Volume 90 1981 > Volume 90, No. 1 > Growth and development of folk zoological life-forms in Polynesian languages, by Cecil H. Brown, p 83-110
                                                                                             Previous | Next   

- 83

Folk zoological life-form terms designate the most inclusive, comprehensive classes of animals regularly found in folk zoological taxonomies. An earlier survey of 112 globally distributed languages showed that five life-forms, “fish”, “bird”, “snake”, “wug” (e.g., American English bug), and “mammal” are added to languages in a highly regular order (Brown 1979a). In addition, subsequent reconstruction of zoological life-form systems of proto-stages in the development of the Mayan language family (Mesoamerican), has been shown to conform with this order (Brown and Witkowski 1980).

Figure 1 presents the lexical encoding sequence for zoological life-forms as revised by Brown (1981a,b) from an earlier formulation (Brown 1979a). It shows that languages lexically encode “bird”, “fish”, and “snake” before “wug” and “mammal”.

Folk zoological life-form encoding sequence (Brown 1979a)

The present study like the Mayan (Brown and Witkowski 1981) treatment undertakes to reconstruct zoological life-form systems of proto-stages of the Polynesian language group. Again, proto-language reconstruction provides evidence consistent with the zoological life-form encoding sequence. However, unlike Mayan languages, which have con- - 84 sistently added animal life-forms over time following a worldwide tendency (Brown and Witkowski 1981), Polynesian languages in some instances have actually lost life-form classes. This is due to the peculiar faunal ecologies encountered by Polynesians as they have spread through their vast Pacific habitat.

In addition to the zoological life-form encoding sequence, an encoding sequence for folk botanical life-form classes has been described (Brown 1977). 1 Both encoding sequences are language change sequences. Languages necessarily add or lose biological life-form terms in the orders proposed. Acquisition rather than loss of life-forms, however, has been the overwhelming tendency through time. This relates to the fact that the two encoding sequences are both strongly related positively to societal complexity (Brown 1977:328-32, 1979a:804-5). Languages having few animal or plant life-form terms tend to be spoken by peoples living in small-scale societies and those having many by peoples of large-scale urban societies. Since societal complexity has generally increased during the course of human history, it follows that biological life-form inventories in most cases have grown rather than shrunk in size. Thus, the animal and plant life-form encoding sequences are primarily additive in nature.

The correlation between number of biological life-form terms and societal complexity indicates the greater usefulness and salience of life-forms in large-scale societies. This greater usefulness relates to the increasing separation of humans from direct reliance and dependence on the natural world as societies become larger and more complex. The typical individual in a small-scale society can commonly name and identify hundreds of separate plant species (Berlin et al. 1974, Conklin 1954, Hays 1976), while typical nonspecialist members of modern urban society might do well to name and identify even 100 (Dougherty 1978). As detailed terms disappear from folk systems of plant and animal classification, general terms, such as life-form names, become increasingly salient and tend to grow in number. Addition of biological life-form classes to languages then indexes an overall decrease of interest in and interaction with plants and animals.

In addition to indicating the additive nature of the biological life-form encoding sequences, the positive associations with societal complexity also imply that earlier language states usually would have had fewer animal and plant life-forms than present states. Since the last several millennia of human history have generally involved increases in societal complexity, it is likely that languages spoken thousands of years ago usually would not have had more animal and plant life-form terms than, or even the same number as, any of their respective contemporary - 85 daughter languages.

These hypotheses have been examined in several studies in which life-form systems of proto-stages in language family histories have been reconstructed (Brown 1979b, Brown and Witkowski 1981, Brown 1980). In each investigation they have been strongly confirmed. For example, reconstruction of plant and animal life-form vocabularies for Proto-Mayan (Brown 1979b and Witkowski and Brown 1981 respectively) shows that this parent language possessed only one plant life-form, “tree”, and only one animal life-form, “snake”, of the encoding sequences. All contemporary Mayan daughter languages have the latter life-forms plus additional ones. Reconstruction of the botanical life-form system of Proto-Polynesian (Brown 1980) similarly shows that the parent language had fewer plant life-forms than contemporary daughter languages have now. Only “tree” reconstructs for Proto-Polynesian while the vast majority of daughter languages have encoded botanical life-forms beyond “tree” such as “grass” and “vine”.

Reconstruction of Polynesian zoological life-form systems indicates a tendency for addition of animal life-forms since Proto-Polynesian times. Three life-forms, “fish”, “bird” and “snake”, unambiguously reconstruct for Proto-Polynesian. The majority of daughter languages have four zoological life-forms, i.e. the latter three and “wug”. Consequently, addition of animal life-form categories has been the prevalent pattern in Polynesian language history. However, a few daughter languages have only three animal life-forms and thus have not increased the size of their zoological life-form inventories beyond that of Proto-Polynesian, an unexpected result in view of findings of earlier studies. In addition, at least two contemporary Polynesian languages have fewer animal life-forms than the group's parent language possessed, an even more surprising result.

If daughter languages possess fewer zoological life-forms than their parent language, it follows that life-form loss must have occurred at some point in the group's history. This, of course, is contrary to the proposal that languages overwhelmingly tend to add biological life-forms over time in response to increases in societal complexity. The unexpected findings reported here, however, are not particularly disturbing since they are clearly traced to ecological conditions, i.e. lack of life-form exemplars, particularly snakes, in the habitats of many Polynesian language speakers. Rather, the validity of the proposal relating life-form growth and societal complexity is even more evident, given the fact that, despite such ecological constraints, most Polynesian languages have still managed to expand their animal life-form vocabularies in size beyond that of Proto-Polynesian.

- 86

In addition to lexical reconstruction, this study treats systematic ways in which Polynesian languages have acquired folk zoological life-forms, continuing the investigation of principles of biological nomenclatural development begun by Brent Berlin (1972) and built upon by Brown (1979b, 1980) and Brown and Witkowski (1981).


The critical features associated with the five zoological life-form classes of the encoding sequence (Figure 1) are as follows (Brown 1979a):

  • “fish” Creature possessing fins, gills and a streamlined body; adapted to an aquatic environment. This life-form is occasionally extended to other aquatic animals lacking some or all of these features, e.g., whales and aquatic crustaceans. In such cases true fish usually constitute the focal members of the class (Hunn 1977:250).
  • “bird” Creature possessing feathers, wings and a bill or beak; adapted to flying. This life-form is occasionally extended to bats or even flying insects. In such cases true birds usually constitute focal members of the class.
  • “snake” Featherless, furless, elongated creature usually lacking appendages; adapted to crawling. This life-form in its greatest extension includes snakes, worms, lizards, eels and occasionally, other elongated creatures such as reptile-like insects.
  • “wug” Small creature other than those included in “fish”, “bird” and “snake”. This life-form always encompasses bugs, i.e. insects and other very small creatures such as spiders and frequently is extended to worms. Occasionally the category also includes other creatures such as lizards, tortoises, and frogs if these are small.
  • “mammal” Large creature other than those included in “fish”, “bird” and “snake”. This life-form is sometimes restricted to mammals, but more often is extended to other large animals such as iguanas and crocodiles and, in addition, to such creatures as tortoises and frogs if these are large. 2

The lexical encoding sequence of Figure 1 represents a series of stages in the growth of folk zoological life-form vocabularies. Languages at Stage O altogether lack life-forms. Languages at Stage 1 encode one, those at Stage 2 encode two and those at Stage 3 encode three of the triad “fish”, “bird” and “snake”, but in no particular order. Stage 4 languages either add “wug” or “mammal” and Stage 5 languages add the remaining item. The encoding priority of “fish”, “bird” and - 87 “snake” may be due to their considerable distinctiveness as natural discontinuities vis-à-vis the relative lack of distinctiveness of “wug” and “mammal” (Brown 1979a). Creatures included in the latter two life-forms are residual animals, those left over after the highly distinctive discontinuities “fish”, “bird” and “snake” are encoded. The encoding of “wug” and “mammal” involves a binary contrast based on residual animal size, i.e. “small residual animal”/“large residual animal”.

While zoological life-forms are typically encoded through use of a single label, they are sometimes lexically realised in other ways. For example, languages may lack a term for “bird”, i.e. a class extended to birds in general, but lexically encode the life-form through the binary opposition “large bird”/“small bird” (Brown 1979a). Among the five zoological life-forms, “snake”, which includes elongated animals in general, is most frequently recognised through binary opposition, i.e. “small elongated animal”/“large elongated animal”. The “small elongated animal” category usually encompasses worms alone while the “large elongated animal” class is usually restricted to snakes. In addition, these two categories always encompass worms and snakes respectively; consequently, a class restricted to lizards, for example, is not considered a “snake” life-form class even though lizards are elongated creatures.


Polynesian languages are dispersed over a vast area of the Pacific Ocean (Figure 2). They are descended from Proto-Polynesian which broke up about 2,500 years ago (Pawley and Green 1971). The Polynesian group and another language group, Fijian, form a Central Pacific branch of Eastern Oceanic within the Oceanic division of the Austronesian language family (Pawley and Green 1973). The break-up of Proto Central Pacific is estimated to have taken place between 3,000 and 4,000 years ago. A genetic classification of Polynesian languages adapted from Biggs (1978) is presented in Table 1.

Table 2 lists folk zoological life-form terms for 24 Polynesian languages. These include representatives from all major and minor subgroups of the family. Language sources, mostly dictionaries, are listed in the Appendix. In Table 2 a question mark (?) appearing in a column indicates that the presence or absence of a life-form term in a language is uncertain, since pertinent sources are not particularly thorough. A line (—) in a column indicates that a life-form term is lacking. A slash (/) occurs between alternative animal terms when these are - 88 found in languages. Terms have been transliterated, when necessary, into standard linguistic symbols. 3

Geographic distribution of Polynesian languages
- 89

Genetic classification of Polynesian languages (adapted from Biggs 1978)

  • Austronesian Family:,
  • Oceanic:
  • Eastern Oceanic:
  • Central Pacific:
  • Fijian
  • Tongic:
  • Niuean
  • Tongan
  • Nuclear Polynesian:
  • Samoic Outlier:
  • East Uvean
  • Samoan
  • East Futunan
  • Rennellese
  • Pileni
  • Anutan
  • Tikopian
  • Mele-Fila
  • Sikaiana
  • Luangiua
  • Kapingamarangi
  • Nukuoro
  • Eastern Polynesian:
  • Easter Island
  • Central Eastern Polynesian:
  • Marquesic:
  • Mangarevan
  • Marquesan
  • Hawaiian
  • Tahitic:
  • Tahitian
  • Mangaia
  • Rarotongan
  • Tuamotuan
  • Rapa
  • Maori
- 90

Polynesian folk zoological life-form lexicons

LANGUAGES “fish” “bird” “snake” (elongated creatures in general)   “wug”
      “snake” “worm”  
Niuean ika manu lele ŋata kelemutu moki
Tongan ika manu/manupuna ŋata ?inise·kite
East Uvean ika manu lele ŋata
Samoan i?a manu/manulele ŋata ?anufe meaola la?itiiti
East Futunan ika moa* ŋata
Rennellese ika/kaui manu gege ŋata kutu
Pileni ika manu ŋata paipe ?
Anutan ika manu
Tikopian ika manu ? unufe nono*
Mele-Fila ika manu ŋata
Sikaiana ika manu ? ? ?
Luangiua i?a/ŋi?a maŋu ?ako ? ?
Kapingamarangi iga manu/manu ma·ŋi
- 91
Table 2 (cont.)
LANGUAGES “fish” “bird” “snake” (elongated creatures in general) “wug”
      “snake” “worm”  
Nukuoro iga/mamu manu lele labodo ilo
Easter Island ika manu koreha koreha ko?ura 4
Mangarevan ika manu
Marquesan ika manu puhi henua/puhi uta i?o
Hawaiian i?a manu naheka/nahesa/mo?o ko?e mea kolo*/holoholona lele*
Tahitian i?a/paru manu/?apa?apa ?o·fi· iro* manumanu
Mangaia ika/mangaika manu ? ? mangaika/potipoti
Rarotongan ika manu-rere o·vi· toke manumanu
Tuamotuan ika/paru/hu·raru manu/kupakupa toke* manu-manu
Rapa ika ? ? ? ?
Maori ika/ŋohi manu na·kahi/neke ŋunu ŋa·rara/mu·*
- 92

All phonological reconstructions presented in following sections are based on regular Polynesian sound correspondences. For an outline of these and the proto-segments they reflect, consult Biggs (1978).


In the following sections, referential histories of Polynesian terms for folk zoological life-forms are reconstructed. In three cases, contemporary terms are derived from ancestral stems, which labelled animal life-forms in Proto-Polynesian. In other cases, however, life-forms have been acquired by languages since Proto-Polynesian times. Acquisition has involved both borrowing and innovation. Most, if not all, life-form terms acquired through borrowing have come from languages outside of the Polynesian group.

Polynesian languages have often innovated zoological life-forms by adapting old terms. As in Mayan (Brown and Witkowski 1981), this has primarily involved three mechanisms of lexical change: (a) use of metaphor, (b) expansion of reference, and (c) restriction of reference. Innovation through metaphor entails drawing parallels between animals and other objects (sometimes other animals), whereby terms for the latter are used to denote the former. An example of a metaphorical equation used by Polynesians in constructing life-forms is “snake” = “eel”. Expansion and restriction of reference involve respectively increasing and decreasing the designative ranges of terms, e.g., extending the referential range of a term for a certain kind of insect, to bugs in general (“wug”), or the reverse, restricting the range of a term for all nonsea creatures (mammals, reptiles, birds and bugs), to just bugs in general (“wug”). Both of the latter example changes have occurred in Polynesian languages.

A fourth method of innovation sometimes employed by Polynesian languages, involves production of descriptive labels for life-form classes. “Wug”, for example, in one language is denoted by an expression translating literally “small alive thing”.


All “fish” terms listed first in the “fish” column of Table 2 are cognate, demonstrating regular sound correspondences. These attest to a Proto-Polynesian term for “fish” which reconstructs *ika (Biggs 1979). 5

Reflexes of *ika in many Polynesian languages are expanded in reference beyond true fish to include other sea creatures such as whales, porpoises, turtles, eels, octopus, etc. This expanded range is rarely the same in any two languages. For example, Mele-Fila ika encompasses whales - 93 and porpoises in addition to fish, but not turtles and octopus. On the other hand, Tahitian i?a includes turtles and octopus in addition to whales, porpoises and fish. In one language, Easter Island, an *ika reflex is even occasionally used in reference to creatures in general. Whatever its exact extension in any one language, the semantic focus of the form always seems to be true fish, a focus which almost certainly pertained to the Proto-Polynesian term. It is, of course, possible that Proto-Polynesian *ika had an expanded referential range similar to that of some daughter language reflexes.

Six Polynesian languages have acquired alternative “fish” labels (Table 2). Two of these, Tahitian and Tuamotuan, share an alternative term, paru. This label traces to Proto-Polynesian *palu (Biggs 1979), which designated some particular fish species. Reflexes of *palu are found in most Polynesian languages and usually denote some specific kind of fish (see Biggs 1979 for reflexes and referents). Use of a reflex of *palu as a term for fish in general, developed through expansion of reference. This may have occurred independently in Tahitian and Tuamotuan. It is, of course, possible that these languages share paru as “fish” label as a result of language contact.

Expansion of reference may also account for another alternative “fish” term, i.e. Nukuoro mamu. An identical term, mamu, designates a variety of fish in Tuamotuan. Possibly related forms denoting specific types of fish occur in several other languages: Tongan mamo “kind of fish”, Samoan mamo “small coral fish”, Hawaiian mamo “fish . . . about seven inches long” and Tahitian mamo “a small fish”. (It is, however, difficult to motivate a change from stem final o to u or vice versa.) If the range of Nukuoro mamu is not expanded from some fish species, it is possible that the term and its meaning are borrowed from outside the Polynesian group. Carroll and Soulik (1973), in their dictionary of Nukuoro, note that neighbouring Ponopean (Micronesian) uses mam in reference to fish in general and suggest this as the source of the Nukuoro term.

The second Mangaia “fish” term, mangaika, apparently combines words for “food” manga, and “fish”, ika. I have been unable to determine probable origins of other alternative “fish” terms listed in Table 2.


A “bird” label, *manu (Biggs 1979), unambiguously reconstructs for Proto-Polynesian. 6 Reflexes of this term in contemporary daughter languages either stand on their own as labels for “bird”, or as constituents of compound terms for “bird” (Table 2).

Reflexes of *manu in some Polynesian languages label broad - 94 zoological classes variously encompassing such creatures as land mammals, reptiles and insects, in addition to birds. Such a category is sometimes overtly characterised as constituting a “nonsea creature” grouping that is in direct contrast with a “sea creature” grouping, e.g., Tahitian (Lemaitre 1973:78). In addition, birds often form the semantic focus of these broad classes; in other words, some *manu reflexes have both the restricted sense of “bird” and the general sense of “nonsea creature”. Polysemy of this type pertains to reflexes of *manu found in Tongan, Samoan, Anutan, Tikopian, Kapingamarangi, Easter Island, Mangarevan, Tahitian, Mangaia and Tuamotuan.

In other Polynesian languages *manu reflexes do not have the restricted sense of “bird”, but rather refer only to nonsea creatures in general. When this is the case, birds are usually designated by compound terms which combine reflexes of *manu with words meaning “to fly” and “flying” thus creating descriptive labels translating “flying nonsea creature”: see terms for “bird” listed in Table 2 for Niuean, East Uvean, Rennellese, Nukuoro and Rarotongan. In addition, some languages having polysemous *manu reflexes also use compound terms as alternative “bird” labels which translate “flying nonsea creature”: see second terms for “bird” listed in Table 2 for Tongan, Samoan and Kapingamarangi. In several Polynesian languages the meaning of *manu reflexes is restricted to “bird” alone: these include Pileni, Mele-Fila, Sikaiana, Luangiua, Marquesan, Hawaiian and Maori. Finally, many Polynesian “bird” terms, both unitary and compound, are extended in reference to other winged creatures such as bats and flying insects.

It is possible that Proto-Polynesian *manu had a wider application than “bird” given the common association of very broad zoological classes with contemporary reflexes. However, there are reasons for believing that “bird” constituted the primary, if not the only, zoological referent of the Proto-Polynesian term and that daughter languages, in some cases independent of one another, expanded their reflexes of *manu to additional creatures.

Most of the islands settled by Polynesian speakers have very limited faunal inventories today and these were even more restricted in the past. Native mammals, other than those adapted to aquatic environments, are rare in Polynesia and in other parts of Oceania. Most of the land mammals that are now common were imported by man either on purpose (domesticated beasts) or accidentally (rats). Other creatures, such as snakes, and even bugs on some islands, are also noteworthy for their scarcity. On the other hand, birds, while also somewhat restricted in number and variety in many parts of Oceania (Pawley and Green 1971:17), nevertheless must have been relatively prominent creatures in - 95 the past compared with exemplars of some other life-form classes.

Assuming the prominence of birds, for Proto-Polynesian speakers and speakers of early daughter languages, animals that were not fish or sea creatures were, for the most part, *manu. This may have meant that things called *manu conceptually were not just “birds”, but also “creatures that do not live in the sea”. If so, it would have been natural to extend referential application of *manu reflexes to other nonsea creatures when these were introduced into Polynesian habitats. Fuentes (1960:782) mentions just this sort of extension for the Easter Island language: “the original meaning of manu was bird, but once animals were brought into the island they became manu since Pascuense was lacking a word to describe them”. In addition, Tregear (1891:207-8) notes the recency of similar extensions of *manu reflexes in Mangaia and Samoan.

If the above interpretation is correct, then some Polynesian languages (Pileni, Mele-Fila, Sikaiana, Luangiua, Marquesan, Hawaiian and Maori) have been conservative, having retained the original range of *manu, i.e. “bird”, without expanding it to additional creatures. Other languages (Tongan, Samoan, Kapingamarangi, Anutan, Tikopian, Easter Island, Mangarevan, Mangaia, Tuamotuan and Tahitian), have been less so, having retained “bird” as a restricted sense of the term while acquiring a second broader application to nonsea creatures in general through expansion of reference. Yet other languages (Niuean, East Uvean, East Futunan, Rennellese, Nukuoro and Rarotongan), have not been conservative at all, having expanded the range of *manu to nonsea creatures in general with the subsequent loss of its original meaning, “bird”. It is interesting to note that Nukuoro has carried referential expansion of its reflex one dramatic step further. Nukuoro manu apparently designates “living things in general” including both animals and plants.

It is inconceivable that initial expansion of *manu to nonsea creatures in general in any Polynesian language would have brought about immediate loss of “bird” as a restricted referent. Expansion at first would create a polysemous condition in which *manu meant both “bird” and “nonsea creature” (a condition pertaining to several contemporary Polynesian languages). The creation of polysemy apparently has led to the linguistic modification of *manu in some languages, possibly as a strategy to disambiguate competing referential applications. Thus, for example, while Tongan manu designates both “bird” and “nonsea creature”, the language provides an alternative compound term, manupuna, literally, “flying nonsea creature”, which unambiguously denotes “bird”. Such a development could free *manu reflexes, as - 96 unitary terms, to function solely as labels for “nonsea creature”. As a consequence, several Polynesian languages use reflexes of *manu only in the expanded sense, while designating “bird” only through use of compound terms derived through linguistic modification of *manu reflexes. 7

Two Polynesian languages, Tahitian and Tuamotuan, have alternative “bird” terms that are not derived from *manu reflexes (Table 2). The Tahitian term, ?apa?apa, is a reflex of Proto-Polynesian *kapakapa which meant “flap wing” (Biggs 1979). Derivation of the Tahitian meaning obviously involved conceptualisation of birds as “wing flapping creatures”. The provenience of the Tuamotuan term, kupakupa, is not so clear. It is probably related to Tahitian ?upa?upa which designates a certain species of bird. This suggests that the Tuamotuan term developed as a “bird” label through expansion of reference. Development of these alternative “bird” terms in Tahitian and Tuamotuan may be a response to the development of application of their reflexes of *manu.

East Futunan differs from all other Polynesian languages in that it apparently does not use its reflex of *manu, either as a unitary lexeme or in a compound label, in reference to “bird”. (East Futunan manu refers to nonsea creatures in general.) The “bird” term listed for the language in Table 2, moa, is marked by an asterisk indicating that it may not be a full-fledged “bird” term. In this case, the word's gloss in the dictionary consulted is somewhat ambiguous but, nonetheless, suggestive of a “bird” application. East Futunan moa is a reflex of Proto-Polynesian *moa which meant “fowl” (Biggs 1979). If the term is a “bird” life-form label, it developed as such through expansion of reference. 8


The third column of Table 2 lists Polynesian terms for elongated creatures in general. Only one Polynesian language, Easter Island, uses a single word for snakes and worms. Nukuoro has a term encompassing snakes, eels and long worms and labels remaining worms with a different word. Other Polynesian languages lexically encode “snake” (large elongated creature) and “worm” (small elongated creature) separately. “Snake” reconstructs for Proto-Polynesian while “worm” does not.

The Proto-Polynesian word for “snake” was *ŋata (Biggs 1979). 9 Nine of the 24 languages surveyed have reflexes of the Proto-Polynesian stem that designate “snake”: Niuean, Tongan, East Uvean, Samoan, East Futunan, Rennellese, Pileni, Mele-Fila and Luangiua (Table 2). These languages are affiliated with both the Tongan and Samoic Outlier subgroups of Polynesian (Table 1). Reflexes of *ŋata are also found in languages of the Eastern Polynesian subgroup, but they do not designate “snake”. These are discussed presently.

- 97

At least four Polynesian languages do not have terms for “snake” (large elongated creature) (Table 2). This means that they have lost the zoological life-form at some point in their development from Proto-Polynesian. In addition, seven of the 24 languages have replaced the original Proto-Polynesian word for “snake” with new terms for the category (Table 2). Both life-form loss and term replacement are traced to the fact that terrestrial snakes are scarce or totally missing in many parts of Oceania settled by speakers of Polynesian languages.

There is abundant evidence, both archaeological and linguistic, suggesting that the islands of Fiji (Figure 2), are the point from which pre-Polynesians moved into Polynesia (Pawley and Green 1971:14-5). Pawley and Green (1971:22) note that Fiji is the only region close to the Polynesian Triangle area of the Pacific that has snakes. (Easter Island, Hawaii and New Zealand constitute the three defining points of Triangle Polynesia). In addition, with the exception of Samoa, all islands of Triangle Polynesia lack snakes and only a few others (e.g. Tonga, Futuna and Uvea), could have maintained some knowledge of snakes by virtue of their continued contact with Fiji or Samoa. It is noteworthy that the Samoan language has a reflex of Proto-Polynesian *ŋata which means “snake” and that most of its genetic cousins in the Samoic Outlier subgroup spoken to the west of the Triangle area, have the same. As mentioned above, languages of the Eastern Polynesian subgroup have reflexes of *ŋata but these do not designate “snake”. These languages are all spoken in parts of Polynesia lacking snakes.

When Polynesian speakers moved into Triangle Polynesia, they diversified linguistically and at the same time tended to lose “snake” life-forms in response to the absence of snakes. In some cases this meant that reflexes of *ŋata were entirely lost and in others that their meanings were shifted to referents other than snakes. At least two Eastern Polynesian languages have completely lost *ŋata reflexes, i.e. Easter Island and Marquesan. The acquisition of new referents took different paths in different languages. Several languages shifted ŋata reflexes to other elongated creatures which are not snakes: Hawaiian naka “a land shell [snail?]”/“a sea shell”, Mangaia ŋata “sea slug”, Rarotongan ŋata “a species of shellfish”, Maori *ŋata “slug, snail”. In another language, Tuamotuan, the reflex shifted to nonzoological living objects characterised by elongation: ŋata “a stem, stalk, vine”. And, finally, in two other languages reflexes totally lost a substantive application: Tahitian ?ata?ata (reduplication) “shocking, disgusting” 10 and Mangarevan ŋatatata (partial reduplication) “to crawl”.

Today, some Eastern Polynesian languages have re-acquired “snake” terms, almost certainly due to contact with Europeans. The fact that very - 98 few of these languages share terms for “snake” (Table 2) is strong evidence that they all lacked such words in the recent past despite the fact that their ancestors of 2,500 years ago had a “snake” term. At least two Eastern Polynesian languages, Tuamotuan and Mangarevan, did not reacquire a “snake” life-form at all.

Two Samoic Outlier languages, Anutan and Kapingamarangi, also lack “snake” terms. This too almost certainly traces to scarcity or lack of snakes in environments of their speakers. These languages and a third member of Samoic Outlier, Nukuoro, have also totally lost reflexes of *ŋata. Nukuoro, however, has a “snake” life-form labelled by a replacement term, labodo.

When Polynesian languages re-acquired “snake” life-forms, they did so in a number of different ways. Most borrowed “snake” terms from Western languages indicating that the primary stimulus for reacquisition was European contact. Both Hawaiian and Maori borrowed “snake” labels from English (Hawaiian naheka/nahesa and Maori neke). 11 According to Williams (1971:503), Maori also acquired a second “snake” term (na·kahi) from Hebrew (nagash), a result of missionary influence. In order to translate the Bible into Polynesian languages missionaries created a large number of animal names from Hebrew, Greek and Latin. Missionary influence also may account for the phonologically similar “snake” terms of Tahitian and Rarotongan, ?o·fi· and o·vi· respectively. Ross Clark (personal communication) notes that these relate to the Greek word for “snake”, ophis. The Easter Island “snake” label, koreha, which encompasses most elongated creatures including snakes eels and worms, is possibly an adaptation of Spanish culebra “snake”.

Use of metaphor underlies derivation of “snake” terms in Marquesan and, possibly, in Nukuoro. Marquesan has two terms for “snake”, puhi henua and puhi uta, both of which translate literally “land eel”. Thus, Marquesan has developed a “snake” life-form by metaphorically associating snakes with other elongated creatures. Nukuoro, labodo, which encompasses snakes, eels and long worms, is possibly related to Kapingamarangi labodu which designates a certain man made object, i.e. a belt used in cloth looms. If related, the metaphoric equation “snake” = “belt”, underlies derivation of the Nukuoro referent. These terms are possibly borrowed from outside the Polynesian group. A phonologically similar term, laboto, designates a flexible elongated object, i.e. a certain vine species, in the Raluana language of the Bismarck Archipelago branch of Oceanic (Lanyon-Orgill 1960). Raluana is spoken on New Britain, which is located due south of the Nukuoro/ Kapingamarangi area.

Finally, Hawaiian has acquired a third “snake” life-form label - 99 through expansion of reference. The term mo?o, extended to snakes and lizards, is a reflex of Proto-Polynesian *moko which designated lizards alone (Biggs 1979). Hawaiian speakers, then, expanded a category of legged elongated creatures to large elongated creatures in general, including those lacking legs.

At least eight Polynesian languages lack a term for “worm”. In addition, very few of those that do have “worm” share related labels for the category (Table 2). This indicates that a term for small elongated creatures in general does not reconstruct for Proto-Polynesian. Nor is there evidence indicating that “worm” reconstructs for any subgroup parent language. Polynesian languages having “worm” terms have all acquired them through expansion of reference.

Three Polynesian languages, Hawaiian, Rarotongan and Tuamotuan, have related terms for “worm” (Table 2). These are all reflexes of Proto-Polynesian *toke (cf. Biggs 1979) which probably referred restrictively to conger eels and expansively to sea eels in general. The development of “worm” as a referent of this form involved a somewhat complex series of lexical changes underlain by both use of metaphor and expansion of reference.

The Niuean reflex, toke, apparently retains both the restrictive and expansive senses of the proto-stem. The Tongan reflex (toke) also seems to have both meanings, but sources consulted inform us only that the term refers to a certain “kind of sea-eel” in addition to sea eels in general. East Uvean and Samoan reflexes designate moray eels and apparently do not have wider applications. The Anutan reflex has developed as a term for eels in general through expansion of reference. Reflexes in Eastern Polynesian languages for the most part designate specific kinds of worm or worms in general rather than sea eels, a development discussed in detail presently. However, the original meaning of the proto-stem is reflected by Tuamotuan and Maori forms which are reduplicated and partially reduplicated: pi·toketoke “a variety of eel” and Maori totoke “a species of conger eel”. The existence of these two forms constitutes the basis for reconstructing “sea eel” rather than “worm” as the principal meaning of Proto-Polynesian *toke. “Conger eel” is tentatively reconstructed as a specific sense of the proto-stem on the basis of meanings associated with reflexes in Niuean and Maori.

Proto-Polynesian *toke “sea eel”/“conger eel” contrasted semantically with Proto-Polynesian *tuna “freshwater eel” (see Biggs 1979 for reflexes and referents). After breakup of Proto-Polynesian, speakers of Proto-Nuclear Polynesian innovated a form which apparently competed with *toke as a label contrasting with *tuna. The innovated form, *pusi, seems to have had the restricted sense of “moray eel” in addition to an - 100 application to sea eels in general. Reflexes of the form are found only in languages of the Nuclear Polynesian subgroup (Table 1). Examples are Samoan pusi “general name of moray eels”, Rennellese pusi “starry or clouded moray eel”, Mangarevan puhi “the sea-eel (Murena) that bites and tears”, Tahitian puhi “an eel, commonly a sea-eel” and Tuamotuan puhi “the eel . . . a general term”.

Competition of reflexes of *toke and *pusi for the semantic space, “sea eel”, apparently led to a meaning shift of the former in Proto-Eastern Polynesian, a daughter language of Proto-Nuclear Polynesian. Proto-Eastern Polynesian *toke lost “sea eel” as a referent and developed the meaning “earthworm”. Reflexes with the latter referent include Marquesan toke, Tahitian to?e and Maori toke. Other Eastern Polynesian languages have reflexes with similar referents: Mangarevan toketoke “a marine worm resembling an earthworm” and iritoke “an earthworm” and Tuamotuan toke “angleworm . . . mosquito larva”. The shift from “sea eel” to “earthworm” was accomplished by metaphor, whereby one type of elongated creature is equated with another type (an equation similar to the one reported above underlying derivation of Marquesan “snake”). This shift parallels that involving another Polynesian eel term. Reflexes of Proto-Polynesian *tuna “freshwater eel” have developed the referents “caterpillar” in Marquesan and “insect larva” in Nukuoro.

The original meaning of Proto-Polynesian *toke “sea eel”/“conger eel”, however, was not entirely lost in Eastern Polynesian. When the shift from “sea eel” to “earthworm” occurred, the original sense was preserved in reduplicated reflexes of the stem. Thus, for the most part, nonreduplicated reflexes of *toke have worm related meanings in contemporary Eastern Polynesian languages, while reduplicated reflexes have eel related meanings: Tuamotuan pi·toketoke “a variety of eel” and Maori totoke (partial reduplication) “a species of conger eel”.

Finally, the contemporary use of Hawaiian, Rarotongan and Tuamotuan reflexes of *toke as labels for small elongated creatures in general is traced to expansion of reference. These reflexes, of course, have developed from Proto-Eastern Polynesian *toke which designated “earthworm” rather than worms in general.

Niuean's “worm” term, kelemutu, also probably developed through expansion of reference from “earthworm”. The same term occurs in neighbouring Tongan, East Uvean and East Futunan with the referent “earthworm” and in Samoan where it designates “grub found in rotton wood”. Kelemutu is, clearly, a diffused item as it is confined to a relatively small, geographically continuous area of Polynesia and crosscuts a subgroup boundary, i.e. that of Samoic Outlier. 12

- 101

Two languages of different Polynesian subgroups, Samoic Outlier and Eastern Polynesian, have independently derived “worm” terms from the same Proto-Polynesian stem. Nukuoro ilo and Tahitian iro (Table 2) respectively are reflexes of Proto-Polynesian *iLo “maggot” (Walsh and Biggs 1966:21). 13 The Nukuoro and Tahitian meanings, of course, have developed through expansion of reference.

Samoan and Tikopian apparently have also independently developed “worm” from the same Proto-Polynesian form. Both languages have expanded reflexes of Proto-Polynesian *?anufe “caterpillar” (Biggs 1979) to small elongated creatures in general. These reflexes, ?anufe and unufe respectively (Table 2), encompass caterpillars in addition to elongated creatures typically included in “worm” life-form classes.

Finally, I have been unable to determine proveniences of “worm” terms in Pileni and Maori (Table 2).


Several Polynesian languages lack “wug” terms. In addition, very few of those having “wug” have related terms for the category (Table 2). “Wug” clearly does not reconstruct for Proto-Polynesian. However, it is possible that a “wug” label pertained to at least one subgroup parent language, i.e. Proto-Tahitic. Polynesian languages have acquired “wug” labels through both expansion and restriction of reference and by innovating descriptive expressions for the category. In addition, at least one language has borrowed its “wug” label from outside the group.

Three members of the Tahitic subgroup, Tuamotuan, Rarotongan and Tahitian, have the same label for “wug”, manumanu. In Tuamotuan and Tahitian the term extends to bugs in general. However, Rarotongan manumanu refers only to all small insects and, therefore, handles only one half of the binary opposition “small wug”/“large wug”. 14 Manumanu is a reduplicated reflex of Proto-Polynesian *manu “bird”. The three languages having it also use a nonreduplicated *manu reflex to refer to nonsea creatures in general including bugs. Thus, development of “wug” in these three languages has involved both reduplication and restriction of meaning from a referential range encompassing all land and air creatures. Reduplication in Oceanic languages is commonly used to indicate “a kind of X”, where the simple base represents X (Andrew Pawley, personal communication). While it is possible the manumanu “wug” pertained to Proto-Tahitic, on distributional grounds it is just as likely that it is a diffused item. For example, it does not occur in Maori, the most geographically remote language of the Tahitic subgroup.

Two languages of different subgroups, Tongic and Eastern Polynesian, have independently developed “wug” through expansion of refer- - 102 ence of the same Proto-Polynesian stem. Both Niuean and Mangaia use reflexes of Proto-Polynesian *moko “lizard” as “wug” labels. (This proto-stem, of course, is mentioned above in connection with derivation of Hawaiian moko “snake”.) The Niuean and Mangaia terms, both moko, similarly extend referentially to insects, grubs, caterpillars, lizards and the like, all of which are small residual animals. In addition, “lizard” seems to constitute the semantic focus of both terms.

One of two Maori “wug” terms, ŋa·rara, is similarly derived through expansion of reference from “lizard”. The Maori label applies to insects and reptiles and has the secondary sense of “a demon, a reptile god, a god of evil”. Related terms referring to lizards are found in other languages of the Tahitic subgroup: Rarotongan ŋarara “a lizard which grows to the length of 12 inches” and Tahitian ?arara “lizard”. These three terms conceivably are all derived from a Proto-Tahitic stem, *ŋarara, which denoted some type of lizard or lizards in general. A fourth Tahitic language, Mangaia, has a related form, karara “a house lizard”. Since the latter is an unexpected reflex of *ŋarara, it is probably borrowed (perhaps from Northern Marquesan where k is an expected reflex of *ŋ). This indicates that the stem possibly did not pertain to Proto-Tahitic but rather diffused to several Tahitic languages. Whatever the exact historical details, the original meaning of the form almost certainly centred on lizards rather than insects.

Proto-Polynesian *iLo “maggot”, referred to above in connection with derivation of Nukuoro and Tahitian “worm” labels, also is ancestral to Marquesan i?o “wug”. The latter encompasses both worms and bugs. 15 Plausibly the Marquesan reflex of *iLo developed first as a label for worms in general through expansion of reference from “maggot” and subsequently expanded further to include bugs.

“Wug” has been innovated in several Polynesian languages by extending the referential application of a term for a specific type of bug to bugs in general. Rennellese kutu, denoting lice, leeches, mites and tiny insects of various kinds, is traced to Proto-Polynesian *kutu “louse” (Biggs 1979). Tikopian nono “gnat, sandfly, insect” is a reflex of Proto-Polynesian *nono which designated some very small flying insect such as a sandfly or midge (Biggs 1979). And Maori mu. is linked to Proto-Polynesian *mu· which denoted some destructive insect such as a moth (Biggs 1979).

Similarly, development of “wug” in Mangaia involved expanding the range of a term which originally designated some type of bug. Mangaia potipoti “any small insect” is a reflex of Proto-Central Eastern Polynesian *potipoti. Other related forms indicate that the proto-stem referred to a specific insect: Maori potipoti “the sandhopper”, Tuamotuan - 103 potipoti “small cockroach”, Rarotongan potipoti “sandflies”, Tahitian popoti “cockroach”, Hawaiian pokipoki “sowbug, pill bug (Isopoda)” and Mangarevan potipoti “the name of an insect”.

The Easter Island “wug” term, ko?ura, refers specifically to “flea” in addition to small insects in general. This suggests that the latter meaning is an expansion from the former. This term traces to Proto-Eastern Polynesian *ko?ura, which probably designated “crayfish” (cf. Green 1966:26-7). 16 The Easter Island “flea” application may have developed through a metaphoric usage equating fleas and crayfish, although the empirical basis for such an association does not seem particularly compelling.

Two Polynesian languages, Samoan and Hawaiian, have innovated descriptive labels for “wug”. These are Samoan meaola la?itiiti “small living thing”, Hawaiian mea kolo “creeping thing” and Hawaiian holoholona lele “jumping creature”.

Finally, Tongan borrowed its “wug” term, ?inise·kite, from English (insect).


Since no Polynesian language has a term for “large residual creature”, a “mammal” column is not included in Table 2. Lack of “mammal” is almost certainly traced in part to the fact that native terrestrial mammals and other large residual animals are not common in areas of the Pacific populated by Polynesian language speakers. As mentioned earlier, introduced species such as rats and domesticated mammals often are included in broad zoological classes encompassing nonsea creatures in general. The latter are usually labelled by reflexes of Proto-Polynesian *manu “bird”.


Known Polynesian folk zoological life-form inventories range in size from two to four life-forms (Table 2). No Polynesian language has a combination of terms for life-form terms of the encoding sequence indicating that encoding paths other than those outlined in Figure 1 have been followed. For example, the combination “bird” and “wug” (without “fish”, “snake” and “mammal”) could occur, but this and other out-of-sequence combinations are not found in Polynesian languages. Evidence of this sort from a large sample of globally distributed languages is, of course, the basis for asserting the existence of a universal encoding sequence for folk zoological life-forms (Brown 1979a, 1981b).

- 104

Language reconstruction also provides evidence consistent with the life-form encoding sequence. Three zoological life-forms, “fish”, “bird” and “snake” (large elongated creature) reconstruct for Proto-Polynesian. On the other hand, if, for example, “fish”, “snake” and “wug” reconstructed, but not “bird”, this would contravene the predicted order in which languages acquire zoological life-forms (Figure 1). Polynesian languages, then, have precisely the kind of reconstructed life-form system one would expect if such systems are, in fact, built up in accordance with the lexical encoding sequence.

My interpretation of the growth of folk zoological life-forms in the Polynesian family is summarised as follows. Proto-Polynesian, spoken at the latest some 2,500 years ago, was at Stage 3 in animal life-form growth, having the initial three life-forms of the encoding sequence. 17 In early post-Proto-Polynesian times, speakers of daughter languages settled numerous islands where snakes were missing or scarce, particularly islands of Triangle Polynesia. In response to this ecological condition, some languages, especially those of the Eastern Polynesian subgroup, lost “snake” life-forms and, thereby, regressed to Stage 2 in life-form growth, having only “fish” and “bird”. After breakup of all subgroup parent languages (e.g. Proto-Samoic Outlier, Proto-Marquesic, Proto-Tahitic), completed roughly 1,000 years ago (Green 1966:34), most Polynesian zoological life-form inventories began to expand. This involved addition of terms for “worm” and “wug” and also terms for “snake” in some of those languages which had lost the life-form. It is probable that much of this late growth took place in response to cultural changes brought about through contact with Europeans during the last several hundred years. Reacquisition of “snake” by Eastern Polynesian languages is clearly indicative of such an influence. Nearly all Eastern Polynesian “snake” terms were borrowed from Western languages. Finally, a few Polynesian languages (Anutan, Kapingamarangi and Mangarevan) did not participate at all in this late phase of life-form growth (Table 2).

As outlined earlier, the world's languages have overwhelmingly tended to add rather than lose biological life-form classes in response to recent massive increases in societal complexity of worldwide scope. Reconstruction of zoological life-form growth in the Polynesian family is exceptional since it shows that some daughter languages have actually lost a life-form found in the family parent language. One finding of the present paper is that ecological conditions, in this case absence of snakes in environments, can bring about reversals in life-form growth. This result is not totally expected, since, plausibly, terms for creatures such as snakes, missing in a habitat but once known to a people on the basis of - 105 first-hand experience, might be perpetuated through stories, myths, legends, rituals and so on for considerable periods of time.

That Polynesian languages do not have “mammal” life-forms is an expected finding of this investigation since native mammals and other large residual creatures are not common in environments of their speakers. An obvious conclusion, then, is that lack of life-form exemplars in an environment always inhibits lexical encoding of that life-form class. However, Polynesian language history itself proves such a conclusion unwarranted. Speakers of several Polynesian languages which lacked “snake” terms in the past nonetheless acquired them despite lack of snakes in their habitats—albeit under European influence.

Given the severely limited faunal inventories experienced by Polynesians, what may be totally unexpected is that a significant number of Polynesian languages have actually expanded their animal life-form vocabularies beyond that of Proto-Polynesian. Despite strong environmental constraints, Polynesian languages have nonetheless managed to participate in the worldwide trend for biological life-form addition documented in several earlier studies (Brown 1979b, Witkowski and Brown 1981, Brown 1980). This itself is testament to the considerable strength of the tendency to increase biological life-form inventories over time.


Sources for the twenty-four Polynesian languages surveyed are as follows: Anutan Feinberg (1977); Easter Island Englert (1938), Fuentes (1960); East Futunan Grézel (1878); East Uvean Bataillon (1932); Hawaiian Pukui and Elbert (1971); Kapingamarangi Lieber and Dikepa (1974); Luangiua Lanyon-Orgill (1944); Mangaia Christian (1924), Tregear (1891); Mangarevan Tregear (1891, 1899); Maori Tregear (1891), Williams (1971); Marquesan Dordillon (1931, 1932), Tregear (1891); Mele-Fila Biggs (1975), Capell (1942), Ross Clark (personal communication); Niuean McEwen (1970), Tregear and Smith (1907); Nukuoro Carroll and Soulik (1973); Pileni Ray (1921); Rapa Stokes (1955); Rarotongan Savage (1962); Rennellese Elbert (1975); Kuschel (1975); Samoan Milner (1966), Neffgen (1918), Tregear (1891); Sikaiana Capell (1936), Ray (1917); Tahitian Andrews and Andrews (1944), Jaussen (1969), Lemaitre (1973), Tregear (1891); Tikopian Durrad (1913), Firth (1940), Williams (1926, 1927); Tongan Churchward (1959), Missionnaires Maristes (1890); Tuamotuan Stimson and Marshall (1964), Tregear (1891).


Bruce Biggs, Robert Blust, Pierre Cabalion, Paul G. Chapin, Paul K. Chase, Ann Chowning, Ross Clark, Christian Clerk, Richard Feinberg, Harry Feldman, J. L. Fischer, Ward Goodenough, George W. Grace, Albert Heinrich, Charles F. Hockett, Rolf Kuschel, Andrew Pawley and Stanley R. Witkowski contributed to this study in various important ways. I am very grateful for their assistance. This material is based upon work supported by the National Science Foundation under Grant No. BNS-7906074.

- 106 Page of endnotes

- 107
  • ANDREWS, Edmund, and Irene D., 1944. A Comparative Dictionary of the Tahitian Language. Chicago, The Chicago Academy of Science.
  • BATAILLON, Pierre, 1932. Langue D'Uvea (Wallis). Paris, Libraire Orientaliste Paul Geuthner.
  • BERLIN, Brent, 1972. “Speculations on the Growth of Ethnobotanical Nomenclature.” Language in Society, 1:51-86.
- 108
  • —— Dennis E. BREEDLOVE, and Peter H. RAVEN, 1974. Principles of Tzeltal Plant Classification: An Introduction to the Botanical Ethnography of a Mayan-Speaking People of Highland Chiapas. New York, Academic Press.
  • BIGGS, Bruce, 1975. A Mele-Fila Vocabulary. Auckland, Linguistic Society of New Zealand.
  • —— 1978. “The History of Polynesian Phonology”, in S. A. Wurm and Lois Carrington (eds), Second International Conference on Austronesian Linguistics: Proceedings, Fascicle 2, Eastern Austronesian (Pacific Linguistics Series C—No. 61), pp.691-716. Canberra, The Australian National University.
  • —— 1979. Proto-Polynesian Word List II. Working Papers in Anthropology, Archaeology, Linguistics, Maori Studies, No. 53. Auckland, University of Auckland.
  • BROWN, Cecil, H., 1977. “Folk Botanical Life-Forms: Their Universality and Growth.” American Anthropologist, 79:317-42.
  • —— 1979a. “Folk Zoological Life-Forms: Their Universality and Growth.” American Anthropologist, 81:791-817.
  • —— 1979b. “Growth and Development of Folk Botanical Life-Forms in the Mayan Language Family.” American Ethnologist, 6:366-85.
  • —— 1980. “Growth and Development of Folk Botanical Life-Forms in the Polynesian Languages.” Unpublished manuscript, Department of Anthropology, Northern Illinois University.
  • —— 1981a. “More on Folk Zoological Life-Forms.” American thropologist 83(2) (in press).
  • —— 1981b. Principles of Folk Animal Classification. Unpublished manuscript, Department of Anthropology, Northern Illinois University.
  • —— and Stanley R. Witkowski, 1981. “Growth and Development of Folk Zoological Life-Forms in the Mayan Language Family.” American Ethnologist (in press).
  • CAPELL, A., 1936. “A Sikayana Vocabulary.” Journal of the Polynesian Society, 45:142-53.
  • —— 1942. “Notes on the Fila Language, New Hebrides.” Journal of the Polynesian Society, 51:153-80.
  • CARROLL, Vern and Tobias SOULIK, 1973. Nukuoro Lexicon. Honolulu, The University Press of Hawaii.
  • CASHMORE, Christine, 1969. “Some Proto-Eastern Oceanic Reconstructions with Reflexes in Southeast Solomon Islands Languages.” Oceanic Linguistics, 8:1-25.
  • CHRISTIAN, F. W., 1924. Vocabulary of the Mangaian Language. Honolulu, Bernice P. Bishop Museum.
  • CHURCHWARD, C. Maxwell, 1959. Tongan Dictionary (Tongan-English and English-Tongan). London, Oxford University Press.
  • CONKLIN, Harold C., 1954. The Relation of Hanunoo Culture to the Plant World. Unpublished PhD dissertation, Yale University.
  • DAHL, Otto Christian, 1973. Proto-Austronesian. Sweden, Studentlitteratur.
  • DORDILLON, Rene Ildefonse, 1931. Grammaire et Dictionnaire de la Langue
- 109
  • des Îles Marquises: Marquisien-Français. Paris, Institut d'Ethnologie.
  • —— 1932. Grammaire et Dictionnaire de la Langue des Îles Marquises: Français-Marquisien. Paris, Institut d'Ethnologie.
  • DOUGHERTY, Janet W. D., 1978. “Salience and Relativity in Classification.” American Ethnololgist, 5:66-80.
  • DURRAD, W. J., 1913. “A Tikopia Vocabulary.” Journal of the Polynesian Society, 22:86-95; 141-8.
  • ELBERT, Samuel H., 1975. Dictionary of the Language of Rennell and Bellona, Part I, Rennellese and Bellonese to English. Copenhagen, National Museum of Denmark.
  • ENGLERT, P. Sebastian, 1938. Diccionario Rapanui-Español: Radactado en la Isla de Pascau. Santiago, Prensas de la Universidad de Chile.
  • FEINBERG, Richard, 1977. The Anutan Language Reconsidered: Lexicon and Grammar of a Polynesian Outlier. New Haven, Human Relations Area Files, Inc.
  • FIRTH, Raymond, 1940. “The Analysis of Mana: An Empirical Approach.” Journal of the Polynesian Society, 31:483-510.
  • FUENTES, Jordi, 1960. Dictionary and Grammar of the Easter Island Language. Santiago, Chile, Editorial Andre Bello.
  • GREEN, Roger, 1966. “Linguistic Subgrouping within Polynesia: The Implications for Prehistoric Settlement.” Journal of the Polynesian Society, 75:6-38.
  • GRÉZEL, Le P., 1878. Dictionnaire Futunien-Français avec Notes Grammaticales. Paris, Maisonneuve et Cie, Libraires-Éditeurs.
  • HAYS, Terence E., 1976. “An Empirical Method for the Identification of Covert Categories in Ethnology.” American Ethnologist, 3:489-507.
  • HOCKETT, C. F., 1976. “The Reconstruction of Proto Central Pacific.” Anthropological Linguistics, 18:187-235.
  • HUNN, Eugene S., 1977. Tzeltal Folk Zoology: The Classification of Discontinuities in Nature. New York, Academic Press.
  • JAUSSEN, Tepano, 1969. Grammaire et Dictionnaire de la Langue Tahitienne. Paris, Musée de l'Homme.
  • KUSCHEL, Rolf, 1975. Animal Stories from Bellona Island (Mungiki). Copenhagen, National Museum of Denmark.
  • LANYON-ORGILL, P. A., 1944. A Study of the Leuangiua Language. London, Luzac & Co.
  • —— 1960. A Dictionary of the Raluana Language (New Britain, S. W. Pacific). Victoria, Canada, Published by the Author.
  • LEMAITRE, Yves, 1973. Lexique du Tahitien Contemporain. Paris, Office de la Recherche Scientific et Technique Outre-Mer.
  • LIEBER, Michael D., and Kalio H. DIKEPA, 1974. Kapingamarangi Lexicon. Honolulu, The University Press of Hawaii.
  • LINCOLN, Peter, 1979. Proto-Oceanic Cognate Sets. Unpublished computer printout.
  • McEWEN, J. M., 1970. Niue Dictionary. Wellington, Department of Maori and Island Affairs.
  • MILKE, Wilhelm, 1968. “Proto-Oceanic Addenda.” Oceanic Linguistics, 7:147-171.
- 110
  • MILNER, G. B., 1966. Samoan Dictionary. London, Oxford University Press.
  • MISSIONNAIRES MARISTES, 1890. Dictionnaire Toga-Français et Français-Toga-Anglais. Paris, Ch. Chadenat, Libraire-Éditeur.
  • NEFFGEN, H., 1918. Grammar and Vocabulary of the Samoan Language. London, Kegan Paul, Trench, Truber.
  • PAWLEY, Andrew and Kaye GREEN, 1971. “Lexical Evidence for the Proto-Polynesian Homeland.” Te Reo, 14:1-35.
  • PAWLEY, Andrew and Roger GREEN, 1973. “Dating the Dispersal of the Oceanic Languages.” Oceanic Linguistics, 12:1-67.
  • PUKUI, Mary K. and Samuel H. ELBERT, 1971. Hawaiian Dictionary. Honolulu, University Press of Hawaii.
  • RAY, Sidney, H., 1917. “Polynesian Linguistics. II.-Polynesian Languages of the Soloman Islands.” Journal of the Polynesian Society, 26:34-105.
  • —— 1921. “Polynesian Linguistics. IV.-Polynesian Languages of the Santa Cruz Archipelago. V. Vocabulary of the Pileni Language.” Journal of the Polynesian Society, 30:103-118.
  • SAVAGE, Stephen, 1962. A Dictionary of the Maori Language of Rarotongan. Wellington, The Department of Island Territories.
  • STIMSON, J. Frank and Donald Stanley MARSHALL, 1964. A Dictionary of Some Tuamotuan Dialects of the Polynesian Language. The Hague, Martinus Nijhoff.
  • STOKES, John F. G., 1955. “Language in Rapa.” Journal of the Polynesian Society, 64:315-40.
  • TREGEAR, Edward, 1891. The Maori-Polynesian Comparative Dictionary. (Reprinted 1969, Oosterhout N.B.) The Netherlands, Anthropological Publications.
  • —— 1899. A Dictionary of Mangareva (or Gambier Islands). Wellington, John MacKay, Government Printing Office.
  • —— and S. P. SMITH, 1907. A Vocabulary and Grammar of the Niue Dialect of the Polynesian Language. Wellington, Government Printer.
  • WALSH, D. S. and Bruce BIGGS, 1966. Proto-Polynesian Word List I. Auckland, Linguistic Society of New Zealand.
  • WILLIAMS, Herbert W., 1926. “A Tikopian Vocabulary.” Journal of the Polynesian Society, 35:267-89.
  • —— 1927. “A Tikopian Vocabulary” (continued). Journal of the Polynesian Society, 36:99-117.
  • —— 1971. A Dictionary of the Maori Language. Wellington, A. R. Shearer, Government Printer.
1   I have made some minor revisions in the folk botanical life-form encoding sequence based on substantial expansion of the cross-language data base (cf. Brown 1980). As originally formulated (Brown 1977), “tree” still proceeds all other botanical life-forms. As revised, both “greb” (small herbaceous plant) and “grass” have essentially equal chances of being added to languages immediately after “tree”. If “grerb” is encoded second, “grass”, “vine” and “bush” will be added subsequently, but in no particular order. If “grass” is second, “grerb”, “vine” and “bush” will be added subsequently, but in no particular order.
2   Animal is more commonly used than mammal as a “mammal” life-form label by speakers of English. Since animal is also used as a unique beginner term to refer to creatures in general, it is not employed as a life-form gloss to avoid ambiguity of reference.
3   Some Polynesian lexical sources inconsistently record vowel length or totally ignore it. Vowel length is given here as reported in sources consulted. In addition, there is occasionally some confusion in sources with respect to the presence or absence of glottal stops.
4   may not be full-fledged life-form term
5   Proto-Polynesian *ika “fish” is traced through Proto-Central Pacific, Proto-Eastern Oceanic and Proto-Oceanic to Proto-Austronesian *ikan “fish” (cf. Hockett 1976, Cashmore 1969, Lincoln 1979).
6   Proto-Polynesian *manu is traced through Proto-Central Pacific, Proto-Eastern Oceanic and Proto-Oceanic to Proto-Austronesian *manuk “bird” (cf. Hockett 1976, Cashmore 1969, Lincoln 1979, Dahl 1973).
7   Research completed just after final journal editing of this paper has led to a revision in the framework for interpreting expanded referential applications of *manu reflexes. A recent survey of a large number of globally distributed languages shows that languages occasionally lump large and small residual creatures (i.e. “mammals” and “wugs”) together and assign a single label to the grouping, thereby creating a combined “wug” + “mammal” category (Brown 1981a,b). Combined “wug” + “mammal” is sometimes acquired by languages by expanding the range of reference of a term for one of the initial triad of the encoding sequence, “bird”, “fish” and “snake” (Brown and Witkowski 1981; Brown 1981b). When such is the case, the “wug” + “mammal” class encompasses not only large and small residual creatures, but also those nonresidual creatures originally designated by the term so expanded. Thus, for example, Tzotzil, a Mayan language of Mexico, has a “wug” + “mammal” grouping encompassing snakes which was acquired by expanding the referential range of its “snake” label (Brown and Witkowski 1981). Other Mayan languages have “wug” + “mammal” categories encompassing birds in addition to residual creatures. These are labelled by terms originally restricted in reference to “birds”. Polynesian languages having *manu reflexes which extend to birds, bugs and mammals (i.e. to so-called “nonsea creatures”) seem to fit this pattern. In light of this, I would now describe the latter languages as having combined “wug” + “mammal” categories rather than “nonsea creature” classes, a treatment which is more in line with the emerging cross-language picture. In addition, Hawaiian holoholona “animal, beast, insect” (Pukui and Elbert 1971:73) is probably also a “wug” + “mammal” class, but one having a somewhat different derivational history than those labelled by *manu reflexes.
8   Moa is assigned the gloss “poul, volailles” in Grézel's (1887) East Futunan to French dictionary. Given the widespread Polynesian practice of constructing “bird” labels through modification of *manu reflexes meaning “nonsea creature”, it is odd that East Futunan should lack such a compound term. It is possible that the absence of the latter in the source consulted (Grézel 1887) is either an oversight or an elicitation gap.
9   Proto-Polynesian *ŋata “snake” is traced through Proto-Central Pacific (Hockett 1976) and Proto-Eastern Oceanic (Cashmore 1969) to Proto-Oceanic *ŋata “snake” (Milke 1968). There is little evidence that an ancestral form pertained to Proto-Austronesian.
10   Polynesians widely regard snakes as repulsive creatures to be carefully avoided.
11   In his dictionary Williams (1971) proposes that neke “snake” is an English loanword. Another possibility is that the form traces to Proto-Nuclear Polynesian *neke “move, edge along, crawl” (Biggs 1979).
12   Evidence assembled in my paper (Brown 1980) on Polynesian folk botanical life-forms, indicates that biological concepts have diffused extensively in that part of the Pacific occupied by speakers of East Uvean, East Futunan, Samoan and the Tongic languages (see map, Figure 2).
13   Walsh and Biggs (1966) use L to indicate that a proto-segment could have been either *l or *r. This ambiguity exists when cognate forms from Tongic languages are missing. Reflexes of Proto-Polynesian *iLo “maggot”, for example, do not occur in either Tongan or Niuean.
14   A convention adopted in Brown (1979a) is that a life-form class is judged present in a language even if only one-half of a binary opposition underlying it is encoded. Thus, for example, language are scored as having “snake” (elongated creatures in general) when they have “worm” (small elongated creature) but not “snake” (large elongated creature). Other Polynesian languages in addition to Rarotongan having terms for “small wug” include Mangaia with potipoti, Rennellese with kutu, and Easter Island with ko?ura.
15   Worms and bugs are typical constituents of “wug” life-forms (Brown 1979a). Wug is, of course, a mnemonic based on worm and bug.
16   Probable reflexes of Proto-Eastern Polynesian *ko?ura in addition to the Easter Island stem include Maori ko·ura “crayfish”, Rapa koura “sea crayfish”, Tuamotuan ko·ura “the rock-lobster, crayfish”, Rarotongan koura “general name for the crayfish species”, Tahitian ?oura “lobster, crayfish, shrimp” and Marquesan kou?a “shrimp, prawn, crayfish”.
17   Proto-Polynesian terms for “fish”, “bird” and “snake” were directly inherited from Proto-Oceanic through Proto-Eastern Oceanic and Proto-Central Pacific (see notes 4, 5 and 8 above). In addition, Proto-Oceanic terms for “fish” and “bird”, but not “snake”, are reflexes of Proto-Austronesian labels for the same life-forms. My survey of lexicons of a reasonably large number of Austronesian languages belonging to most of the other major and minor subgroups of the family, indicates the unlikelihood that “snake” and “worm” terms reconstruct for Proto-Austronesian. Therefore, it appears that Proto-Austronesian was affiliated with Stage 2 of zoological life-form growth having only “fish” and “bird”. Development of Proto-Oceanic from the family parent language involved an advance to Stage 3. This order of life-form growth, of course, is totally in accord with the zoological life-form encoding sequence.