Rethinking "Mana"

Author(s): Roger M. Keesing

Source: Journal of Anthropological Research, Vol. 40, No. 1, Fortieth Anniversary Issue 1944-
1984 (Spring, 1984), pp. 137-156
Published by: University of New Mexico
Stable URL: http://www.jstor.org/stable/3629696
Accessed: 19-11-2015 10:07 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

University of New Mexico is collaborating with JSTOR to digitize, preserve and extend access to Journal of Anthropological
Research.

http://www.jstor.org

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKINGMANA

Roger M. Keesing
Departmentof Anthropology,The AustralianNational University,
GPOBox 4, Canberra,ACT 2601

Comparativedata are assembled to suggest that in Proto-Oceanic,mana was canonicallya stative

verb meaning 'be efficacious, be successful,be realized, "work."' Wheremanawas used as a noun,
it was (and in most daughterlanguagesis) not a substantivebut an abstractverbalnoun: 'efficacy,'
'success,' 'potency.' Anthropological misinterpretationsof mana from Codringtononward as a
medium of power represent,for most daughter languages,pervasivetranslationerror.Mana is a
condition, not a "thing": a state inferred retrospectivelyfrom the outcome of events. Where
mana was used as a substantive,it seems to accompany the emergenceeitherof extremepolitical
hierarchyand the sanctity of chiefs (easternPolynesia)or secret religiouscultism (partsof eastern
Melanesia).

EARLY IN THIS CENTURY mana became part of the metalanguageof anthropolo-

gy. Following Codrington and ethnologists of Polynesia, anthropologists assumed
that mana in Oceanic religion was a kind of invisible medium of power, a spiritual
energy manifest in sacred objects, a potency radiated by humans.
Yet this Codringtonian view of mana rested on very insecure ethnographic
foundations. Codrington himself, in his initial explication on mana to the Anthro-
pological Institute of Great Britian and Ireland (1881:278-79) had conceded that
It would be very difficult to ascertainwhether ... mana ... is thought to originatein a con-
nection with . . . spiritual beings. The notion conveyed by the word... is vague, and the
origin of the power not likely to be clearly conceivedin the native mind.
It was not until a decade later, when debates about the evolution of "primitive"
religion had reached their peak, that the view of mana as animatistic spiritual energy,
as power manifest in things and persons, came clearly into focus: mana was, as
Codrington (1891:118-19) put it, "a power or influence," which "attaches itself
to persons, and to things, and is manifested by results which can only be ascribed
to its operation."
When the nature of mana in "primitive" thought was still genuinely at issue
among such armchair students of Oceanic religion as Lehmann, Marett, Hubert,
and Mauss, the debates focused partly on the elusive facts of language: mana seemed
to be used not only as noun, but also as adjective and verb. Hubert and Mauss
noted (1902-3:108, as translated in Firth 1940:484) that mana is "not only a
force, a being; it is also an action, a quality and a state. In other words, the term is at
once a noun, an adjective and a verb." Firth (1940:484-85) commented that "this
seeming grammaticalconfusion has been responsible for much laborious theorizing.
The elaborate arguments . . . seem to turn largely on the question as to whether it is
more nearly correct to say that an object is mana or has mana."
When these issues were being debated at the turn of the century, little was
known about Oceanic languages-other than their fluidity, which seemingly allowed
terms like mana to flow about in a kind of conceptual soup. When the debates
were resolved in favor of mana as a noun, labeling a diffuse spiritual energy or
power, it was more by virtue of rhetorical persuasivenessand the sheer intelligibility
of such an imagined medium of spiritual potency to European philosophical imagi-
nation than because of solid textual, linguistic, or ethnographic evidence. But for
137

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
138 JOURNAL OF ANTHROPOLOGICALRESEARCH
subsequent generations of anthropologists, mana has been a "thing"; a substantive
if not quite a substance, something manifest in objects, something people had more
of or less of.
Three-quartersof a century later, we finally know a good deal about the struc-
ture of Oceanic languages;and despite the wholesale destruction of Oceanic religions
by Christianity, we have further ethnographic evidence on mana. I will suggest,
drawing on linguistic and ethnographic evidence, that the Codringtonianinterpreta-
tion of mana is deeply flawed.1 The linguistic doubts about mana as a substantive
were well taken, and the Codringtonianresolution was fundamentally erroneous.
Yet anthropological orthodoxies have deeply influenced and often distorted the
way subsequent ethnographers have understood and translated mana. We have
rendered folk concepts into Codringtonianterms, often doing violence to indigenous
usages. Rethinking mana demands a critical hermeneutics in which cultural transla-
tion is cast deeply in doubt.
The first steps toward this hermeneutic reassessment must take us to the lan-
guages of Oceania, their interrelationshipsand some aspects of their grammar.

THE LINGUISTICBACKGROUND

Let me begin by making my claims explicit. Mana, I argue, is in Oceanic lan-

guages canonically a stative verb, not a noun: things and human enterprises and
efforts are mana. Mana is used as a transitive verb as well: ancestors and gods mana-
ize people and their efforts. Where mana is used as a noun, it is (usually) not as a
substantive but as an abstract verbal noun denoting the state or quality of mana-ness
(of a thing or act) or being-mana(of a person). Things that are mana are efficacious,
potent, successful, true, fulfilled, realized: they "work." Mana-nessis a state of
efficacy, success, truth, potency, blessing, luck, realization-an abstract state or
quality, not an invisible spiritual substance or medium.
My claim requires twofold linguistic explication: on the one hand, a sketch of
Oceanic languages and their subgrouping (for Codrington (1891:119) erred in
surmising that "the word [mana] is common ... to the whole Pacific"; and on the
other, a glance at the grammatical classes and derivational processes of Oceanic
Austronesian languages.
All Austronesian languageseast of 1360 E apparently comprise a single subgroup,
now conventionally labeled Oceanic (OC). Daughter languages of Proto-Oceanic
have not been subgrouped with any certainty. But one large putative subgroup
labeled by Pawley (1972) EasternOceanic (EO) is of special interest. These languages,
strikingly conservative of OCgrammaticalpatterns and lexicon, include the languages
of northern Vanuatu, Fiji, Rotuma, Polynesia, apparently the southeast Solomons,
and "Nuclear Micronesia" (the Marshalls, Kiribati, and the central and eastern
Carolines).
Mana and cognate forms turn up in a few Oceanic languagesoutside the putative
EO subgroup, in the western Solomons and probably the Loyalties and one of the
Massim islands off the Milne Bay coast of Papua New Guinea. Thus a form of mana
will probably have to be reconstructed for POC. The attestations in languagesof the
western Solomons provide important evidence on mana in historical perspective.

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 MANA
RETHINKING 139
Early students of Oceanic languages struggled with the seeming amorphousness
of their grammars:the same word, in root form, often seemed to function as noun,
adjective, verb, and sometimes adverb. This problem looms large in the early in-
terpretation of mana. Modem grammatical studies have greatly clarified matters,
however. There are some forms reconstructed for POC that in root form fit into two
or more form classes. Thus *ngmane is 'male' (as adjective) and 'man' (as noun);
*mate is 'be dead/comatose' (as stative verb) and 'die' (as intransitive verb). But
while root forms themselves may be semantically and syntactically multivalent, in
sentences their function is unambiguous. A verb, for example, is marked as such
by a preceding pronominal form that introduces the verb phrase.
To understand the evidence on mana we need to look at verbs in Oceanic lan-
guages. Verbs describing states, stative verbs, are a major class in most daughter
languages. Thus in Gela (southeast Solomons): mbou, 'be hard'; para, 'be hot';
puku, 'be true'; tambu, 'be sacred'; uto, 'be good'; kama, 'be big'.2 Stative verbs
are distinguished from active verbs. Verbs in POC were used in root form in in-
transitive constructions; in transitive constructions they were canonically marked
with transitive suffixes. Thus in Kwaio (southeast Solomons), aga, 'look'; aga-si, 'see
(something).' Whereasmany semantic classes of verbs could be used either intransi-
tively (in root form) or transitively (with transitive suffixes), some semantic classes
were canonically intransitive. Thus, again in Kwaio, leka, 'go'; to'oru, 'stay, live.'
In most daughter languages of POC, intransitive verb roots are converted into
abstract verbal nouns by adding a nominalizing suffix. Thus in Kwaio, to'oru, 'live,
stay,' becomes to'oru-ngaa, 'living'; oso, 'eat,' becomes oso-nga, 'meal'; and fata,
'speak,' becomes fata-nga, 'speech, talking.' Oceanic languages also create abstract
verbal nouns from stative verbs, forms we can usually translate into English with
something like "(adjective)-ness." Some daughter languages use the same nominal-
izing suffix that creates verbal nouns from intransitive verbs. Thus in Kwaio, fii,
'be sick, be in pain,' becomes fii-nga, 'illness, pain'; tegela, 'be strong,' becomes
tegela-ngaa,'strength';le 'a, 'be good,' becomes le 'a-nga,'goodness, virtue.'
Other daughterlanguageshave two classes of statives: one forming abstract verbal
nouns by affixing a nominalizing suffix to the root, the other forming abstract
verbal nouns with zero morphological marking. Thus in Gela (Solomons), (class I)
tambu, 'be sacred,' becomes tambu-ga, 'holiness,' and uto, 'be good,' becomes uto-
ga, 'goodness'; while (class II) mbou, 'be hard,' becomes mbou, 'hardness,'and rusu,
'be sad,' becomes rusu, 'sadness, grief.'
Many daughter languages use only the second pattern (zero marking)for stative
verbs, whereas they use a nominalizing suffix for intransitive verbs. Thus in Bugotu
(Solomons), tutuni, 'be true,' becomes tutuni, 'truth'; thaba, 'be great,' becomes
thaba, 'greatness'; and thangga, ' be confident,' becomes thangga, 'confidence.'
The distributional facts suggest that in POC there may have been a class of stative
verbs marked with nominalizing suffixes when used as verbal nouns and a class of
stative verbs used with zero marking as verbal nouns. Daughter languages have
either maintained these subclasses or neutralized the distinction, universalizing the
derivational process by using either zero marking or the nominalizing suffix for all
stative verbs.
We can now come back to mana. The distributional and linguistic evidence
suggests that:

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
140 JOURNALOFANTHROPOLOGICAL RESEARCH
(1) Manain POCwas in root form a stative verb (with meanings of 'be efficacious,
be potent, be true, be realized, be successful, "work"').3
(2) As stative verb, it fell into the subclass forming abstract verbal nouns with
zero morphological marking; mana as abstract verbal noun in POC apparently
carried meanings of 'efficacy, potency, success, realization, luck' (with the
assumption that these entailed both support or activation by ancestors or gods
and proper human performance and skill). That is, mana as verbal noun labeled
states of 'mana-ness.'It would seem that in conventional metaphors, things and
people could be spoken of as "havingmana-ness"and as "being mana."
(3) Mana in POC seems also to have functioned as a verb (with god/ancestor as
subject), carrying meanings of 'support, protect, empower'; used with a transi-
tive suffix it meant 'mana-ize (us)'; used in root form as verbal imperative, as an
invocation in prayer or magic, it was like 'Amen.'4

THE LINGUISTIC EVIDENCE

First, it is revealing to look at two clusters of daughter languages, widely sepa-

rated in time though not in space, that provide crucial evidence on mana. They
are the languages in which the derivational process for forming abstract verbal
nouns from stative verbs has been generalized by giving all such verbal nouns a
nominalizing affix. One cluster comprises languages of Malaita, in the southeastern
Solomons.5
All Malaita languages use a reduplicated form of mana (either mamana or the
metathesized nanama) as stative verb. When used as nouns, these forms are always
marked with the nominalizing suffixes used to create abstract verbal nouns from
statives. In many of these languages, mamana (or nanama) is used with a transitive
suffix (as "mana-ize") and/or causative prefix ("cause-to-be-mana"),and as verbal
noun ("mana-ness", "mana-ization"). The hermeneutic task of inferring meanings
is complicated by the fact that most of the lexicographers of these languages have
been missionaries working in the shadow of Codrington (notably the Anglicans
Ivens and Fox); even so, the semantic pattern is fairly clear.
Because many daughter languages use stative forms mainly adjectivally rather
than verbally, I shall refer generically to "statives" rather than "stative verbs" in
the citations to follow (this pattern of adjectival usage prevails increasingly as we
move eastward from the Solomons).
Taking the languages and dialects of Malaita from northwest to southeast, we
find the following stative uses:
To'abaita: mamana
'be true, real, fulfilled' (I. Frazer,personal communication)
'be successful (of a man)' (Hogbin 1936:259)
Lau: mamana
'be efficacious, (of medicine); grow well, of trees; be good, of news; be
prosperous, lucky, in good health; be true, come true, be fulfilled' (Fox
1974)
'to be true, come true, be fulfilled; to be spiritually or magically powerful'
(Ivens 1934)

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 141

Baegu: mamana
'[be] holy, true, effective' (Ross 1973:235)
Fataleka: mamana
'authentique' (Guidieri 1972:334)
Kwaiot nanama
'be effective, fulfilled, confirmed, realized; "work"
'Are'are: nanama
'be strong, powerful, in metaphysical sense' (Geerts 1970)
Sa'a: nanama
'be powerful' (Ivens 1918)
For verbal forms (intransitive, transitive, or indirectly transitive), we find (as a
partial list):
To'abaita: mamana
'impart spiritual or magical power' (Hogbin 1936:259; this meaning for
"the verbal form of the word" is inferential)
Lau: mamana
'impart spiritual or magical power' (Fox 1974)
mamana fua-
'of ghost, empower (a person)'
Kwaio: nanama
'of ancestor, support, protect, empower'
nanama fa-
'of ancestor, support, protect, empower (a person)'
nanama-nge 'e-ni
'of ancestor, support, protect, empower (a person)'
'Are'are:nanama'ini-
'give power to, cause to be powerful' (Geerts 1970)
'empower, of ghostly action' (Ivens 1930-32)
Sa'a: nanama'ini-
'to put power into' (Ivens 1918)
The last three forms represent transitive suffixation.
As nominalized forms we have:
To'abaita: mamana-a
'blessing, prosperity' (I. Frazer, personal communication)
'ancestrally conferred power' (Hogbin 1936, 1939)
Baegu: mamana-a
'blessing, efficacy, power' (Ross 1973:235)
Lau: mamana-a
'spiritual or magical power' (Fox 1974)
'spiritual or magical power' (Ivens 1934)
mamana-laa
'good health, good luck, success; truth' (Fox 1974)

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
142 JOURNAL OF ANTHROPOLOGICALRESEARCH
Kwaio: nanama-ngaa
'protection, efficacy, good luck, blessing, success, realization; conveying of
protection'
'Are'are: nanama-ha
'spiritual power' (Geerts 1970)
Sa'a: nanama-nga
'power' (Ivens 1918)
Fortunately some Malaita people continue to practice their ancestral religion,
so contemporary ethnographic evidence can shed light on the meanings of mamana,
nanama, and derived forms. I will shortly return to this and other modern ethno-
graphic data on mana in Oceanic religion. First, we can usefully turn to the Roviana
language of the western Solomons, one of a western Solomons subgroup of Oceanic
languages separated (it would seem) by well over four thousand years from the
southeast Solomonic languages.
In Roviana the derivation of abstract verbal nouns from stative verbs has been
generalized by the use of the infix -in- to mark nominalization of both intransitive
verbs and stative verbs (Waterhouse 1949:228). Thus for intransitive verbs, mate,
'die,' becomes m-in-ate, 'death'. For stative verbs, mangini, 'be hot,' becomes m-in-
angini, 'heat'; and malahoro, 'be soft, weak,' becomes m-in-alahoro, 'weakness.'
For mana in Roviana, mana, 'be potent, effectual,' becomes m-in-ana, 'a blessing.'
Waterhouse's definition of m-in-ana is somewhat suspect (a common problem with
missionary renderings, especially where mana has been used in Bible translations).
My own data (W. Paia, personal communication) indicate that in Roviana m-in-ana
is better rendered as "mana-ness" (i.e., efficacy, potency) and was used both in
secular and religious contexts.
In Roviana and in the closely related language of Simbo, mana was often used
as a verbal imperative in prayer and magic. Thus, in his appendix (1949:150),
Waterhouse notes that mana is "used in invocation, as when placing offerings for
tomate [spirits] ." The verbal imperative form ("make it mana!") addressed to spirits
is mana tu! in both Roviana and Simbo, where tu is a sort of exclamation point.
For Roviana we also find a verbal usage employing the transitive suffix: mana-ni-a,
'may it be "mana"!' 'bless it!' (Waterhouse 1931:125).
Rather than try to list all the linguistic evidence that the stative use of mana
('be effective, be potent, be realized, be true'), the verbal invocatory form ('may it
be realized'; 'mana-ize us'), and the nominalized form ("mana-ness") comprise the
canonical ancient, Oceanic pattern, let me simply provide a few examples from
widespread Eastern Oceanic Austronesian languages. For what is surprising about
mana, once we embark on this hermeneutic task with the current linguistic evidence
as text, is the stability of this complex of meanings across more than four thousand
years of the dispersion of Austronesian speakers in the island Pacific.6
I will start with a Cristobal-Malaita language closely related to those of Malaita,
then work toward the northern New Hebrides, Fiji, Micronesia, Rotuma, and parts
of Polynesia.
Longgu (Guadalcanal-Cristobal-Malaita subgroup; Ivens 1937-39)
Stative: nanama, 'be powerful'

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 143
Verb: nanama-ni-, 'empower'
Noun: nanama, 'power'
Gela (Guadalcanal-Gelicsubgroup of Southeast Solomonic; Fox 1955)
Stative: mana, '[be] efficacious, from spiritual power, obtained from charms,
prayers, intercourse with [ancestors or spirits]'
Verb: mana-ngi-, 'to make successful, efficient; to empower, authorize; to rule
over; to put to rights, clear up'
Noun: mana, 'efficacy, success, power, authority'
For Mota (Banks Islands) our lexicographersare Codrington and Palmer (1896);
so it is not surprising to find a Codringtonian rendering. But the same linguistic
pattern still turns up:
Mota (Northern New Hebrideansubgroup of EO)
Stative: mana, 'to have an invisible spiritual force or influence'
Verb: mana, 'to influence, work upon, with an invisible spiritual force'
mana-gi-, 'convey mana to, make to be mana, influence with mana'
Noun: mana, 'an invisible spiritual force or influence'
In another dictionary entry for Mota, we find striking evidence of a semantic pattern
already encountered in the Solomons and widespread farther to the east and north
of Mota, as will be seen: meserere, 'banana leaves made mana with fire and rubbed
on the arms before fighting, for strength and valour.' This stative usage of mana
is clarified by a brief text from the northern New Hebridean language of Maewo.
Codrington (1891:311) quotes a student describing magically poisoned arrows
(toto): "And this Maewo toto is exceedingly mana; if it hits any one by chance,
without being shot at him, he dies..."
Let me move onward to Fiji. Although the Codringtonian usage of mana was
given by Capell (1941), the 1850 dictionary published by Hazelwood, which reflects
usages of a pre-Christianand preanthropologicalera in Fiji, is more interesting.
Fijian
Stative: mana, '[be] effectual; efficient, as a remedy; wonder working' [Capell
notes that "the opposite of mana is drevi, 'useless, inefficacious'"']
Verb: mana, 'a word used when addressinga heathen deity: so be it, let it be so'
Noun: mana, 'a sign, or omen; a wonder, or miracle'
Here three points are worth noting. First, despite all the subsequent attributions
of mana as a noun in the Codringtonian sense to Fiji on the part of ethnographers
such as Sahlins (1963) and Thompson (1940), I see no evidence of such usage in
early texts from Lau, Viti Levu, or other parts of Fiji. Second, the stative usages
of mana as 'be effective, true, realized' and the invocatory usages ['let it be so!']
are well attested by early texts. Hocart, who worked in Fiji early in the century
after doing research on Simbo (western Solomons) was struck by the similarity in
usage of mana (1914:9 8):
Fijians [like Simboese] . . . do not distinguish"true" and "right".Says one informant: "If
it is true (ndina), it is mana;if it is not true, it is not mana." In fact the words are almost
interchangeable,and natives will speak of a sacred stone as mnanaor as ndina ("true")... In
winding up a prayer the words manaand ndina are alwayscoupled; "manae i ndina" ("Let it
[be] mana,let it be true"), is the Fijian"Amen."

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
144 JOURNAL OF ANTHROPOLOGICALRESEARCH
"A thing is mana if it operates;it is not mana, if it does not," says one [Fijian] authority.A
Fijiancure, which is manafor one complaint,may not be manafor another...
A third point is the absence, in early Fiji, of a nominalized form of mana as
"mana-ness." Rather, we find the specialized usage of mana as 'sign, omen, miracle,'
apparently mainly with reference to natural (meteorological?) phenomena. This may
relate to a widespread usage of mana in Western Polynesian languages to refer to
thunder and/or lightning.
Let us return from this brief but illuminating textual examination to trace usages
of mana in other Eastern Oceanic languages. For Rotuman, an Oceanic language
separate in origin from both Polynesian and Nuclear Micronesian, and probably
derivative from pre-Polynesian Fiji, we have (from Churchward 1940), the stative
form mana, '[be] supernatural, miraculous, possessed of or manifesting superhuman
power or extraordinary efficacy.' No usages of mana as a noun are given.
Before turning to Polynesia, where concepts of mana are extensively developed,
a brief excursion into the Nuclear Micronesian languages will further illustrate
the pervasiveness of the pattern.
Mokilese (Harrison and Albert 1977)
Stative: manman, 'spiritually powerful, able to do magic without artifice'
Noun: manman, 'magic, spiritual power'

Ponapean (Rehg and Sohl 1979)

Stative: manaman, 'magical, mysterious, spiritual; official'
Verb: man, 'to take hold, to stick, to be effective'
Noun: manaman, 'magic, mysterious or spiritual power; miracle, authority'
Puluwat (Elbert 1972)
Stative: manaman, 'have divine, supernatural, or miraculous power'
Noun: manaman, 'divine, supernatural, or miraculous power'
Marshallese (Abo et al. 1976)
Stative: monmon, 'having supernatural powers'
Trukese (Goodenough and Sugita 1980)
Stative: mana, 'have divine, magical or supernatural power (a quality of being,
not a form of knowledge); be miraculous, magically powerful'
manaman, 'have divine, magical or supernatural power'
Noun: manaman, 'divine, magical or supernatural power (as distinct from
knowledge)'
Goodenough and Sugita exemplify the stative usage with aa mana ewe safey, 'The
medicine has become powerful'-a usage that would fit perfectly in Fiji or the
Solomons. Goodenough has advised me that this stative verbal usage in Trukese is
the most common, and clearly conceptually primary. He feels that the noun form in
Trukese is best rendered as 'being mana'; it is used in ways that parallel another
nominalized stative, "being right."7
What may be a mana cognate in Yapese, a Micronesian language of uncertain
affinity that will probably turn out to be Oceanic but not Nuclear Micronesian,
conveys a strikingly similar meaning:
Yapese (Jensen 1977)
Stative: mangiing, 'effective, powerful, of medicine'8

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 145
But what about Polynesia, where so many metaphysical ideas about mana have
been reported? We have all learned that in Polynesian languages mana is a noun.
Before considering the Polynesian evidence, an entailment of the linguistic pattern
I have described requires explication. If, as I surmise, mana in POC was canonically
verbal, used as a stative and as an active verb (with god or spirit as subject), and
mana without morphological marking was an abstract verbal noun derived from the
stative ("mana-ness", "being mana"), then the possibility lies open in daughter
languages for this verbal noun to be, in effect, substantivized. Mana can (and in
some languages undoubtedly did) become substantivized as a noun, which can then
displace the verbal usages as conceptually primary, or even replace them altogether.
Polynesian languages are heavily nominalizing, so that the substantivization of mana
would scarcely be surprising.
Although we find evidence of such a process in Polynesian languages, what I
take to be the original pattern still emerges clearly in many parts of Polynesia.
Particularly in Western Polynesian languages (i.e., those most directly derivative
from an earlier Fijian speech community), mana as noun has a limited semantic
domain, seemingly related to Fijian usages.
We can start with Tongan and with Hocart's accounts of Tongan usages in Fiji,
Tonga, and west Uvea:
A Tongan will mention the name of all the leaves and then ask the spirit of the charmto be
graciousand make the charmeffective (mana) (Hocart1929:176).
WallisIsland is a Tongan colony. They are all Roman Catholicsand have almost entirely for-
gotten heathendom, or else do not like to talk about it. But they still use the word mana...
Sosofo gavein his own languagethe following illustration:if you go to the Fatherand ask him
to pray that I shall die, and the Father consents, he holds a Mass that I may die; suddenlyI
die, and the people say, "The Father'sMassis manasince a boy has died." The same informant
explained the expression, "the medicine is mana," as meaning "it is effective" or "useful".
He paraphrasedmanaby lave, which means "to strike," "to hit the mark"(Hocart 1914:99).
Or again, from Tonga:
In Tonga I got a definition in Englishfrom an intelligentTonganwho was educatedin Sydney.
"Manameans if a man puts medicine it takes effect ... ". Another Tongan explained it in
Tongan thus, "If I am angry with my relations and say I wish them to die and they die, it is
mana"(Hocart1914:99).
Tongan dictionary entries give other usages:
Tongan
Stative: mana, 'supernatural,superhuman,miraculous;attended or accompanied
by supernaturalor apparently supernaturalevents' (Churchward1959)
Verb: mana, mana-i, 'to bewitch; to cause something to happen to another'
(Baker 1897)
Noun: mana, 'miracle,supernaturalact or event; supernaturalpower or influence
or attendant circumstances' (Churchward1959)
'thunder, a miracle' (Baker 1897)
For Niue, another daughter language of Proto-Tongic (Tregearand Smith 1907;
McEwen 1970) we have:
Niue
Stative: mana, 'powerful' (as in "I am a powerful god")
Noun: mana, 'power, authority; miracle' (as in "He performed a miracle on a
blind person")

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
146 JOURNALOFANTHROPOLOGICAL RESEARCH
For daughterlanguagesof Proto-NuclearPN we find:
Samoan (Violette 1879)
Stative: mana, 'puissant'
Verb: mana, 'exercer un pouvoir surnaturel'
Noun: mana, 'puissance surnaturelle,prodige, operation divine'
In many languages the noun form is used only for thunder and/or lightning
(Futunan, Sikaiana, Rennell/Bellona). As Burrows (1938:64-65) summed up the
situation for what was then classed as "WesternPolynesia":
the word [mana] [is] . . . used with a connotation of the mysteriousand portentous, the
concept in the West [of Polynesia] seems to be less explicit, less emphasized,and less abstract.
It is applied ratherto specific phenomenasuch as thunder, omens, and spells, than to gener-
alizedpower embodiedin divinitiesor humanbeings.
Firth's detailed account (1940) of mana cognates in Tikopia, another Nuclear PN
language, shows exactly the same pattern of stative usage found in Fiji and the
Solomons. From Firth's texts and analysis we can glean these glosses:
Tikopia
Stative: mana, 'efficacious'
manu, '[be] efficacious, potent, powerful, successful, realized'
Noun: mana, 'thunder'
manu, 'efficacy, power, success'
Firth's analysis of the Tikopian usage anticipates and closely parallelsmy general
one. We can well wonder why its implications have been so little recognized; perhaps
Tikopia, only a tiny speck in the midst of Melanesia,has been seen as deviant.
I shall shortly return to Firth's interpretation, but first I will quickly glance
at some Eastern Polynesian languages. For Marquesiclanguageswe have:
Hawaiian
Stative: mana, 'powerful, strong, influential; able to produce effects' (Andrews
and Parker1922)
'possessed of mana, miraculous, divinely powerful' (Elbert and Pukui 1971)
Verb: mana, 'give mana to'
Noun: mana, 'supernaturalpower; power strength, might' (Andrews and Parker
1922)
Mangarevan(Tregear1899)
Stative: mana, 'powerful, mighty'
Noun: mana, 'power'
Marquesan(Tregear1891)
Stative: mana, 'strong, of gods'
Noun: mana, 'power, dominion, divinity'
For Central East Polynesian languageswe find, as dictionary entries:
Tahitian (Davies 1851)
Stative: mana, 'powerful, mighty, affluent'
Verb: mana, 'be in power, possess influence'
Noun: mana, 'power, might, influence'
Raratongan(Savage 1962)
Stative: mana, 'effectual, binding, authoritative, having influence or power'

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 147
manamana, 'powerful, having great power, possessed of great magical
power'
Verb: mana, 'to take effect, to be effectual'
Noun: mana, 'power, might, authority, influence'
Tuamotuan (Stimson and Marshall1964)
Stative: mana, 'be possessed of mana; having power, authority, or prestige'
Noun: mana, 'an extraphysical power or potentiality, either naturally or magi-
cally derived'
Maori (Williams1957)
Stative: mana, 'effectual, binding, authoritative; having influence or power;
vested with effective authority; be effectual; take effect, be avenged'
Verb: mana, 'support' (Gudgeon 1885)
Noun: mana, 'authority, control; influence, prestige, power; psychic force'
Whereasethnographersof the Maori, portraying a mystical world view that seems
as much a product of European as Maori imagination, emphasize the noun usage,
many texts and illustrations reveal stative and verbal usages. Thus from Gudgeon
(1885:217):
Mana has many and various meanings;for instance it meansfulfil ... ka manataku kupu i au
(I will fulfil my word); and it means potent, as he karakiamana(a potent charm);and it also
means effective, as he kupu mana tana kupu (his word is effective); it also means granted,
as e kore to tono e whakamana(your request will not be granted);it also meanssupport,as
mawaie manaai tau kupu (who will supportyou that your word will be effective).
We can well note, in these Eastern Polynesian usages of mana, a greater empha-
sis on the sanctity and authority of (increasinglygodlike) chiefs. I will return to this
shift. What I find most striking, however, are the continuities-the uniformities in
usage extending from the Solomons through Fiji and northward to Truk, southward
to New Zealand, and out to the Marquesasand Hawaii.
I have found only one probable mana cognate among Oceanic languageswest of
the Solomons. But here again the continuity in usage is striking. Seligman, seeking
his own peculiar version of the Codringtonian concept of mana in the islands near
Milne Bay, New Guinea, had observed (1910:576) that
Neither at Wagawaga,Tubetube nor elsewhere in the district does there seem to be any
development of that system of personal influence (mana) taboo whereby the thing made
taboo receives,as it were, a dynamicchargefrom contact with an individual.
Yet the one apparent cognate I have found comes from Tubetube. MarthaMacintyre
(personal communication) cites what seems to be a metathesized and reduplicated
form of mana: naManaMa,'be efficacious, "work", be good, be true, have positive
qualities, fulfil potential (that is, of an animate or inanimate entity, to manifest
qualities appropriate to one's nature).'9 A Tubetube folk healer told Macintyre:
naManaMane nima-gu, 'my hands are mana.' Almost at the same time, the ethnog-
rapher Barbara Herr, citing her field data on folk healing and ethnopsychology in
the Lau Islands of Fiji, wrote to me of a folk healer who told her: sa mana na
liga-qu, 'my hands are mana.' Fiji and Tubetube are separated by more than four
thousand years of time and some two thousand miles of ocean. Yet these two
folk healers, practicing their art on islands Christianized for decades, are using the
same ancient concept in exactly the same way.

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
148 JOURNAL OF ANTHROPOLOGICALRESEARCH
AN ANTHROPOLOGICAL
INTERPRETATION

I argue elsewhere (Keesing 1979, 1981) that many modem ethnographic in-
terpretations of mana cognates by social anthropologists have been distorted by
Codringtonian orthodoxies; we all learned what mana meant when we were graduate
students. In fact I began this hermeneutic task of reinterpretationwhen I discovered
my own pervasiveerrorsof translation.
Nonetheless, field data and texts allow a cautious reinterpretationof how mana
was conceptualized by POC speakers and how these conceptualizations changed
among some of their scattered descendants, despite the striking continuities we
have seen. Mana is a concept that addresses two circumstances of life among early
settlers of the Pacific: first, the essential unpredictability of the outcomes of human
effort-in war, fishing, gardening, feastgiving, curing, and other activities; and
second, inequalities among humans-in their attainments and success, in their
rank, and in their access to the gods and spirits (these were, of course, intertwined,
interdependent but separable). POC-speakingpeoples very probably had hereditary
chiefs, but rank had always to be validated by political and military success, by
responsible use of authority.
In every realm of worldly effort, these early Oceanic speakerssought the support
of ancestral and other spirits. In prayer they sought protection and blessing; in
magic the objects and spells humans used required validation and potentiation by
the spirits. Having done the part humans must do to win wars, catch fish, grow
taro, give successful feasts, cure the sick, and conduct divinations, they waited
to see whether the gods and spirits had done what they must do-the invisible
complement of what humans do. The concept of mana is an explanation of the
difference between a stone used magically that "works" and an ordinary stone; of
the difference between a medicinal potion that cures the patient and a potion that
does not; a divination that turns out to be true and one that does not; of the differ-
ence between a raid that ends in victory and a raid that ends in disaster;between a
fishermanwho brings home a successful catch and one who returns empty-handed.
The stone or potion that "works" magically looks the same as an ordinary stone
or potion. The difference is invisible, a potentiation by the spirits. Not surprisingly,
conventional metaphors of heat are often used to describe this invisible potentia-
tion.10 The stone or magical potion or poisoned arrow is mana: it works by virtue
of its potentiation by spirit beings, complementary to human skill and knowledge.
Codrington saw this idea of potentiation as evidence for a notion of an in-
visible medium of power, some invisible ingredient the stone, the potion, and the
arrow have. But I believe that, like the "soul substance" Indonesian headhunters were
inferred to be acquiring in taking heads (see Needham 1977), mana as invisible me-
dium of power was an invention of Europeans, drawing on their own folk metaphors
of power and the theories of nineteenth-century physics. Europeans invented a
metaphysic which seemed to be implied by the way Oceanic peoples talked. This
point was perceived by Firth (1940:497-98) years earlier:
To the Tikopia, manu I am sure has not the connotation of an isolatable principle,a power,
or any other metaphysical abstraction-though it may be conceived of as a specific quality.
The interpretationin termsof such abstractioncan only be the work of the anthropologist.
I see no evidence that early Oceanic speakers had any developed theology about
what the spirits do that complements human activity and allows its realization. In

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 149
some cases, acts, objects, and words were perceived as potentiated; in other realms-
living free from illness, growing successfully (of humans, taro plants, pigs)-mana-ing
by the spirits seemingly entailed no potentiation, but only a protection that allowed
the natural realization of life processes. As one of my Kwaio informants put it:
What we see is that something happens. Something recovers.The taro growswell. Unless the
adalo [ancestralspirit] nanama'sthe pig will die, the taro will grow badly. We don't see how
it happens.What we see is the man who recovers.We see that he is alive. Or we see that the
taro or pigs are good. Or that a man has won his court case. Then we know that the adalo
has nanama-ed.
Or another:
You see that the taro corms are well formed, you see that the bodies of the pigs are sound,
and you say, "Oh, the thingsI've done are nanama."
Or another:
If you [RMK] give white man'smedicine to me, and I am cured,then the medicineis nanama.
If I don't get better, then it isn't nanama.
The latter passage suggests that for Kwaio, at least, the "element" of success, the
unknown and unseen complement to human effort, need not be "metaphysical"
or conveyed by the spirits; similar usages have turned up all over modern Oceania.
Coastal, Christian Kwaio pull the rope on an outboard engine: if it doesn't start,
it isn't nanama.
Such "secular" usages of mana seem to have been widespread in pre-Christian
Oceanic societies; they posed something of a puzzle to early observers. But they
pose no contradiction to the pagan Kwaio, still sacrificing pigs to their ancestors. If
we suppose that mana is some invisible medium of power conveyed by gods or
spirits, then how the makers of penicillin or the doctors who give the injections put
this into the medicine poses a metaphysical contradiction. But there is no such
ingredient, only a quality of efficacy manifest in visible results.
The same retrospective pragmatism, of trying magic or medicine to see if it
works, of attributing mana-ness or its absence on the basis of perceived outcomes,
turns up clearly in Firth's account of Tikopia (1940:497):
[manu] is correlatedalways with concrete situations, falling of rain, growth of food, advent
of calms,relief of sickness.In fact its very existence is inferredby such concrete results.Again
and again I hammeredaway at my informantstrying to find what was the meaningof manu
itself apartfrom the evidenceof it in crops, fish, and the like. But all my inquiries... came to
nothing. Always it was insisted that the crops and the fish were manu.
The same thing is seen in Hocart's account of Fiji (1929:186):
Tui Tumbou gives the following explanation of the word: A mana thing is a thing which ...
is mana; as for a spirit thing, if we speak to it, then the words of the spirits... mana;but if
it does not come to pass they do not mana. If it is true (ndina)it is mana;if not true then it
is not mana ... A Fijianmedicine does manaif it works;it does not manaif it does not work.
The European error lay in inferring some substance-like medium, mana, the
invisible element the spirits and gods gave or withheld, the universal medium of
power and success. Pacific Islanders knew that what made a magical stone mana was
not the same as what made a war canoe or a fishing expedition mana: different
ancestors or gods conveyed different powers to different people in different (though
unknown) ways. But with few exceptions, Pacific Islanders have been unsuccessful
in explaining that to theologically minded Europeans, caught up in their own
conventional metaphors of power, notions of a universally explained universe, and

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
150 JOURNAL OF ANTHROPOLOGICALRESEARCH
physical models of electricity and hydraulics. We have not understood that mana-ness
represented a common quality of efficacy or success, retrospectively interpreted,
not a universalmedium of it.
The nominalized usages reflect no substance or even a substantive in most
Oceanic languages, and especially not in those from Samoa and Tonga westward;
rather, I think, they are more like "success" or "luck" than like "power" in a
quasi-physical sense. Kwaio pray for nanama-ngaa:"Nanama-izeus! Nanama for us!
Give us nanama-ngaa."A pair of men lost at sea in a canoe, then saved, in a Kwaio
tale say, when they are rescued "nanama-ngaangai a-da'a," 'good fortune is with
us!' But to imagine that this represents a metaphysical transaction in which the
ancestors convey mana as a "thing" to their descendants is to misunderstandtotally
a pragmatic world view in which the living seek to channel the outcomes of their
efforts, not to explain them metaphysically (Keesing 1982). Needham (1977) notes
that if we insist on seeking an underlying metaphysic in nonwestem religions, we
may be providing answers to questions that are simply not posed within another
people's world view (just as we ourselves invoke a concept of "accident" for those
co-occurrences of events for which we seek no explanation). Hogbin (1936:245)
notes of his efforts to uncover a philosophy of mana on Guadalcanalthat "nobody
knows how nanama works, and I gathered that the thought had never occurred to
anyone before I made inquiries."
Manamay metaphoricallybe substantivized:a magical stone may "have mana. 1
But the interpretation by missionaries (and anthropologists) that the stone thereby
"contains" some invisible medium of power seems to me fundamentally erroneous.
C.E. Fox, an Anglican missionary working, in the shadow of Codrington, on San
Cristobal, in the southeast Solomons, wrote (1924:251, 252) of the Arosi concept
of mena:
Mena . . .seems to be conceived of as an invisible spiritualsubstancein which objects may
be immersed...
A great warrioris seen to have mena and all his possessionsare soaked in it, so his club is
treasuredand handeddown ... ;certain placesare impregnatedwith mena.
This seemsanalyticallyequivalentto inferringthat the Pope'scrucifixis soakedin
sanctityand a rabbit'sfoot is soakedin good luck.
Here is where our conventionalmetaphorsabout "power" come into play.
Whereaswe do not metaphoricallysubstantivize"success"or "sanctity,"we per-
vasivelyrender"power"as if it were a quantifiableentity, a "thing"people have
more or less of: someone who has "it" is powerful,"full" of "it." Yet "having
power," cut loose from this metaphorical substantivization, is a relationship, always
contextual and two-sided. Because of the way we metaphorically substantivize
"power," the term was adopted to label quantifiable electrical energy as a medium
whose flow could be channeled through cables and directed to human ends. This
physicalist conception of electrical energy as power, like Fox's physicalist metaphors
of immersion, has affected characterizationsof mana. Handy's interpretation (1927:
28) of Polynesian religion is a striking example:
The Polynesianideas concerningthe behaviorof and laws governingmanaare analogousin so
many ways to the known nature of electricity that the simplest and clearest method of
illustratingthe native concept of the all pervasivepsychic force is by describingit in terms
of this all pervasivephysicalenergy.

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 151
He goes on to describe "psychic magnetic influence," flows between "transmitter"
and "reservoir," "insulation," and "positive" and "negative" poles (1927:28). Of
Handy's metaphysics of psychic dynamism, Douglas Oliver (1974:55) cogently notes
that
no recorded texts or accounts of behaviorthat I know of provideevidencethat the [Society
Islanders] entertained such a generalized,all-embracing,energetic animatisticview of their
universe . . . I believe that [Handy's] view is the construct of the philosophicallyminded
ethnologistand not, even implicitly, of the [islanders]themselves.12
I suggest that a minimal first step toward a hermeneutic reinterpretationof mana
in Oceanic religion would be to change every gloss of mana as "power" to "potency,"
and every gloss of "powerful" to "potent." "Potent" as stative and "potency" as
derived abstract noun (and the parallel series needed to capture other senses of
mana: "effective" and "efficacy"; "true" and "truth"; "realized" and "realization";
"sacred" and "sanctity"; "confirmed" and "confirmation"; etc.) begin to capture
the linguistic, semantic, and ethnographic facts at hand.
Yet in some parts of eastern Polynesia, mana clearly had become incorporated
into systems of social hierarchy in which the high chiefs approached the status of
gods, in which the sanctity of chiefs was expressed and ritualized in concrete images
of mana and tapu attaching to people and things. That takes us back to the second
aspect of mana as a key concept in early Oceania: its reference to human inequality,
to differential rank and differential success.
It seems probablethat speakers of POChad a stratified social system with heredi-
tary chiefs (see, e.g., Pawley 1981, 1982). But those entitled to lead presumably had
to demonstrate, by success in war, skill in leadership and resource management, and
proper conduct, that they had the support of the superhumaninvisible beings, gods
and ancestors, on whom life depended.
Such success was continuous visible evidence that the leader himself was mana or
"had mana." We speak metaphorically of success "following" someone. "Having
mana" I take to have originally been such a conventional metaphor. The hereditary
chief was mana, or had mana, as long as he fulfilled the expectations and responsibili-
ties of his position. The feared warrior,the skilled navigator, the successful gardener
or hunter or fisherman "had mana" in the sense that good fortune "went with him"
(with the Oceanic implication that good fortune is not a matter of chance but of
support from unseen beings).13 The individual who in day-to-day life did not stand
out from his or her fellows would see in particular outcomes-a successful catch,
a propitious omen, a lucky escape-evidence of being mana or having mana: recall
the Kwaio sailors who, when saved, proclaimed that "nanama-ngaais with us!"
But I do not see in the distributional evidence grounds for inferring that early
Oceanic speakers had a substantivized concept of mana of the kind widely reported
for Polynesia, in which the head, the person, the words of a chief contained or
emanated sacred power. Such usages of mana are relatively rare in western Polynesia,
where mana as noun more often refers to miracles, to thunder and lightning, and
other evidence of superhuman beings than to substantivized "power." In pre-
Christian Fiji, mana was rarely used with reference to the potency or authority of
chiefs; when it was, it seems to have been used in stative form (Hocart 1929:38):
"Leading nobles are mana, that is they cause supernatural effects, only to their

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
152 JOURNAL OF ANTHROPOLOGICALRESEARCH
juniors." Some Micronesian peoples with hereditary chiefs (e.g., Marshallese)use
mana cognates only as statives.
However, we cannot dismiss the evidence from Polynesia, especially eastern
Polynesia. Firth's account of the manu of Tikopian chiefs, the most reliable textual
evidence on pre-EuropeanPolynesian religion, indicates that at the very least, manu
had become a kind of concretized metaphor (Firth 1940:502, 504):
"The manu is there in you, there in your hand which touches and your outer lips." [Pa
Rangifuri]14
"The Gods take and place [manu] on the head of him who has asked for the mana to be
given to him."
In Tikopia, at least, stative usages regarding chiefly potency seem at least as
common as nominalizations (Firth 1940:491, 494):
"Whenwe see that the rain has come we say: 'the manu chief.' If we say also 'the chief is
manu' it is correct. If he asks for the breadfruitto come, for it to fruit, and then it fruits, we
say 'he has been manu'; the asking of the chief has been made manu ... He is termeda manu
chief, a manuman."
"In this land manu is there in the lips of the chief. In his speech whateverhe may ask for, if
a chief is manu then when he asks for fish, they will come; when he speaksrequestinga calm
it falls. That is a manu chief.. . Whenhe asks it of the spirits, if the spirits wish to give it
hither, they give it, and therefore I say that the chief is manu. A chief who is manu-the
spiritsjust continuallyrejoicein their desiretowardsthe chief" [Pa Fenuatara]
That in Tahiti, the Marquesas, New Zealand, and Hawaii, Polynesians had
further substantivized mana as an invisible medium of power connecting godlike
chiefs with the gods seems, translation errors and European-imposed mysticism
notwithstanding, beyond question. Moreover, there are some parts of Melanesia
where mana has apparently been substantivized, where a metaphysic that lies latent
in conventional metaphors of "having mana" was apparently created. The next
stage in a reinterpretation of mana is to connect the substantivizations of mana in
metaphysical terms with the worldly circumstances that have given rise to such
cosmologies.
The evidence points toward the processes whereby hereditary chiefs became
transmuted into an aristocratic class in Polynesia (and some parts of Micronesia)
as a key to the emergence of a metaphysic of mana. It seems not simply to have
been the increasingly godlike status of the highest chiefs that progressively led to
the sanctification of their persons (to which mana and tapu in their most fully
developed forms give expression). I believe that the creation of developed theologies
among Oceanic speakers reflects the emergence of a class of theologians, out of the
older Polynesian custodians of lore, ritual, magic, and genealogical knowledge.
Theologians as part of the sacred chief's entourage, celebrating and rationalizingthe
chief's sanctity, seem to have elaborated the cosmological implications of mana as
metaphoric "power." Whether in the process they created the full-blown mystical
theory of "psychic dynamism" imputed to them, we may well doubt, as Oliver
(1974) does.15
What, then, about the apparently substantivized and cosmologically elaborated
conceptions of mana reported from a few parts of Melanesia?It is hard to reconcile
what we know of, say, San Cristobalin the Solomons with an image of a small band
of professional theologians validating the godlike status of their sacred chiefs, when

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 153
in fact there were no godlike chiefs. Perhaps Fox, his fanciful accounts of "immersion
in mana" notwithstanding, gives us the answer (1924:252):
The ... Arosi ... priest of the Serpent ... will say that there is a great stock of mena, whose
source is the serpent figona, and in the last instance in a personal centre of mena, the spirit
Agunaa; from this central source mena is imparted to adaro [ghosts] and to material objects.
The ordinary Arosi native will say the source of mena is in the ghosts.
Cosmologies substantivizing mana have rarely emerged in Melanesia, I think, as
validations of chiefly rank; but more probably in contexts of religious cultism,
especially secret cultism. Where there is a cult, an inner sanctum, there must be
Inner Mysteries; where knowledge is a source of power (Keesing 1982) and access to
it is controlled and restricted, then cult secrets and esoteric knowledge must be
elaborated.
Tracing out the development of mana as a concept in time and space, anchoring
it in social systems rather than disembodied philosophies, is a continuing task. The
task demands not only painstaking hermeneutic skill, but comparative sociological
vision. The coparticipation in this search of scholars who have the needed command
of Oceanic languages and ethnographic evidence is urgent.

NOTES

1. This paper presents part of the argument 2. In most daughter languages, some stative
and evidence set out in a book whose comple- verbs function as adjectives as well, modifying
tion has long been delayed. My rethinking of nouns directly. As stative verbs they always
mana began early in 1979, when I discovered follow a preverbal pronoun as copula. Thus in
with shock (on going back to original tapes and Kwaio (Solomon Islands), ruma lo'oo e ba'ita,
texts while writing Kwaio Religion, Keesing 'this house is big' ("house here (it) be big") and
1982) my own mistranslations and misrender- ruma ba'ita, 'a big house' ("house big").
ings of Kwaio nanama. I have given parts of the 3. For simplicity of explication, I ignore
argument since in seminars at the University of the probability that a final consonant will have
California, Santa Cruz, the University of to be reconstructed for mana (*manang?) in
Chicago, the University of California, San POC, and the possibility that as subgrouping
Diego, the Australian National University, the proceeds the pattern I describe will have to be
University of Sydney, the University of Ade- attributed to an interstage proto-language
laide, the University of the South Pacific derived from POC.
Centre, Honiara, and the University of Hawaii; 4. Most daughter languages of POC use a
and have given papers touching on mana at causative prefix to form causative verbs from
annual meetings of the Southwestern Anthro- intransitive and stative verbs. With statives they
pological Association (Berkeley 1979) and the form meanings such as 'heat (something)' (from
American Anthropological Association (Los the stative 'be hot') and 'strengthen something'
Angeles 1981). For helpful comments and/or (from the stative 'be strong'). Causative con-
data I am grateful to a great many scholars, structions using mana are widespread in Mela-
among them Michael Allen, Bob Blust, Ann nesia and Polynesia (e.g., Lau faa-mamana
Chowning, Jim Clifford, Roy d'Andrade, 'establish, make true, make successful, effi-
William Davenport, Raymond Firth, Ray cient' and Maori whaka-mana 'give effect to,
Fogelson, Ian Frazer, Ward Goodenough, give prestige to, make effective, rectify'). For
George Grace, Barbara Herr, Alan Howard, the sake of brevity, I will not deal with causa-
MargaretJolly, Bruce Kapferer, Peter Lawrence, tives formed from mana as stative verb.
Dennis Lulei, Martha Macintyre, Nancy Munn, 5. The position of Southeast Solomonic
Rodney Needham, Douglas Oliver, Warren Paia, languages, including those of Malaita, within
T.N. Pandey, Andrew Pawley, Michelle Rosaldo, Oceanic is not yet clear. Pawley (1972) in-
Francis Saemala, Harold Scheffler, Milton cluded them within his putative EO subgroup, a
Sibisopere, Esau Tuza, Darrel Tryon, Terry hypothesis that in the long run will probably be
Turner, Jim Urry, and David Walsh. sustained despite his later doubts (Pawley 1977,

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
154 JOURNAL OF ANTHROPOLOGICALRESEARCH
Grace 1976) and the lack of firm supporting mana, is said to be mana, with the use of the
evidence. verb."
6. The definitions given for the noun forms 12. Oliver (personal communication, 20
by lexicographers who assumed these were the May 1983) tells me that in his exhaustive search
primary senses of mana and who were in many of early Tahitian texts, he found only a handful
cases influenced by anthropological ortho- of occurrences of mana; he infers that the term
doxies and Biblical translations, are often was rarely used, and (contra Handy) had a very
suspect; but the pattern is nonetheless clear. minor place in Tahitian religious beliefs and
7. Personal communication, 20 May 1983. political ideology.
8. Discussions with Yapese specialists in 13. Ward Goodenough (personal communi-
Hawaii (18 May 1983) have convinced me that cation, December 1981) observes that "being
this almost certainly is a mana cognate, showing mana" in Truk was like "having a green thumb"
the expected final consonant reconstructed by in Philadelphia.
Grace for POC (*manang) and a vowel shift 14. Even Firth slips into errors of substan-
attested elsewhere in Yapese (POC *mata 'eye' tivization in translation, however: he translates
> Yapese mitt). (1940:504) the causative verbal usage ke
9. The M is labialized as mw in medial fakamana i toku rima as "to give mana from
position; it corresponds to m in neighboring my hand" ("cause to be mana through my
languages. hand" would seem more apt). Such translation
10. On conventional metaphors see Lakoff errors, as I have shown elsewhere (Keesing
and Johnson (1980) and Keesing (1983). 1979, 1981), are rampant in ethnographic
11. But note that even Codrington's 1891 accounts of Melanesia (including my own).
description sets out the stative usage, albeit 15. David Walsh (personal communication),
mainly in a footnote (1891:119): "an abundant on the basis of his study of Maori texts, has
crop on the tree or in the garden shows that he expressed to me his conviction that much of
is right, the stone is mana"-to which he the mysticism attributed to the Maori is the
appends the footnote that "an object in which product of European, not Polynesian, imagina-
mana resides, and a spirit which naturally has tion.

REFERENCES CITED

Abo, T., B.W. Bender, A. Capell, and T. Practices in Melanesia. Journal of the Royal
Debrum, 1976, Marshallese-English Dictionary. Anthropological Institute 10:261-315.
PALI Language Texts. Honolulu: University of Codrington, R., 1891, The Melanesians:
Hawaii Press. Studies in their Anthropology and Folk-lore.
Andrews, L., and H.H. Parker, 1922, A Oxford: Clarendon Press.
Dictionary of the Hawaiian Language. Honolu- Codrington, R., and J. Palmer, 1896, A
lu: Bishop Museum. Dictionary of the Language of Mota. London:
Baker, S.W., 1897, An English and Tongan Society for Promoting Christian Knowledge.
Vocabulary and a Tongan and English Vocabu- Davies, J., 1851, A Grammar of the Tahi-
lary, with . . . Phrases and Tongan Grammar. tian Dialect of the Polynesian Langauge. Tahiti:
Auckland, N.Z.: Wilson & Horton. Mission Press.
Burrows, E.G., 1938, Western Polynesia: A Elbert, S.H., 1972, Puluwat Dictionary.
Study in Cultural Differentiation. Etnologiska Pacific Linguistics, C-24. Canberra, Australian
Studier, no. 7. Goteborg: Goteborgs Museum. National University.
Capell, A., 1941, A New Fijian Dictionary Elbert, S.H., and M.K. Pukui, 1971,
compiled... for the Government of Fiji. Syd- Hawaiian Dictionary. Honolulu: University of
ney: Australasian Medical Publishing Company. Hawaii Press.
Churchward, C.M., 1940, Rotuman Gram- Firth, R., 1940, The Analysis of Mana: An
mar and Dictionary. Sydney: Australasian Empirical Approach. Journal of the Polynesian
Medical Publishing Company. Society 49:483-510.
Churchward, C.M., 1959, Tongan Diction- Fox, C.E., 1924, The Threshold of the
ary. London: Oxford University Press. Pacific: An Account of the Social Organization,
Codrington, R., 1881, Religious Beliefs and Magic and Religion of the People of San Cristo-

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
 RETHINKING MANA 155
val in the Solomon Islands. London: Kegan Journal of the Polynesian Society, Supplement,
Paul. Memoir No. 11. Wellington, New Zealand.
Fox, C.E., 1955, A Dictionary of the Ivens, W.G., 1937-39, A Vocabulary of the
Nggela Language (Florida, British Solomon Language of Longgu, Guadalcanal, Solomon Is-
Islands). Auckland, N.Z.: Unity Press. lands. Bulletin of the School of Oriental and
Fox, C.E., 1974, Lau Dictionary. Pacific African Studies 9:165-93.
Linguistics, C-25. Canberra: Australian National Jensen, J.T., 1977, Yapese-English Diction-
University. ary. Honolulu: University of Hawaii Press.
Geerts, P., 1970, 'Are'are Dictionary. Keesing, R.M., 1979, Rethinking Mana:
Pacific Linguistics, C-14. Canberra: Australian The Kwaio Case. Paper presented at annual
National University. meetings of Southwestern Anthropological
Goodenough, W.H., and H. Sugita, 1980, Association, Berkeley, California.
Trukese-English Dictionary. Philadelphia: Amer- Keesing, R.M., 1981, Mana in Nuclear
ican Philosophical Society. Oceania. Unpub. ms. in author's possession.
Grace, G., 1976, Review Article on Studies Keesing, R.M., 1982, Kwaio Religion: The
in Oceanic Culture History, vol. 3 (ed. by R.C. Living and the Dead in a Solomon Island
Green and M. Kelly). (Pacific Anthropological Society. New York: Columbia University Press.
Records, no. 13. Honolulu: Bishop Museum.) Keesing, R.M., 1983, Conventional Meta-
Journal of the Polynesian Society 85(1):103-12. phors and Anthropological Metaphysics: The
Gudgeon, T.W., 1885, History and Doings Problematic of Cultural Translation. Paper
of the Maoris from 1820 to 1840. Auckland, presented at Conference on Folk Models,
N. Z.: Brett. Institute for Advanced Study, Princeton, N.J.,
Guidieri, 'R., 1972, La route des morts: 14 May 1983.
introduction la vie ceremonielle Fataleka. Lakoff, G., and M. Johnson, 1980, Meta-
Journal de la Societ6 des Oceanistes 18:323-36. phors We Live By. Chicago: University of
Handy, E.S.C., 1927, Polynesian Religion. Chicago Press.
Bulletin no. 34. Honolulu: Bishop Museum. McEwen, J.M., 1970, Niue Dictionary.
Harrison, S.P., and S. Albert, 1977, Mo- Wellington, N.Z.: Department of Maori and
kilese-English Dictionary. PALI Language Texts: Island Affairs.
Micronesia. Honolulu: University of Hawaii Needham, R., 1977, Skulls and Causality.
Press. Man, n.s., 11:71-78.
Hazelwood, D., 1850, A Feejeean and Eng- Oliver, D.L., 1974, Ancient Tahitian
lish Dictionary with Examples of Common and Society. Honolulu: University of Hawaii Press.
Peculiar Modes of Expression . . . Native Cus- Pawley, A., 1972, On the Internal Rela-
toms, Proverbs, Native Names (etc.). Vewa: tionships of Eastern Oceanic Languages. Pp.
Weslayan Mission Press. 1-142 in Studies in Oceanic Culture History,
Hocart, A.M., 1914, Mana. Man 14:97-101. vol. 3 (ed. by R.C. Green and M. Kelly). Pacific
Hocart, A.M., 1929, Lau Islands, Fiji. Anthropological Records, no. 13. Honolulu:
Bulletin No. 62. Honolulu: Bishop Museum. Bishop Museum.
Hogbin, I., 1936, Mana. Oceania 6:241-74. Pawley, A., 1977, On Redefining 'Eastern
Hogbin, I., 1939, Experiments in Civiliza- Oceanic.' Unpub. ms., University of Auckland,
tion. London: Routledge. N.Z.
Hubert, H., and M. Mauss, 1902-3, Es- Pawley, A., 1981, Melanesian Diversity and
quisse d'une theorie generale de la magie. Polynesian Homogeneity: A Unified Explana-
L'Annee Sociologique 7:1-146. tion for Language. Pp. 269-309 in Studies in
Ivens, W.G., 1918, Grammar and Vocabu- Pacific Languages and Cultures in Honour of
lary of the Language of Sa'a and Ulawa, Solo- Bruce Biggs (ed. by J. Hollyman and A. Pawley).
mon Islands. Washington, D.C.: Carnegie Auckland: Linguistic Society of New Zealand.
Institution. Pawley, A., 1982, Rubbish-man Commoner,
Ivens, W.G., 1930-32, A Vocabulary of the Big-man Chief? Linguistic Evidence for Heredi-
Language of Marau Sound, Guadalcanal, tary Chieftainship in Proto-Oceanic Society. In
Solomon Islands. Bulletin of the School of Oceanic Studies (ed. by J. Siikala). Helsinki:
Oriental and African Studies 6:963-1002. Finnish Anthropological Society, Transactions,
Ivens, W.G., 1934, A Vocabulary of the no. 11.
Lau Language, Big Mala, Solomon Islands. Rehg, K.L., and D.G. Sohl, 1979, Pona-

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions
156 JOURNAL OF ANTHROPOLOGICAL RESEARCH

pean-English Dictionary. PALI Language Texts: Comparative Dictionary. Wellington, N.Z.:

Micronesia. Honolulu: University of Hawaii Lyon & Blair.
Press. Tregear, E., 1899, A Dictionary of Man-
Ross, H., 1973, Baegu: Social and Ecologi- gareva (or Gambier Islands). Wellington, N.Z.:
cal Organization in Malaita, Solomon Islands. Government Printer.
Urbana, Ill.: University of Illinois Press. Tregear, E., and S.P. Smith, 1907, A
Sahlins, M., 1963, Poor Man, Rich Man, Vocabulary and Grammar of the Niue Dialect
Big Man, Chief: Political Types in Melanesia of the Polynesian Language. Wellington, N.Z.:
and Polynesia. Comparative Studies in Society Government Printer.
and History 5:285-300. Violette, L., 1879, Dictionnaire samoa-
Savage, S., 1962, A Dictionary of the frangais-anglais et frangais-samoa-anglais precede
Maori Language of Rarotonga. Wellington, d'une grammaire de la langue samoa. Paris:
N.Z.: Department of Island Territories. Maisonneuve.
Seligman, C.G., 1910, The Melanesians of Waterhouse, J.H.L., 1931, The Kazukuru
British New Guinea. Cambridge, Eng.: Cam- Language of New Georgia. Man 31:123-26.
bridge University Press. Waterhouse, J.H.L., 1949, A Roviana and
Stimson, J.F., and D.S. Marshall, 1964, A English Dictionary, with English-Roviana Index.
Dictionary of Some Tuamotuan Dialects of the Revised by L.M. Jones. (Earlier editions 1923,
Polynesian Language. The Hague: Martinus 1928.) Sydney: Epworth Press.
Nijhoff. Williams, H.W., 1957, A Dictionary of the
Thompson, L., 1940, Southern Lau, Fiji: Maori Language. 6th ed. Wellington, N.Z.:
An Ethnography. Honolulu: Bishop Museum. Government Printer.
Tregear, E., 1891, The Maori-Polynesian

This content downloaded from 138.73.1.36 on Thu, 19 Nov 2015 10:07:06 UTC
All use subject to JSTOR Terms and Conditions