of gardening activity is preceded by ritual, the garden magician, prior to the gathering of taro and kuvi (large yams), cuts off the top of a taro plant in each holding and places it in his house rafters as an offering to the ancestral spirit. In the third day following, each man pulls up a few taro plants and digs a few yams. These ‘first fruits’ are brought to the village where some are displayed and others placed on the graves of recently dead relatives (Malinowski 1935:166). In Madak, New Ireland, where each man carries out elaborate ritual while cultivating his own taro garden, “a small first fruits feast can be held to celebrate the garden” once the taros are ready (Eves 1998:210).
In the southeast Solomons, Fiji, Polynesia, Micronesia and parts of Vanuatu, the offering of first fruits to the ancestors is typically carried out by priests or village elders on behalf of the community rather than by individuals, as is often the case in western Oceanic communities. Codrington (1891:132) notes that in Gela and Sa’a, when canarium nuts were ripe, no-one might partake until sacrifice of the first fruits. (In Gela the term for ‘first fruits’ is hinava.) In Ulawa, people were free to dig their yams only after the first fruits had been offered by the priest. This offering is called toli uhi [lay/place yams] (Ivens 1927:363). Similarly in Kwaio, when taro and yams were cultivated, a priest was required to make an offering of first fruits from the garden to the ancestral owners, before the living owner of the garden could partake of the food (Keesing 1982:119–121). Separate rituals were undertaken for each. Keesing’s Kwaio dictionary (1975) lists fafiʔalo ‘the shrine, ritual complex etc. involving presentation of first fruits of taro to the spirits’. To’aba’ita differs in that it is the male owner of a new garden rather than the priest who takes one small taro and roasts and eats it by himself: this opens the garden, allowing it to be harvested by other family members. This is the ceremony of gwa lusu abu ‘first fruit of taro’ (Lichtenberk 2008:172). lusu is also the term for ‘first fruits’ in Lau.
Durrad (1940:398) describes the gardening activities on Lo in the Torres Islands in northern Vanuatu.
The time for digging [the yams] is decided upon by the village elders, and there is a certain amount of ceremonial associated with the lifting of the first yam, which is offered as a sacrifice to the ancestral gods. When this first yam has been lifted, and the ritual connected therewith finished, everyone can lift his yams as he requires them.
Petersen (2009:196) describes the situation in Micronesia.
The most extensive, conspicuous and consequential of all Micronesian religious practices are the first fruits rituals. Although known by a wide variety of names – for example, Lamotrek maulmei (Alkire 1974), Chuuk wumwusomwoon (Goodenough 2002:262–5), Marshalls’ ekan (Carucci 1997:168), Pohnpei nohpwei, and Kiribati inagu (Lundsgaarde 1978:71) – they are nearly identical throughout the region.
Although first fruits presentations and feasts may be directed specifically towards the chiefs, ultimately they are a form of homage to the spirits, a way of safeguarding the land from the force of storms and drought. In Micronesia the first fruits rites are performed in stages throughout the seasons for most staple food crops, this as a way of being ever-mindful of the protection of the spirits (Petersen 2009:196).
In Moala (Fiji) the yearly tribute of first fruits was given to the chief “that the land might be prosperous” (Sahlins 1962:319). “In the old days the presentation of first fruits (sevu) was of great magical and religious as well as political significance” (p.343). The village priest, bete, traditionally directed the collection of the yams before they went to the chief. The latter was required to contribute his own yams because the recipient was ultimately the god.
Williamson (1937:121) reports that the periodic offering of first fruits was widespread throughout Polynesia. In Futuna, the main ritual season began with a major feast and with the offering of the first yams to the gods (Kirch 1994b:275). The Maori offered a portion of the first fruits of each season – fish, fowl, and vegetable – to the departmental gods; of birds to Tane, of fish and seaweed to Tangoroa, and cultivated foods to Rongo (Best 1934:69).
Kirch & Green (2001:274, Fig. 9.5) include PPn *quinati (?) ‘first fruits rituals’ in their diagrammatic summary of the ancestral Polynesian ritual cycle, but that meaning may be a narrower example of a more general PPn term for any allocated portion of food.
| PPn | *qinati | ‘share, allocated portion of food’; ‘first fruits rituals?’ (Kirch and Green 2001) | |
| Pn | Tongan | ʔinasi | ‘share, allotted portion, quota’; ‘presentation of food to the Tuʔitonga [supreme chief] in a way that came to be regarded as inconsistent with the Christian religion’ (in old Tonga) |
| Pn | Rennellese | ʔinati | ‘food share’ |
| Pn | Samoan | inati | ‘part, portion, share, first fruits’ |
| Pn | Tokelauan | inati | ‘group of people who receive a share from a community-owned asset; the share received’ |
| Pn | Māori | inati | ‘portion, share of food at a feast’ |
Anyone in a community would be able to call on their ancestors for such matters as success in love or fishing. But communities generally have recognised purveyors whose assistance may be bought for more serious concerns. Spells are typically handed down from parent to child, and spell owners who have a reputation for successful outcomes, particularly in areas such as childbirth or control of the weather will be much sought after.
A difficulty in dealing with the role of practitioners in magic lies in the somewhat indiscriminate use in ethnographies and wordlists of the terms ‘priest’ and ‘sorcerer’, and the related role of shaman. All may have a claimed special ability to communicate with the gods. Although sorcery is concerned with doing harm, some so-defined sorcerers have powers that serve positive ends. Stephen (1987a:44) writes that in Mekeo “a sorcerer may well spend far more of his time in healing than in doing harm.” In contrast, those serving as priests in the southeast Solomons and Polynesia do not deal in harmful magic. They fill a pre-determined role in the community, often inherited, and have regular ritual duties to undertake. Effectively, their duties are to maintain communication with the ancestors and ensure their blessings. In return, they may be accorded certain privileges. A shaman, sometimes referred to as a spirit medium, usually fills a lesser role, underpinned by the ability to fall into a trance, during which he may transmit answers from the spirits as to certain queries, perhaps identifying the source of a wrong-doing. It should be noted however that these roles are not clearcut. The ability to perform while in a trance is also noted in people including the sorcerers in Mekeo and those serving as priests in Fiji and Polynesia (see Stephens 1987a, Williams & Calvert 1859 and Williamson 1937 below).
In western Oceanic communities there is little evidence of a specialised priestly role such as is found in the southeast Solomons and Polynesia. Superior practice of magic appears to lie in western Oceanic either with the headman or with specialist magicians like the garden magician of Kilivila, who fills a role in the hierarchy second only to the chief, with duties that also serve social and administrative ends (Malinowski 1935:66). In Wogeo, for instance, the kokwal (hereditary clan headman) “has a far more extensive knowledge of magic, including sorcery, than other people. … Apart from the kokwal there is little or no specialisation.” His area of expertise is with matters that concern the whole village, such as weather magic and for success in trading (Hogbin 1935a:319). In Mekeo there are, among practitioners of harmful magic of all kinds, departmental experts who can summon stronger spirits than non-specialists and are in command of more powerful medicines (Hau’ofa 1981:240). In Dobu, although ownership of incantations is tightly held by family lines, reputation may be built upon demonstrated skill in a particular area (Fortune 1963:135). Chowning (1973:65) writes that “it is true that Melanesia generally lacks full-time specialists of any sort. On the other hand, part-time priests, often the heads of kin groups, who make sacrifices to ancestral ghosts are not uncommon in eastern Melanesia.” Presumably she is here including the southeast Solomons.
In the southeast Solomons are communities which include within their descent groups a person who officiates at the offering of sacrifices to the gods. Wordlists typically gloss the native term for such people as ‘priests’. In northern Malaita they are referred to as fata-abu (‘speak sacred’) (Keesing 1982:11). In Kwaio each descent group has a ‘shrine-man’ (wane naa baʔe) [baʔe ‘shrine’], ideally agnatically descended from the founding ancestors, who, although not a full-time religious specialist, acts as the principal officiant in sacrifices to ancestors (Keesing 1982:19, 87–91). Hogbin (1970a:106) writes that in To’aba’ita “each cemetery has its own priest [aofia] who has to give his approval before any offering can be made to the spirits of the persons buried there. The office is held by hereditary right, being transmitted from father to son, but it is by no means a full-time job, and apart from his special position at ceremonies the priest lives the life of an ordinary man.” Lau also has aofia whose duties are to officiate at sacrifices (Fox 1974). In Sa’a, the priest (ora-ora) officiates at the sacrifice of burnt offerings made to ancestors, and shares some of the privileges due also to the priest, such as being exempt from the obligation to make a return for gifts received. As in To’aba’ita and Kwaio, the position is hereditary (Ivens 1927:8, 242).
Codrington (1891:127) writes of Vanuatu:
There is no priestly order, and no persons who can properly be called priests…. If the object of worship … is one common to the members of a community, the man who knows how to approach that object is in a way their priest and sacrifices or them all; but it is in respect of that particular function only that he has a sacred character.
Although there is little information on the practice of magic in Micronesia, Linton (1926:175, 176) notes that “in all the groups there were individuals who combined the duties of priest and shaman, serving the gods, curing the sick and working magic. … Throughout Micronesia offerings were made and rites performed at uncarved stones believed to be the dwelling places of spirits.”
Fijians had bete, men with special powers for communicating with the gods. Williams (Williams & Calvert 1859:178) writes that they “exercise a powerful influence over the people, an influence which the Chiefs employ for the strengthening of their own, by securing the divine sanction for their plans.” The priesthood was generally, but not invariably, hereditary, a vital part of a bete’s role being the ability to communicate the wishes of the god he serves while in a trance. As Williams puts it (p178), he needed to “shake well and speculate shrewdly.”
Kirch & Green (2001:249) write that “the evidence is compelling that in ancestral Polynesian societies the principal ritual leaders were simultaneously the main secular leaders, the *qariki.” However, a separate functional class of priests indicated by PPn *taaula, developed largely in hierarchically elaborated societies like those of Tonga, Samoa and Hawaii.
Williamson (1937:288) describes the religious function of the taaula
as being personally to organize and reinforce the religious beliefs and sentiments of the people. They were the holders of the ancient traditions, who were well versed in the impressive legends of the gods and creation. Ritual too was their province, and they directed ceremonial behavior along lines pleasing to the gods and spectacular to the worshippers, common participation in which served to keep alive religious sentiments. When inspired by the gods they manifested abnormal or unusual forms of behavior which provided visible evidence of their close communion with supernatural forces. … By making sacrifices and directing ritual they pleased the gods, and made them favourably inclined towards their worshippers; by prayer and intercession they sought for blessings or the aversion of evil.
| PPn | *taaula[-qatua] | ‘priest, medium, shaman’ (POc *qatuan ‘deity, supernatural being’) | |
| Pn | Tongan | taula | ‘priest or priestess’ |
| Pn | Niuean | taula-atua | ‘shaman or priest of heathen times’ |
| Pn | Rennellese | tauga | ‘a medium, one possessed; a prophet’ |
| Pn | Pukapukan | taula-atua | ‘a person who can perform miracles or foresee the future’ |
| Pn | Samoan | taula(aitu) | ‘priest, only with reference to the old religion’ (Williamson 1967:407; POc *qanitu ’spirit of the dead) |
| Pn | Rarotongan | taura-atua | ‘sorcerer, priest of the ancient gods’ |
| Pn | Tikopia | taura-atua | ‘traditional spirit medium’ |
| Pn | Māori | taaura | ‘priest who accompanies an army’ |
| Pn | Hawaiian | kaaula | ‘prophet, seer’ |
Some Eastern Polynesian languages use reflexes of *tafuŋa, a variant form of PPn *tufuŋa ‘expert, skilled craftsman’ to refer specifically to those with priestly duties.
| Pn | Tahitian | tahuʔa | ‘expert craftsman’; ‘specialist in magic, primarily a spirit medium and curer’ |
| Pn | Tuamotuan | tahuuŋa | ‘expert; priest’ |
| Pn | Māori | tohuŋa | ‘skilled person’; ‘wizard, priest’ |
| Pn | Hawaiian | kahuna | ‘priest, minister, sorcerer, expert in any profession’ |
In practice, most magic is performed by men, since nearly all the tasks for which it is appropriate are those normally done by men. Women priests were not unknown in Polynesia. However, the magical procedures of protection and healing are often more highly individualised than those of production, and across the Oceanic world these may often be performed by women.
The ability of some people to fall into a trance-like state as a means of accessing the spirit world is reported from many communities. Ethnographers may refer to such practitioners as shamans or spirit mediums. They may enter a state of disassociation or semi- consciousness, usually ritually induced, as a form of self-hypnosis, perhaps preceded by prolonged fasting. In this condition they are recognised as possessed by the god, and believed capable of super-human powers. A person’s utterance is then thought to be the direct voice of the ancestor or god.2 Elsewhere, the role of shaman varies greatly.
Sometimes a medical condition expressed by semi-consciousness or other abnormal behaviour will be interpreted as due to the possession by spirits.
Fortune (1935:9) summarises belief in Manus. “The Manus have the familiar concept of … a certain kind of soul or vital essence in the living, … the concept that in trance or swoon this soul or essence approaches the ghosts.”
In Mekeo, it is the sorcerer himself who, like a shaman in some respects, communicates with the spirits in visions, trance and dreams, although he avoids the more overt displays of spirit possession often associated with the latter (Stephen 1987a:66–67).
Keesing (1982:108) writes that although in Kwaio there is no culturally defined role for a person with shamanistic powers, “the rare man with such psychic bent … is attributed special sacredness in life.”
Across Micronesia there are spirit mediums – individuals through whom the spirit of the dead are able to speak to the living, while the living are, in turn, able to petition ancestral spirits for information and assistance (Petersen 2009:195).
Sahlins (1962:359) describes those he calls well-rounded shamans (dau ðaka wai) in Moala Fijian “who not only cure, sorcerize and counter-sorcerize, but also find lost things and foretell events.”
In Fijian and parts of Polynesia, certain people identified by their ability to enter trance-like states are known by terms reflecting Proto Central Pacific *waga (or PNPn *waka-qatua):
| PCP | *waga | ‘spirit medium’ (from POc *waga ‘canoe’?) | |
| Fij | Wayan | waga-waga | ‘spirit medium, person through whom a spirit speaks’ |
| Fij | Bauan | waga-waga | ‘the body assumed by a kalou-vū [ancestral spirit] for purposes of self-manifestations’ |
| PPn | *waka | ‘medium or bodily abode of a god’ | |
| Pn | Tongan | vaka | ‘bodily abode of a supernatural being’ |
| Pn | East Uvean | vaka | ‘spirit medium’ |
| Pn | Māori | waka | ‘medium of an atua’ |
| PNPn | *waka atua | ‘spirit medium’ (POc *qatuan ‘deity, supernatural being’) | |
| Pn | Anutan | vaka atua | ‘spirit medium’ |
| Pn | Rennellese | baka ʔatua | ‘representative of the gods’ |
| Pn | Tikopia | vaka atua | ‘spirit medium’ |
One role commonly employed by shamans is that of identifying the person or spirit responsible for perceived harm. Some do it by divination, testing hypotheses by way of yes/ no questions, but there is little commonality in the methods recorded. Fortune (1963:154) describes ways of divining in Dobu by water-gazing or crystal gazing, Chowning (1987:165) describes a pole-in-house technique whereby answers are sought as to persons responsible for deaths. In Manam, where serious illness is believed due to theft of the spirit by sorcery, Wedgwood (1934-35:293) describes various ways to divine whether the patient will live or die. Stephen (1987:56) mentions divination as another means for a Mekeo sorcerer to communicate directly with the spirits. Hogbin describes a kind of diviner in Longgu, the toʔiai, who possesses spells that enable him to conjure the dead man’s soul into an areca nut. By means of various yes/no questions and the resulting movement of the areca nut the guilty party is located (1964:58). In Kwaio, dried cordyline leaves called felo are commonly used in divination (Keesing 1982:113). In Sa’a, various implements including bows, spears and bonito rods may be utilised (Ivens 1927:345). Tonkinson (1981:82) describes individuals in SE Ambrym known as lele who claim the ability to discern by divination whether or not particular illnesses are sorcery-related. Williams (1858:228) describes various methods of divination used in Fiji to decide between two options.
Although the belief that certain people have the ability to intercede with the gods while in a trance or other inexplicable state appears to be widespread, as is the practice of divination, no reconstructions other than the above have been possible. Wordlists rarely include terms such as ‘shaman’ or ‘spirit medium’ or ‘trance’ or ‘divination’.
There is widespread belief that magic will be strengthened or made effective if the protagonist undergoes certain privations before he can carry out his activities. Similar privations must be undertaken by anyone about to sacrifice to the gods. As summarised by Eves (1998:58) in Madak, “all of the powerful forms of magic, and some minor forms, require the magician to undergo a regime of fasting and sexual abstinence.” To illustrate, we offer examples from three Western Oceanic subgroups, NNG (Wogeo); PT (Mekeo, Roro); and MM (Tabar, Petats).
In Wogeo, most men think it advisable to refrain from sexual intercourse and from eating nuts for a full twenty-four hours if they are expecting to perform gardening or fishing magic (Hogbin 1970b:181).
An extreme form of preparation is practised by a sorcerer in Mekeo, who must be in a constant state of ritual preparedness. Stephen (1987:62) points out that the inflicting of serious illness, injury and death differs from other magic only in that the sorcerer deals with more dangerous entities … and he must prepare himself more stringently for the task.
Sexual abstinence is the most important restriction and this may be necessary for only a few days, several weeks or many months. It is also essential not to immerse the body in cold water, wash in it or drink it. Hot water must be used for washing; and in the most rigorous forms of gope [the magical act] no washing at all is permitted. Only hot liquids may be drunk; and under rigorous gope one drinks as little as possible. … What little food is consumed must be taken with plenty of ginger or chilli to make it hot. This regime is said to render the adept’s body light and hot and dry.
Seligman (1910:292) describes preparation for a wallaby hunt in Roro where “not only is co-habitation forbidden to the expert, but he may not eat food cooked by his wife or any other woman, he may not eat yams, nor the flesh of wallaby, nor pig, though he may eat the flesh of the kangaroo-rat and drink the milk of unripe coconuts which have been more or less roasted.”.
In Tabar, a fishing net magician neither eats nor drinks during three days of rites before initial use (Groves 1934a:449). In Petats, when bonito magic rites are performed, the principal performer must neither eat nor sleep until the rites are over, and along with all who take part in the bonito fishing, must abstain from sexual intercourse. (Blackwood 1935:480).
No examples have been located from Eastern Oceanic subgroups.
Sorcery here refers to magic used for harmful purposes. In its most extreme form, sometimes referred to as black magic, it was used to cause a person’s death. Such knowledge was almost never admitted to. It was the most secret of possessions. As dangerous knowledge, it was essential that it be kept within the hands of a chosen few. It was strongly discouraged by missionaries as incompatible with Christian beliefs. In British New Guinea from the 1880’s on, there was also a very real fear of the results of government interference, for sorcery became an indictable offence (Seligman 1910:278). It would seem that in general it is desirable that those who profess knowledge of sorcery do not draw attention to their powers, lest it leave them open to retaliation. Sorcery works best in a climate of rumour and innuendo. Occasionally, however, as described by Seligman (1910:279) in Roro, “a sorcerer may … be regarded generally as real protection, for, besides being able to thwart the acts of sorcerers of other villages, the latter will, it is supposed, refrain from hostile magic in order not to provoke reprisals”, while Hogbin (1964a:57) noted that the Longgu of Guadalcanal attributed a knowledge of sorcery to most village headmen and many elders, “to the extent that man’s identification as a sorcerer was a measure of his social distinction.”
Sorcerers may be the most powerful people in a community, their activities serving purposes for both good and ill. Their legitimate role may lie in regulating social life, a form of coersive social control. Hogbin (1935b:18) writes that in Wogeo, yabou (black magic) helps to ensure that individual rights are respected and obligations carried out, and that the chief men are obeyed. Disadvantages lie in that it occasionally leads to murders and tends to increase local hostilities. Firth (1967:211) defines the reasons for sorcery from the perspective of Tikopia as “mainly those of economics or of personal status: desire for land; wish to punish for theft of food; jealousy of competitive achievement; resentment at a personal affront.”
Although there is widespread belief across communities (e.g. Roro (Seligman 1910:279), Mekeo (Stephen 1987b:252), Kove (Chowning 1987:157), Longgu (Hogbin 1964:58)) that except in the case of infants and very old folk and those killed in warfare, death is the result of sorcery, much of sorcery consists of attack and counter-attack, without resulting in death. Kin of sick individuals may approach those believed to have the power to cause harm and offer them payment to remove their spell (Stephen 1987:43). A sufferer may have a good idea of someone he or his close kin have offended, or know of those who hold a grudge against him, and he may have a good idea of those with the ability to practise sorcery and identify the likely sorcerer. Chowning (1989:224) notes that in Kove, if an illness was drawn out, it was assumed that the sorcerer wanted to be bought off rather than to kill the victim. There it was common practice to approach a number of known sorcerers with offerings of shell money and to ask them to undo the spell.
Much sorcery was carried out for personal or private reasons. In Dobu, it was used “for collecting bad debts and enforcing social obligation, in vendetta to avenge one’s own sickness or one kinsman’s death, to wipe out any serious insult” (Fortune 1963:175). Chowning writes that in Kove sorcery threats were used to enforce many prohibitions, such as those that prevented a woman from showing disrespect for the rituals of the men’s houses, and they also made the younger men obey their elders. Chowning considers that it was threats of sorcery rather than powerful spirits or gods that upheld general moral standards in Kove (1989:225).
Sometimes, as in the Trobriands (Malinowski 1922:75), Mekeo (Stephen 1987b:270) and Kaoka/Longgu (Hogbin 1964), a sorcerer will reserve the right to heal by counteracting sorcery imposed by another, although more often this is done by specialist healers. He may also be prevailed upon to remove his own spell by means of payment by kin of the affected person.
While lesser illness may have been attributed to attack by spirits or to the breaking of a taboo, severe illness was more likely to be considered due to sorcery involving theft of the victim’s soul. As Stephen describes it, the soul journeys of the Mekeo sorcerer are the means of not only ensnaring the souls of others but also of restoring them (p.270).
As might be expected, no POc terms are reconstructable. A term for ‘sorcery’ and related ‘sorcerer’ is reconstructable for PPT.
| PPT | *baravu | ‘sorcery’ | |
| PPT | *(tau)baravu | ‘sorcerer’ | |
| PT | Dobu | barau | [N] ‘sorcery’; [VI, VT] ‘kill or afflict by sorcery’ |
| PT | Dobu | (to)barau | ‘sorcerer’ (on to- see vol.5, §2.2.1.2) |
| PT | Molima | balawu | ‘sorcery’ |
| PT | Molima | (to)balawu | ‘sorcerer’ |
| PT | Wedau | baravu | [N] ‘sorcery’; [V] ‘to practise sorcery’ (Seligman 1910) |
| PT | Gumawana | balau | ‘sorcery’ |
| PT | Motu | ba-balau (tau-na) | ‘sorcerer’ (for exp. †balahu) |
| PT | Motu | ba-balau (kara-na) | ‘sorcery’ (kara ‘conduct, customs’) |
However, a search through available sources throws up the terms listed below. They all bear some resemblance to PPT *baravu above. Assuming regular sound correspondences, the PPT reconstruction suggests a POc †*ba(r,R)apu, but none of the forms below exactly reflects this. Lou pʷalop comes close, but Lou -l- reflects POc *-l-, not POc *-r- or *-R- (Blust 1998b). Intriguingly, the PT language Kilivila has apparently replaced this same consonant with -g-. Similar issues affect the cognacy of the other items below, yet the likelihood of chance resemblances of several trisyllables with similar senses is close to zero. A plausible inference is that the word itself was considered to have evil power and was distorted for that reason.
| Adm | Lou | pʷalop | ‘sorcerer’s magic’ |
| NNG | Mapos Buang | paɾaʁək | ‘sorcery, black magic’ |
| PT | Kilivila | bʷagau | ‘sorcery’ |
| PT | Kilivila | (to)bu-bʷagau | ‘sorcerer’ |
| NCV | Lewo | pʷuruwap | ‘sorcerer’ |
cf. also:
| NCV | Raga | barahuva | ‘salvation’ |
Raga barahuva appears under ‘cf. also’ because it is glossed ‘salvation’: is this a Christian missionary re-use of an old word?
The following reconstruction is given a tentative gloss. It may be connected with the PPT *baravu above.
| PWOc | *bara | ‘poison, magic employed to affect another person’ | |
| NNG | Lukep | bar | [N] ‘poison, magic; ancestral relic’ |
| NNG | Takia | bar | ‘incantation, blessing, religious ceremony, song’ |
| MM | Nehan | (uel)bara(ŋa) | [N] ‘power; poison, magical power, love magic, seducing charm, war magic; hot as a pepper, strong’ (uel- RECIPROCAL, -ŋa NOMINALISER) |
Without further reflexes it is impossible to know if the following form, reconstructable to PNGOc, relates only to black magic or to spells in general.
| PNGOc | *nabʷa | ‘a spell’ | |
| NNG | Manam | nabʷa | ‘spell usually resulting in rapid death’ (Wedgwood 1934-35) |
| PT | Dobu | nabʷa(sua) | [N] ‘magic spell’; [V] ‘to utter magic spell’ (-sua ‘?’) |
| Adm | Titan | nam | ‘sorcery, magic, spirits’ |
Examples from NNG, PT, MM and SES show consistency of belief regarding the methods used by those wishing to inflict harm on another through use of leavings.
In Manam, “when a man wishes to work dzere against an enemy, he takes something which has been in close contact with the latter – it may be a piece of chewed areca nut, some lime, a piece of clothing or some hair clippings” (Wedgwood 1934-35:71). In Dobu, “personal leavings may be remains of food, excreta, footprints in sand, body dirt, or a bush creeper with a malevolent charm first breathed into it which the sorcerer watched his victim brush against and which he subsequently took to his house to treat further” (Fortune 1963:150). Eves (1998:63) writes that in Madak “personal leavings sorcery involves the sorcerer manipulating the victim’s personal exuviae (excreta, hair, fingernails etc.) or items which the victim has used or touched (such as a discarded areca nut shell or the soil from a footprint).”
In Sa’a,
the things used for magical charms to work harm with were skins of areca nuts which a person had eaten, cutting of hair or nails, excrement, spittle, earth on which the intended victim had trodden, the strip of coconut leaf with which he had rubbed himself down after bathing, fragments of his food, the afterbirth of children. In most cases the object was breathed upon in order to impart virtue to it. Certain things were considered as possessing in themselves the power to do harm, and an incantation was said over such charms. Lime, ginger and dracaena leaves were the most commonly used of these. (Ivens 1927:324)
Our efforts to identify the kind of magic practices performed in POc times are based on descriptions from ethnographies. These show considerable resemblances, but the terminology of magic is more diverse, making reliable reconstruction difficult. Where communities carry out comparable rituals they do so either because they are continuing the practice of their forebears, or because the practices are borrowed or because they have been independently innovated. Borrowings in matters magical could be presumed minimal in view of the secrecy surrounding them in practice, at least prior to western contact. But when such beliefs as that in which a curer could travel in dreams to recover the soul of an ill person are identified across subgroups, and are unrelated to any comparable Christian belief, then evidence favours the conclusion that this was the practice of their forebears.
Furthermore, over time and place, the very belief system can undergo reinterpretation, with communities modifying beliefs in such things as the role of the ancestors and the purpose of traditional rituals. As an example of the former, Hogbin (1935a:330) writes that in Wogeo the people believed that the ancestors were barely involved there in daily life. “The spirits of the dead are able to cause the death of small infants, but otherwise they are completely powerless. It follows from this that as reciprocity is the keynote of native life, sacrifices are held to be entirely superfluous and are never in fact carried out.” As an example of the latter we can look to the tendency in places for offerings to be seen no longer as primarily to the ancestors, but rather to the aggrandisement of the occasion, as in Mekeo (Hau’ofa 1981:71) or to the giver, as in To’aba’ita (Hogbin 1970a:105). But when offerings are made to the magician or priest, apparently as thanks for their intervention, we still find that as in the Trobriands (§8.2.4), in Micronesia and in Moala Fiji (§8.2.4.1), some part is set aside as a form of homage to the ancestors.
Nonetheless, the practice of magic has many similarities across the Oceanic world. The following apply consistently across subgroups and are thought likely to accord with the practice of POc society.