Version classiqueVersion mobile
OpenEdition Books

Maenge gardens

Françoise Panoff

Chapter 3. Maenge systems of classification and appellation of domesticates

Texte intégral

  • 16 (Publisher’s note). From here to the end of the chapter, we have added subtitles or clarified the (...)

1In this chapter I shall deal first with Maenge systems of classification of domesticates and then with their nomenclature.16

Systems of classification

2Before describing Maenge systems of classification, it is necessary to define clearly two terms: key and taxonomy, because superficial similarities between these two types of classification still lead some authors to confuse one with the other. The definition adopted here derives in part from that of Conklin (1964: 39-41).

3Taxonomy. In a taxonomy, the constituent entities or taxa (i.e. labelled categories) are arranged vertically based on their degree of inclusiveness, the categories at one level being included in a category at the next higher level. Hierarchic positions in a taxonomy are thus not permutable.

4Key. A key is a succession of attribute oppositions which, by their hierarchical application, help to identify entities. Positions in a key may be permutable.

5In the next section I shall present a general outline of Maenge taxonomy of plants to provide background information concerning the main features of their classification of this sphere of the natural world. I shall then turn to their taxonomy of domesticates. In the second part of this chapter I shall discuss in detail the attributes (or elements of a key) which the Maenge use to identify the cultivars of two of their most important cultigens: taro and cordyline.

Maenge taxonomy of plants

6The Maenge do not have a word for plant (or indeed, for animal). The word ura, which refers to the area covered with vegetation, by contrast with the word for village site, maga, derives from the proto-Austronesian form uta, probably meaning vegetation. As used in present-day speech, however, it does not convey the same meaning as the word plant. The Maenge divide the plant kingdom into superordinate categories: tree (vega) and vine (oalo). They lack any category of the same order that would correspond to herbaceous plants, a category that the Hanunóo for instance do possess (Conklin 1954: 90). The terminology for herbaceous plants varies considerably from village to village and even from informant to informant, a point which later investigations should elucidate. For instance, Kutaena from Malakuru village in Jacquinot Bay calls paili small herbaceous plants which appear in short fallow gardens, while Kaeliuna from Pomio calls valinglala both small and tall herbaceous plants which appear in short fallow gardens. Neither term can be applied to herbaceous plants which appear in another environment, while ligneous plants which appear in the same environment but are too small to be classified as trees can be called vega tuna, “tree son” or vega tuna e sina, “small tree son”. The division of species following a generation principle also applies to fish or sea mammals: the bigger species (shark, dolphin, porpoise, tortoise) are the grandmothers of the fish, lea suguname, while the Caranga is their mother. The same contrast mother (name)/son (tuna) is also used to express the different parts of a vine.

7With only two superordinate taxa, the Maenge leave aside plants such as bamboos, ferns, wild Zingiberaceae etc. These are included in categories higher than that of species, but are not however co-ordinate with categories like tree or vine. With three superordinate categories, the Hanunoo were faced with the same problem (Conklin 1954). Most labelled taxa, in the Maenge language, correspond to the level of species of Western taxonomy, at least with respect to wild species. The case of the cultivated ones will be discussed below. In a few instances, one finds intermediate taxa that may correspond to the levels of family, genus or species. In this last case, the terminal taxa would correspond to the level of variety. For example, several Myristica spp., Myristica hollrungii and two unidentified Myristica spp. are called tagataga, together with another member of the Myristicaceae, possibly a Horsfieldia sp. An attribute added to the first name indicates which tolova one is talking about. The fact that there is a limited number of these intermediate labelled categories does not mean that the Maenge do not recognize the similarities between plants belonging to the same group, a point that has been clearly demonstrated for the Tzeltal Maya by Berlin, Breedlove and Raven (1968). These similarities are simply not expressed within the taxonomic classification. Thus, while there is no shared name to label two different species belonging to the genus Barringtonia (Barringtonia niedenzuana and Barringtonia asiatica), all informants can say that these species are akin to one another and have more in common between themselves than with any species outside this genus. The same applies to the various Ficus spp., Euodia spp., Pipturus spp. and Boerlagiodendron spp. among others.

  • 17 (Editor’s note). Later published as: Berlin, B., Breedlove, D., & Raven, P. (1973). General Princi (...)

8With usually no more than three levels, Maenge taxonomy of wild species falls into one of the evolutionary steps described by Berlin in his attempt to reconstruct the possible evolution of the vocabulary related to plant taxonomy (Berlin, 1969 n.d.)17. One cannot follow Berlin, however, when he tries to argue that the starting point in the evolution of this part of the lexicon must have been the genus. In the light of empirical evidence, as well as on grounds of logic, this position seems to be erroneous. When they are hunting, or collecting wild species, humans need to identify those individuals which present the same characteristics: they will look for Barringtonia niedenzuana, to return to our earlier example, when wanting to feed themselves, and not to Barringtonia asiatica, to which they will turn when needing a fish-poison. In no case, it seems, will humans look merely for a Barringtonia. It is the species that concerns humans and which they perceive as having the most striking characteristics (on this point, see Lévi-Strauss 1966[1962]). Moreover, it seems that while investigating nature and classifying it, humans must have gone from the less general to the more general so that they would have built the category of genus after having built that of species. This does not mean that all the necessary steps in the evolution of the lexicon should be one-way oriented. On the contrary, it seems quite normal that the superordinate categories could be introduced before some of the categories belonging to lower levels in the taxonomic scale; such a dialectical process is characteristic of the way the human mind works. Yet while it does not seem necessary to have built the category genus to build categories corresponding to tree or vine, clearly one needs the concept of species to build the more elaborate category of genus, which is only a bundle of species sharing identical attributes, just as one needs the concept of individual to build up the category of species.

Taxonomy of cultigens

9The Maenge call all edible plants kaninga, a noun derived from the verb kania, to eat, and which can be glossed as “vegetable food” in contrast with posi, which refers to protein food, usually game. Edible shells are called posi nga vai (posi of the sand), and fish can be called posi nga pelau (posi of the sea), in addition to their usual name lea, fish. These names indicate that vegetable food, mostly coming from root crops, constitute the most important part of the diet, while protein food is a delicacy. Kaninga, with no other qualification, usually refers to edible plants. Famine food is called kaninga ora urapuna (vegetable food of the forest), or pangaura (found in the forest), or more specifically, kaninga a pitolo (famine food). Trees with edible food, whether wild or cultivated, can be called vegame onra kaniriame (trees the fruit of which is eaten) while the noun kaoluma refers only to cultivates.

10The kaninga group, understood as referring to edible plants only, is divided into taxa that correspond to the level of our species. These may in turn be sub-divided in sub-groups. For instance, the taxa corresponding to taro, mao, is divided into two sub-groups, the “taro of the fire” (I.2) and the “rubbish taro” (I.1). Nonedible cultigens that do not bear a name which could contrast with kaninga, are classified following the same principles. For example, the cordyline, ela, is also divided into two sub-groups, the big (II.2) and the small (II.1) cordyline. As these cultigens play such an important role in Maenge life, it is necessary to examine in detail the characteristics of these sub-groups before proceeding further. The figures used here simply reflect distinctions that the Maenge make.

Taro (mao)

I.1. “Rubbish taro” (mao e kasau)

11Characteristics: number of cormlets above five; buds scattered on the tuber; petioles under five feet when reaching maximum height, petioles often dropping, petioles often green; leaves measuring around one foot in length; flesh of the corm soft, maturing quickly (some in four or five months), some causing skin irritation. With regards to quick-maturing varieties of taro, see Burkill (1935: 645) and Whitney (1939: 21-70). The word kasau usually serves to describe a “rubbish man” in contrast to a “big man”.

I.2. “Taro of the fire” (mao a poi)

12Characteristics: number of cormlets under five or cormlets absent; buds arranged in concentric rings on the upper part of the corm; petioles above five feet when reaching maximum height, petioles erect, often with bright colours; leaves longer than one foot; flesh of the corm hard, maturing slowly; not causing skin irritation.

13Taro belonging to the first category are planted in those parts of the garden where no fire has been made, the “ground nothing”. They are harvested first and not offered as gifts of raw food (gona e mata) since they rot quickly. Taro of the second category are planted where primary and secondary firing places have been made and around tree stumps, hence their name. They are harvested last and given in ceremonial exchanges as gifts of raw food since they can remain unspoilt for several weeks after being harvested.

Cordyline (ela)

II.1. “small cordyline” (ela e sina)

14Characteristics: stem under six feet, diameter around one inch, leaves under 16 inches in length and 2.4 inches in breadth.

II.2. “big cordyline” (ela e volau)

15Characteristics: stem above six feet, diameter over one inch, leaves longer than 18 inches and usually 4 inches in breadth.

16Cordylines of the first category are usually planted at the limits of the garden and alongside the divisions which run across a Maenge garden. Cordylines of the second category are planted in the main and secondary firing places and around tree stumps. As will be shown later, the two categories have different utilisations.

17Each of the sub-groups is subdivided into taxa which correspond to the level of cultivar. These taxa are usually terminal, but in some cases, they are further divided and this is reflected in the nomenclature. Taxa of the order corresponding to cultivars usually have simple names, but in the few cases discussed below, they have compound names (simple names plus attributes); the simple name is identical for two or more cultivars while only the attribute varies (Figure 6 below). Cultivars can be referred to by the word mata, that can be glossed as “kind”. This word is used to refer to species as well as to varieties: one can speak of vegamataname, kinds of tree; maomataname, kinds of taro, and namvumataname, kinds of the namvu cordyline.

18Throughout the Maenge area as well as among the neighbouring tribes, the gardeners grow three cordyline cultivars that bear the same name in Maenge: namvu. They only differ in their colours but are similar in all other respects. Namvu e kavauvaunga is “white” (kavavaunga), in fact light green in colour. Namvu e tente is “red” (tente). The third cordyline is light green with conspicuous red stripes and is called namvu e sisikaunini after sisikaunini, the striped chambered Nautilus shell. The Kol grow a cordyline identical to the Maenge namvu e tente, which they call “nomvule”, and another identical to namvu e sisikaunini, which they call “nomvukaia”. The Tomove equivalents are “nomvung” and “nomvungkaia”. The part which these plants play in magic will be discussed later.

19While the number of taxa that are further sub-divided in this way is not high (eight for taro, six for cordyline), these taxa present an interesting problem. Quite often, one of these cultivars is referred to as the “true” (manangana) one. Berlin has shown how universal this way of using the attribute “true” or “original” is (1969 n.d.). In the case of cultivated plants, the attribute may point to a priority in origin, the other cultivars having appeared after the one which is called the “true” one. The attributes of the other cultivars in the same sub-group refer to differences in colour or size. Thus, the palakanra manangana cordyline contrasts with palakanra marea, from marea, a parrot with a red plumage; they are similar in all respects except that the latter has more red stripes than the former.

20Another cordyline, motepa manangana, contrasts with motepa kavauna, from kavau, ash, referring to its whitish colour. The latter has a “white” surface while the former has large red stripes.

21It is rare to find more than two or three cultivars bearing the same initial name and usually these taxa belong to the same level in the taxonomy. In one case, however, that of the “kova” cordyline, there are some further subdivisions. In Malakuru there are two “kova”, one called kova e tente, the “red kova” and the other kova e ngango, the “yellow kova”. The latter has a yellow upper blade, the former a red one. They are otherwise similar and characterized by a large red purple margin from which their name is derived: kova is the name of the blood-thirsty spirits of the dead. In Matong, the people grow two distinct cultivars of kova, both belonging to the “yellow” sub-group. They differ in size and the smaller one is called the “female kova” (kova avale) while the larger one is called the “male kova” (kova apanung). This further subdivision however is not reflected in the nomenclature. One does not find a name such as kova e ngango avale or kova e ngango apanung, with a succession of two attributes and most probably the inhabitants of Matong are not aware that a “red” variety of kova is grown in Malakuru.

Figure 6. Taxonomy of taro and cordyline

Figure 6. Taxonomy of taro and cordyline

Taxonomy of tree crops

22The main tree crops, the kaoluma, are divided into sub-groups following the same principles, which apply in the classification of the cultigens. In the simplest cases, one finds only one level below that of the species. For instance, Barringtonia niedenzuana (paua), is divided into two terminal taxa: paua a toto, with red fruits (toto means blood) and paua a kera, with green fruits (kerakerang means green). Quite often it is the colour, the size, or some other prominent characteristic of the fruit (texture, colour of the seeds etc.), which is used as the basis for classification. For example, the Maenge distinguish between two kinds of mango trees: kinga a kula, the fruit of which has numerous fibres looking like roots (kula) and kinga a gopu, the fruit of which has few fibres and is almost as soft as the cucumber (gopu). The cultivated Areca palms are divided into two sub-groups based on the colour of the seed that is “white” in one group and “black” on the other. In some villages, the people also grow an Areca palm with “yellow” nuts, while the previous ones have “green” nuts and a “red” spathe, contrasting with the green spathe of the other types. The wild Areca palms are classified in three groups based on the colour of the nuts, which can be “green” or “yellow”. In the case of the most important tree crops, the coconut, the Canarium and the breadfruit tree, classification is based on several dimensions. The Maenge divide coconut trees into three categories: those bearing “red” nuts (lamasi a toto), “green” nuts (lamasi a kera) and “yellow” nuts. These are further sub-divided, based on the size of the nuts.

Maenge keys

23To illustrate the keys used by the Maenge to identify domesticates, I have taken the case of taro and cordyline, since they are the cultigens represented by the highest number of cultivars. The coastal Maenge plant at least 130 taro cultivars and the Longueinga over 200. The coastal Maenge plant at least 50 cordyline cultivars. Rather than present the keys themselves, I have chosen to present the attributes which are used to identify the cultivars. These can be listed in a few pages, while the keys would require more space. The list of attributes suffices to show both the minute character of Maenge observation and the fact that their keys are not essentially different from those built by Western scientists. Before proceeding to the actual data, it is necessary to say a few words about the perceptual attributes, which constitute the elements of the keys and the procedure used in the field to collect the relevant information.

Perceptual attributes

24In his paper on the “Diagnosis of Disease among the Subanum of Mindanao” (1961: 122-123), Frake sums up conveniently the distinction to be made between various ways of categorising. In his componential analysis of kin types, the investigator resorts to an analytical derivation of meanings. This procedure supposes an independent way of coding recorded instances of a category; this produces “distinctive features”, which Frake differentiates between “cues” and “criteria”. He describes as “cues” the perceptual attributes relevant to categorisation, such as the characteristics of plants, while he calls “criteria” a culture’s explicit definition of categories. Referring to categories applying to supernatural phenomena, he notes that a deity or a goblin can only be what his informants tell him it is. Here, I am concerned only with the analysis of perceptual attributes or cues. A cultivar name labels a class of cultivars. The task of the investigator is then to find out what are the perceptual attributes used by informants to name a specific instance of a cultivar.


25In the same paper, Frake remarks: “discovering cues in ethnographic settings requires yet largely unformulated procedures of perceptual testing that do not replace the culturally relevant stimuli with artificial laboratory stimuli” (1961: 123). Since this article was written, methods have been designed to try and eliminate errors due to the ethnographer’s cultural bias. The most interesting are probably those methods worked out by Berlin, Breedlove and Raven (1968), but they have been designed for literate societies. Working in an illiterate group, I have been using the method described by Conklin in his classic study of the Hanunóo (1954). I have asked informants to compare cultivars and describe their characteristic features. In gardens, I had to work mostly with single informants, as it is not customary for gardeners to walk into each other’s gardens, but whenever possible – and this was particularly easy for the cordylines – specimens were brought back to the village to be examined and compared by several other informants. Depending on circumstances (whether it was still planted or unrooted, with or without its petioles and leaves) or on individual characteristics of the plant, the identification could be done by beginning at a wide range of levels. The following list of attributes cannot thus be taken to represent any preferential order. The attributes are listed based on a three or two-part division of the plant that is common among the Maenge. The descriptive vocabulary of a cordyline leaf is given in Figure 7. (See Figure 5 in chapter 2 for descriptive vocabulary of taro.) Drawings illustrate some of the main contrasts between taro and cordyline cultivars in Figures 8 to 16.

Figure 7. Descriptive vocabulary of a cordyline leaf

Figure 7. Descriptive vocabulary of a cordyline leaf

26Before presenting the relevant data, one must say a few words about Maenge colour categories, as colour is one of the most important dimensions when discussing the specific attributes of taro and cordyline cultivars. Maenge colour terminology comprises basically five terms, thus belonging to stage IV in the evolutionary sequence proposed by Berlin and Kay (1969). These terms are: “white” (kavauvaunga), “black” (seisenga or kakaso), “red” (tente), “yellow” (ngango), and “green” or “blue” (kekaregang). “White” and “black” indicate value. Besides referring to white, “white” includes other colours such as grey, pink, light green and light yellow. Besides black and brown, “black” includes dark colours such as dark green or dark purple. “Yellow”, “red” and “green” refer respectively to bright yellow, red and green. All these colours can be modified by using the adverbs isura (slightly) and raumana (very). Other terms are borrowed from paints, man-made objects or natural objects and phenomena.


27I shall follow the Maenge habit of examining separately the main parts of the plant: leaf, petiole and corm.


28Size. It can be small or big (sina/volau). Taro of the group I.2 have large leaves but among them, some, like the cultivar opemagope, are conspicuous for their very large leaves.

29Shape. The leaf can be either round or elongated (taliliu/gavili). The round leaves correspond to ovate leaves, elongated ones to sagittate leaves. This contrast is sometimes expressed with the words moro (short) and gavili (elongated), that indicate differences in length, as shown below.

30Surface aspect. The Maenge contrast leaves which are flat (karepapa) with those which have a cup-shape or curved margins (tonrolongkala, concave) and with those that have margins dropping towards the ground (savalu, convex). Characteristic of taro with cup-shaped leaves are the cultivars kaluva and kerengkea, which can retain water in their leaves.

31Position of the leaf. The leaf can be erect (meisi, to stand), horizontal (keno, to sleep) or drooping (pu, to fall).

32Position of the point of insertion. In Maenge, the point of insertion is called either sipona (navel) or matana (eye). The Maenge only grow one non-peltate leaf cultivar, called mataupatalea, from matau (to be afraid) and patalea (to stand in a public place), a name recalling its appearance.

33Colour. When discussing the colour of a leaf, the Maenge consider the prevailing colour of the leaf as well as the colour patterns, the colour of the point of insertion and those of the various veins.

  • Prevailing colour. Some leaves are “white”, light green in colour. Others are “green”, others “black” (dark green).
  • Colour patterns. Some cultivars have distinct patterns either on the upper face of the blade or on the lower one. For instance, the cultivar gauveraguna, (gauve, dog and raguna, face) presents a large white spot on the upper surface of its leaves, evocative of a dog’s face. On another cultivar, malangana, one finds a white design which looks like a fish bone on each side of the mid-rib.
  • Colour of the veins. The veins of many cultivars are coloured. Colour can occur independently on the marginal vein or on the mid-rib. It can also appear on the primary veins or on the V-shaped veins of the lobes.

34Marginal vein. Some cultivars have a conspicuous marginal vein (“black”, dark purple, in colour). This is the case of the cordyline cultivar named kova, not only because the marginal vein but the whole system of veins, the point of insertion of the petiole, the petiole itself and the corm are of the same dark purple reminiscent of the colour of blood.

35Mid-rib. In some cultivars, the mid-rib is coloured. It is either “black” (purple) or “red” on the lower surface of the leaf. This coloration can appear by itself or be accompanied by the two following combinations: first with primary veins, in which case the whole system of primary veins is of the same colour as the mid-rib; second, with V-shaped veins on the lobes, in which case only the veins on the lobes (one for each lobe) are coloured rather than all the primary veins.


36Length. The petiole can be short or elongated (moro/gavili). Taro in group I.2 have elongated petioles. Some taro in group I.1 have unusually short petioles, for instance the cultivar kamoromoro (from kamana, petiole and moro, short) is named after this characteristic.

37Position of the petiole. It can be erect (meisi, to stand), drooping (pu, to fall) or even, in some cases, lying on the ground (keno, to sleep). The cultivar muelunroga, grown by the Longueinga, has a stem which grows in an oblique position so that it gives the impression of lying on the ground.

38Colour. When discussing the colour of the petiole, one must consider the prevailing colour, its distribution and the colour of the margins of the sinus.

  • Prevailing colour. The petiole can be “green”, “red”, “white” (either light green or ivory white), “black” (dark purple, dark green or black) or “striped” (magelegele). This last word refers to the “marks” or signs (panga) of various colours to be found on the petioles. They are often arranged in parallel stripes either of the same colour on a uniform background or of different colours. Since the colour of the stem is of special importance in the eyes of the Maenge, the cultivar’s name is often derived from it. For instance, gaiva has a petiole of the colour of the Malay apple (gaiva, Eugenia malaccensis); paua has a petiole of the dark purple colour of the fruit of paua (Barringtonia niedenzuana). Kasona, a cultivar with a black petiole, derives its name from the root of the term kakaso (black), while kerangana, a cultivar with a green petiole, derives its name from kekaregang (green). Of course, the colour categories mentioned above are very broad, and when discussing the nuances of the cultivars belonging to the same category, the Maenge will resort to various lexical devices.
  • Distribution of colour. The colour can be evenly distributed on the petiole or found mostly at the base of the petiole (puna) or at the apex (kuna).
  • Coloration of the margins of the sinus. Some cultivars are identified by the presence of coloured sinus margins. These can be “black”, (dark purple) “red” or “white” (light green or white).


Figure 8. A round taro corm

Figure 8. A round taro corm

Figure 9. An elongated taro corm

Figure 9. An elongated taro corm

39Cormlets. Some cultivars such as taro belonging to group I.1 produce numerous cormlets, while others such as taro in group I.2 produce very few or none. The cultivars producing cormlets are described as females (avale) while those producing none are called males (apanung).

40Distribution of buds. In taro of the group I.2, buds are arranged in concentric circles on the upper part of the tuber, while in cultivars of group I.1 they are scattered on the corms with no obvious pattern.

41Shape. The tuber can be of round or elongated shape (taliliu/gavili). This contrast can also be expressed by the opposition short/elongated (moro/gavili).

42Branched corms. Some cultivars (for example mao a vesangana, from vesangana, fork) produce branched tubers, while others never do.

43Colour of the flesh. It can be “red”, “yellow”, “black” (dark purple), “white”, “multicoloured” (konko) or “dazzling” (laulauga). Konko applies to cultivars with a white flesh and conspicuous black fibres (laulauga, fibres), and to cultivars with white flesh and conspicuous “yellow” fibres. The colour of the flesh has sometimes influenced the choice of the name as in the case of lonango (from lona, inside and ngango, yellow).

44Texture of the flesh. It can be hard (ginging) or soft (same). Taro in group I.2 usually have a hard flesh.

45Colour of the skin. It can be “red”, “white”, “yellow”, “striped”, “multicoloured” or “dazzling”.

46Skin Irritating properties. A few cultivars are more irritating (masikili) than others. They all belong to group I.1 and many of them are quick-maturing. They often bear names indicating their poor quality: one is called pangaura, the usual name for famine food; another is momolisi, the name of a wild Dioscorea sp. collected during famines; another is sese, the name of a wild Colocasia sp.

Figure 10. A round taro leaf

Figure 10. A round taro leaf

Figure 11. A sagittate taro leaf

Figure 11. A sagittate taro leaf

Figure 12. A cup-shaped taro leaf

Figure 12. A cup-shaped taro leaf


  • 18 (Editor’s note). The reader may recall that not all cordylines are cultivates.

47To distinguish between cordyline cultivars18, the Maenge consider some characteristics of the stem and various features of the leaf.


48Length. It can be short or elongated (moro/gavili). Cordylines of the group II.1 have short stems while cordylines of the group II.2 have elongated ones.

49Diameter. The stem can be small or big (sina/volau). Cordylines of group II.1 have small stems while cordylines of group II.2 have bigger ones.


50Size. It can be small or large (sina/volau). To decide whether a cordyline leaf is to be classified as small or large, length as well as breadth are considered. The leaf of the cultivar raga, though longer than the average leaves of plants in group II.1, is nevertheless considered to be a small leaf because of its narrow aspect. Cordylines of group I.1 have small leaves, while cordylines of group I.2 have large leaves.

51Length. Leaves can be short or elongated (moro/gavili). This contrast is used to express relative differences of length inside groups II.1 and II.2. Here length is considered as an independent variable. One is then allowed to contrast the cultivars vena and lota as short, with guruguru and patalea as elongated, although these four cultivars all belong to the group of cordylines with large leaves. One can contrast in the same way two cultivars with a small leaf: one, pepe, has a short leaf while the other, motepa, has an elongated one. This difference in length results in a difference in shape, long tapering leaves contrasting with shorter and rounder ones.

52Surface aspect. The leaf can be flat (karepapa), curved (tonrolongkala) or twisted (lulukala). In this context, curved means that the margins tend to roll up. In three varieties that are very slender (less than one inch in breadth), the margins overlap with one another and the leaf is twisted (lulukala) at the same time. In flat leaves, the tip of the leaf may be rolled up, but this is not taken into consideration.

53Colour. When discussing colour, one must consider a) the prevailing colour; b) the presence of patches and c) the colour of the marginal vein.

  • Prevailing colour. The leaf can be “white” (light green), “black” (dark green or dark purple), “red”, “yellow” or “striped”; only the wild cordyline kerakera is “green”. To classify a cultivar in one of these categories, the Maenge consider only mature leaves and note the predominant colour on the upper blade. The presence of stripes on the upper blade does not necessarily mean that the leaf is described as “striped”. For this, the stripes must be large and conspicuous. If stripes of one colour are numerous and large, then the cultivar will not be considered as “striped” but classified using the predominant colour.
  • Patches. It often happens that even when a leaf is mature, colour is not evenly distributed on the upper blade. There can be patches of different colours and these are used for discriminating purposes. These patches can be either at the base, at the tip or at the centre of the leaf. They can occur near the mid-rib or near the margins. For instance, the cultivar lote is easily identified among other “red” cordylines because of a “yellow” patch that remains on the leaf, around the mid-rib, towards the tip. Not only there are minute differences used as cues but they are also called into play when composing costumes.
  • Marginal vein. Some cultivars, like kova, have a very large and conspicuous marginal vein, and this is a sufficient characteristic to identify them readily. In others, a very thin marginal vein can be the only cue allowing discrimination between two cultivars. The cultivars kaokina and lolongotulu are identical in all respects but kaokina has a “red” marginal vein while lolongotulu has none. The marginal vein can be “red” or “black” (dark purple); it can be thin or wide (sina/volau).

Figure 13. pepe, a cordyline cultivar belonging to Group II.1 with short leaves

Figure 13. pepe, a cordyline cultivar belonging to Group II.1 with short leaves

Figure 14. vena, a cordyline cultivar belonging to Group II.1 with long leaves

Figure 14. vena, a cordyline cultivar belonging to Group II.1 with long leaves

Figure 15. namvu, a cordyline cultivar belonging to Group II.2 with medium-size leaves

Figure 15. namvu, a cordyline cultivar belonging to Group II.2 with medium-size leaves

Figure 16. guruguru, a cordyline cultivar belonging to Group II.2 with long leaves

Figure 16. guruguru, a cordyline cultivar belonging to Group II.2 with long leaves

54All the attributes listed above are cues: in specific contexts, they can be used to discriminate between cultivars. I hope this list has shown the detailed character of Maenge observation as well as the fact that it does not differ essentially from our own. Maenge data stand very well the comparison with the key that the Western botanist Whitney (1939) elaborated when studying Hawaiian taro cultivars. The attributes used in both cases are nearly identical, the main exception being the Maenge’s lack of interest in leaf texture. While the Maenge know that the colours of both taro and cordyline flowers can vary, they do not take these variations into consideration when identifying cultivars. The reason for this is obvious, since cultivars must be identified daily, throughout the year, but flowering occurs only during a limited period. The key must use characters that are easily observable. Again, since they never dig out the rhizomes of cordylines, their key rests only on the aerial parts of the plant. These limitations are inherent in all systems; one has only to remember how the study of cultivated plants has gradually disappeared from the sphere of interest of Western taxonomists during the last centuries because of the material difficulties presented by the storage of specimens (Anderson 1952: 30-48) to appreciate Maenge performance in this field.

Systems of appellation

55A plant appears to be fully domesticated only when it has received a name. A plant identified as a domesticate must be endowed with a soul and a name, a process which should be understood as an effort to humanise the natural world. More will be said on the soul of domesticates, especially of cultigens, in Chapter 5. Only what concerns naming directly is noted in this section.

56Among the Longueinga, it is said that when guided by the guardian spirit of his clan, a man discovers a new variety of taro, banana, sugar cane or pitpit that can be domesticated, he takes it back to the village. During the next night, he is visited in his dreams by the guardian spirit who reveals to him the name of this variety. To secure the plant is not enough; one must know its identity. Similarly, a Tomoive tale recounts how a man stole a banana plant from the garden of the spirits of the dead, located on the slopes of the Father. He plants it in his own garden but he is quite unhappy not to know its name. On the following night, he goes back to the volcano and waits until the woman spirit who had owned the banana plant comes. She discovers the sad fact and laments over the loss of her banana. As she cries, she utters its name, kagirale. The man then goes away, rejoicing to have overheard it.

Cultivar names

57Many cultivar names are no longer understood by present day Maenge. For others, which have an obvious meaning like keno (it sleeps), lolo (stone) kaonamena (saliva), it was not possible to collect information on the reason why these names were given to varieties. In many cases, however, the information is available. Among the classes of names which are regularly given to cultivars, we find place of origin, plant and animal species, and man-made objects. The last classes are of interest as they are given to underline a characteristic feature of the plant. This type of name giving is frequently used by Western as well as by primitive taxonomists (Dournes 1968: 143-145).

58Out of the 127 names of taro cultivars collected in Malakuru, 13 (10%) are names of plant species, 2 (1.65%) of animal species; 3 (2.3%) of man-made objects and 3 of places. Out of the 57 cordyline cultivars grown in the same village, 2 (3.5%) bear names of plant species, 5 (8.7%) of animal species, 2 (3.5%) of man-made objects and 1 (1.7%) of a place of origin. Out of the 142 taro names collected among the Longueinga, 14 (5.7%) bear plant names, 4 (1.6%) bear names of animal species, 1 (0.4%) bears the name of a man-made object and 2 (0.8%) names of places of origin. Out of their 32 cordyline cultivars, 5 (15%) bear names of plant species, 1 (3%) of animal species and 1 (3%) of a place of origin. The same type of name giving can be observed with the other cultigens. Among the 36 names of banana cultivars collected in Malakuru, 2 (5.5%) bear names of plant species; 3 (8.3%) of animal species, 2 of man-made objects and 1 (2.7%) of a place of origin. Among the 37 cultivars of sugar cane, 4 (10.8%) bear names of plant species. These names emphasize a feature of the cultivar. The taro cultivar gaiva (Eugenia malaccensis) is called by this name because its petiole is of the same pink colour as the Malay apple; the taro cultivar paua has a petiole of the same purple colour as the fruit of the paua tree (Barringtonia niedenzuana). A cordyline cultivar is called nagutu (Ficus itoana) because its leaves are of the same colour as the fruits of this fig tree; another is called magu because its leaves are of the same colour as the stems of Diplocaulobium sp. (magu). Among the Kol, one finds a cordyline cultivar with a yellow leaf called kokona after Parartocarpus venenosus, a wild breadfruit tree that produces a yellow breadfruit, while the name of this species, ivo, has been given by the Maenge to a sweet potato with a yellow flesh. Animal and man-made objects reflect the same tendency. A cordyline with red leaves is named roga after the land crab; a taro with a white flesh is named kapulu (lime), etc.

  • 19 (Editor’s note.) A venomous species of sea snail, a marine gastropod mollusc in the family Conidae
  • 20 Nautilus pompilius.

59In the Linnean classification, the use of names borrowed from plants, animals or man-made objects for the second term of the binomial is frequent (to quote only Conus textile19). The similarities with the Maenge system are obvious if one underlines the fact that the common name gaiva for a taro cultivar is but an abbreviation of the longer mao e gaiva, the taro named after Eugenia malaccensis. The same remark applies for binomial expressions such as namvu e sisikaunini, the cordyline namvu named after the chambered Nautilus sp.20 Here again an animal name is used in the second term of the binomial as a way of differentiating this namvu cultivar from the two others. With respect to primitive classifications, one can quote for the Pacific area the varieties of Hawaiian taro described by Whitney (1939: 48-53). One variety is called “makoko”, (a reddish fish); another one “manini”, (a small striped fish); yet another one “kumu”, (a bright red fish) etc. In Iruk, one finds an “octopus Colocasia” and another Colocasia sp. is called “return to clam” as its leaves look like clams. In Ponape, a Cyrtosperma sp. is named “lizard” (Mahony, 1960 p. 14 and 125).

60With respect to the Maenge taxonomy, the number of names derived from the plant or the animal kingdoms would be higher if one considered not only the names derived from terminal taxa (such as gaiva, paua etc.) but all those which refer to taxa of a higher order (such as frog, snake etc.) as well as those which are derived from the parts of plants or animals. For instance, a banana cultivar with its fruits turned upwards is called manuina (bird’s bill). Other names may be descriptive as well. Thus, a banana cultivar with whitish leaves is called kavau (ashes), while the name kova referring to the blood-thirsty spirits of the dead is given to specific banana, taro and cordyline cultivars because of the presence of dark purple on their petiole and/or leaves. Instead of pointing to an anatomical characteristic of the plant, its name may refer to other characteristics such as the time it takes to mature, the size of the harvest and the way it was dispersed. For example, a banana cultivar is called parimao (maturing before taro), because although it is planted at the same time as the taro cultivars, it is ripe after six months. Another banana cultivar is called luvolua (two luvo). The luvo is the rachis of the Archontophoenix sp., which the Maenge use as a tray. One needs not one luvo as usual, but two, to carry the yield of this banana cultivar as it is so abundant. Another taro cultivar is named manutaena from manu and taena (faeces), because it was found in the forest and supposed to have been brought there by a bird. Yet another taro cultivar, dispersed in the same way is called pige (Eucalyptus deglupta), because it was found near a pige tree. A sweet potato cultivar is called manupilia, from manu and pilia (to defecate), for the same reason that the taro cultivar mentioned above is called manutaena.

Exchange of names

61We have seen that a small but significant percentage of cultivar names are names of plant or animal species. If one now turns to the names given to human beings, one finds that among these, a small percentage is represented by the names of cultivars. Maenge personal names are infinite in number. They are made up during the days following childbirth, either by the parents or some close relative, and commemorate a specific action or event that has already taken place or is contemplated at the time of the birth. Personal names thus form a kind of chronicle, a record of the most important events that take place within the village. They also serve as a kind of memento when they remind their bearer or his or her kin of, for instance, some revenge. Except for a few instances, personal names tend to be unique in genealogical records. The main exceptions are due to the tendency to give a child the name of a maternal grandfather, and the relationship through which a big-man can bestow his personal name to a child who is not related to him, so that his name should not perish with him.

62Personal names are usually compound names and often include a verbal form. A few Maenge personal names are derived from the natural world. They come mostly from the flora and usually they are names of cultivated plants. There is no example of fish names and only one bird name was recorded. This does not mean that bird species cannot be used as a way of classifying: they are used as totems and play an important part in the mythology, together with sea-mammals or fish. The fact that these names are found at the sociological level of the division of the Maenge into clans, and not at the level of individuals, probably points to a greater identification between humans and plants – mostly cultivated plants – than between humans and other parts of the natural kingdom.

63The personal names that are derived from wild plant species are records of specific events. If a man or a child falls from a tree, he may be named after that tree. Thus, one finds names such as Paugala (Albizia falcataria) or Vavalupuna (base of Gmelina moluccana). Usually this class of names is constructed as in the following examples: Pungataua: “he has fallen from a taua” (Pometia pinnata); Pungapaua, “he has fallen from a paua” etc. These personal names however are nicknames which are given at any time during one’s life rather than at birth. Although there is no regular pattern for changes of names, it is not unusual to find a Maenge with several personal names given to him or her at different moments of his or her life. The first one may be entirely forgotten except by the oldest people. Names of cultivars are also given to commemorate specific events. When a woman has been successful in growing a variety of cordyline, sugar cane, banana or taro, she may give the name of this variety to her child at birth. Later in life, the child should make an offering of this specific variety to his mother. The taro and cordyline cultivars the names of which are given to children always belong to the best varieties, in groups I.2 and II.2. One should note that this custom is found only in the western part of the Maenge country and that, though a few instances of cordyline, sugar cane and banana cultivars’s names are known, most of the names are names of taro cultivars. One finds names like Isingana, Ngarole, Laviauna, Levolong, Songetuna, Galipulia, Vagoragili etc. In the genealogical records, 54 names of taro were found, 7 of cordyline, 6 of banana, 6 of sugar cane and 1 of pitpit.

64The symbolic meaning of this type of name-giving will be analysed later. Briefly, one can say that this act can be interpreted as the wish that the child be as strong and fecund as this cultivar. This type of name-giving is possible only because of the similarities between the taro soul and the soul of human beings that the Maenge assume. What interests us at this stage is the logical process which is involved here. As seen in the previous pages, names given to cultivars could be names of plant or animal species. Here, the names given to individuals are the names of cultivars. As one proceeds from relatively high levels of generality (species, cultivar) to the level of individuals, names at one level are borrowed from the level immediately above in the following scheme:

65The technique emphasizes the discrimination either between varieties or between individuals.

Cultivates as individuals

66While humans borrow their names from domesticates, mainly from cultivars, they also bestow their own names on domesticates, mainly on cultivates. Most cultivates in a village territory, especially in the western part of the Maenge area, bear names which are personal names of human beings. When a child is born, it is customary to bestow his or her name on a coconut tree, Canarium sp., breadfruit tree or mango tree. The tree will belong to the child and will perpetuate the name of its owner after death. Given that the population has a very short genealogical memory (most individuals are only able to name their grandparents, and only one informant could give the name of an ancestress belonging to generation 5), these trees form a kind of genealogical record which helps to memorize the names of the ancestors of a given clan. On Figure 17 below are shown the members of the Lengelenge clan whose names were bestowed on coconut trees still to be found in the territory of Malakuru.

67As a proof that the Maenge tend to consider cultivates individually, one should note that individual cultivates tend to be described in terms of the characteristics contrasting with those of other individuals of the same species. It was mentioned above that the Maenge divide the coconut trees into three groups based on the colour of the nuts. These groups can be further subdivided based on the size of the nuts. When describing an individual coconut tree, however, people will not point only to those characteristics which make it possible to classify it into one of the groups or sub-groups mentioned. They will also insist on features such as the length of the stem, the way it has been bent by the wind, the shape of the nuts etc. All these features, which are not systematically considered and which vary among individuals, do not belong to the classificatory system as such but reveal the Maenge’s tendency to treat such trees as individuals, with individual characteristics, and not only as members of the same species, or of the same groups or sub-groups bearing the same uniform characteristics.

68In conclusion of this chapter, one should insist on the importance that the notion of cultivar has for the Maenge. This has been apparent when discussing the main features of a Maenge taro garden as well as their systems of classification and appellation. In the next chapters, when describing techniques such as cooking, the feminine costume and the main characteristics of Maenge magic and of the gardening rites, it will appear again and again that it is not the species as such which plays the most important part in the Maenge conceptual world but the taxa that come just below it, the cultivars.

Figure 17. Genealogy of Kelatape, the Mother of the village of Malakuru, showing members of her clan, the Lengelenge, whose names are still attached to coconuts growing on the village territory

Figure 17. Genealogy of Kelatape, the Mother of the village of Malakuru, showing members of her clan, the Lengelenge, whose names are still attached to coconuts growing on the village territory


16 (Publisher’s note). From here to the end of the chapter, we have added subtitles or clarified the already-existing ones in order to bring out the structure of the text and make it easier to follow.

17 (Editor’s note). Later published as: Berlin, B., Breedlove, D., & Raven, P. (1973). General Principles of Classification and Nomenclature in Folk Biology. American Anthropologist 75(1): 214-242.

18 (Editor’s note). The reader may recall that not all cordylines are cultivates.

19 (Editor’s note.) A venomous species of sea snail, a marine gastropod mollusc in the family Conidae.

20 Nautilus pompilius.

Table des illustrations

Titre Figure 6. Taxonomy of taro and cordyline
Fichier image/jpeg, 167k
Titre Figure 7. Descriptive vocabulary of a cordyline leaf
Fichier image/jpeg, 73k
Titre Figure 8. A round taro corm
Fichier image/jpeg, 125k
Titre Figure 9. An elongated taro corm
Fichier image/jpeg, 71k
Titre Figure 10. A round taro leaf
Fichier image/jpeg, 124k
Titre Figure 11. A sagittate taro leaf
Fichier image/jpeg, 78k
Titre Figure 12. A cup-shaped taro leaf
Fichier image/jpeg, 56k
Titre Figure 13. pepe, a cordyline cultivar belonging to Group II.1 with short leaves
Fichier image/jpeg, 82k
Titre Figure 14. vena, a cordyline cultivar belonging to Group II.1 with long leaves
Fichier image/jpeg, 85k
Titre Figure 15. namvu, a cordyline cultivar belonging to Group II.2 with medium-size leaves
Fichier image/jpeg, 73k
Titre Figure 16. guruguru, a cordyline cultivar belonging to Group II.2 with long leaves
Fichier image/jpeg, 77k
Fichier image/jpeg, 22k
Titre Figure 17. Genealogy of Kelatape, the Mother of the village of Malakuru, showing members of her clan, the Lengelenge, whose names are still attached to coconuts growing on the village territory
Fichier image/jpeg, 304k

© pacific-credo Publications, 2018

Conditions d’utilisation :


Volume papier