Volume 95 1986 > Volume 95, No. 2 > Completely by accident I discovered its meaning: the iconography of New Ireland malagan, by S. M. Albert, p 239-252
                                                                                             Previous | Next   

- 239

The first part of the title of this paper comes from Parkinson. His experience with attempts to pin down the meanings of the malagan (malanggan) 1 carvings of northern New Ireland (Papua New Guinea) is instructive. It shows us what kinds of problems stand in the way of a more adequate approach to the significance of the carvings.

Parkinson begins by mentioning the complexity of a particular turu, one of the many carved boards or “friezes” called by this general term (Parkinson 1907:647). Two birds are portrayed, each dipping its head into a hole in the centre of the frieze. A snake is coiled at one end. Evidently questioning people about the carving revealed little, but a myth or tale told in another context gave Parkinson what he thought was the key to its meaning. Hence his claim of accidental discovery. Parkinson found that the birds represented two pigeons who came to a water-hole to quench their thirst after a long flight. While they were drinking, a snake slithered their way to try and devour them; but they were able to see it through their legs and thus escape. This, according to Parkinson, is its significance. Thus, he concludes, “so long as we do not know these tales, so long must the significance of the carvings remain dark to us” (Parkinson 1907:651).

Certainly there is merit to Parkinson's claim. But even with the tale before us, one wants to ask, “Can that be its full significance? Do we really know so much more about the meaning of the carving even with the tale?” If this is the “meaning” of the carving, it is an unsatisfying one, for even with the tale we know little more about the carvings than before.

Parkinson's experience with malagan sculptures presents one classic interpretative finding, namely that indigenous tales, myths and cosmological beliefs simply do not explain the carvings. The tale associated with this turu carving itself requires explanation. Rather than take it as the meaning of the carving, it appears that we ought really to add it to what needs explaining. For the significance of the tale is unclear, nor is it apparent why the tale should merit a visual realisation. I - 240 conclude two things. First, the carvings are underdetermined by the myths or tales associated with them. These shed light on some aspects of the carvings but not all; and when they do present ideas that appear in the carvings, so that one can really say that a carving illustrates a theme from myth, much still remains unexplained. Second, if this is the case, I claim that we cannot expect to find a belief or myth that explains a carving. Or rather such tales or myths are as obscure as the carvings they are meant to explain. Both carving and belief are data for analysis, and these shed light on meanings or significance of another order. If I am right, not only did Parkinson fail to find the meaning of this turu, but also he had to fail, for this kind of meaning does not exist.

I return to these claims below. At this point it is best to tender the other classic interpretative finding associated with the sculptures. To put it simply, when we find a variety of information on the sculptures—say, names from different locales, myths, associated rites—and when our hopes for a fuller explanation are raised, we also find that our sources conflict. That is, two very different origin tales might be advanced to explain one named carving; or worse, two myths involving the same character as founder of the carving may themselves differ in radical ways. No doubt some of this conflict can be traced to variations in locale and temporal distance, according to when and where the information was gathered. But even this explains only so much.

One example here will speak for many. Thus, while turu is a general name for a type of frieze (Gunn 1982), the hole in Parkinson's turu is described as a water-pond, but Krämer presents evidence to show that the circular form actually represents a hearth. Krämer's Amba informants spoke of the circle figure in another carving as mbane ‘hearth’ (Krämer 1925:82). I shall not even consider Peekel's claim that it is neither, but rather a moon, for Peekel's ability to find moon imagery everywhere is hardly credible, as Bühler (1933) has shown. More disconcerting, since perhaps less tinted by analytical bias, is variation in the meanings given for the name of one type of turu, the walik malagan. Peekel, on the one hand, claims that the name means “two ancestors”, and he supports the claim with linguistic evidence and the form of walik, a board with fish heads of some sort at each end (Peekel 1927:17). Walden, on the other hand, traces walik back to a historical figure whose name means ‘proper man’ (Walden 1940:21). Not even the general names for the sculptures are consistent; Parkinson's turu must be laid next to Peekel's kata.

I mention turu, kata and walik here because these will receive some attention below. In the face of the conflicting information we have about them, one is tempted to claim that New Irelanders are unsure of the significance of the carvings, as Peekel did; or that the traditions are waning, as Lewis (1973) has done; or that one of our sources must be wrong; or that the whole effort of untangling the sculptures is simply not feasible. Now, no doubt some of the information we have on the sculptures is wrong, and no doubt the degree of knowledge concerning carvings must vary according to time and place. But this, I would claim, is less an obstacle to an understanding of the carvings than an indication of what kinds of meanings we ought to look for. No doubt different informants and different - 241 elicitation techniques gave rise to different data of varying quality. But the inadequacy of Parkinson's discovery and the variations between Krämer, Peekel and Walden show that particular referents for the forms that appear in the carvings, even if discoverable, will not exhaust the significance of the carvings. Given the type of meanings expressed in malagan, I claim that differences between informants should be expected. That malagan forms can have more than one referent is itself meaningful. And differences in the referents assigned to the circle form, as well as variations in the design of such a form, for example, may be as revealing as a more uniform exegesis. Thus, rather than look for single and unvarying referents, we must turn instead to a more sensitive treatment of the organisation of form in the carvings, to the sense of a carving or class, rather than to the referents of particular forms.

We have already seen something of the empirical justification for this approach. Let me turn now to its philosophical basis. I refer readers to the work of Wollheim (1980) and Gombrich (1960, 1963). These show that understanding an expressive form involves much more than knowing what the form represents. For the ability of a form to represent something depends upon a wide range of knowledge and behaviour that links artist, spectator, medium and meaning. As Wollheim has shown, this network of knowledge and behaviour need not be complete or fully shared, and probably is not; but some level of overlap, some degree of sharing, is essential for expression and, indeed, is presupposed in the concept of art.

I suggest that this basic condition for the expressivity of malagan is to be found in the organisation of form in the carvings, and not in some relation between particular forms and their referents. Hence my emphasis on “sense” Moreover; I suggest that the principles that govern this organisation of form, when specified, will give the minimal level of consensus in the interpretation, use and understanding of the carvings that allows them to express what they do. 2 Individuals may know more or less, claim different meanings, or see none at all; but all, I would claim, share elements of knowledge that allow the sculptures to carry meanings and give the forms a degree of system. The principles governing the organisation of form in the carvings give us this level of overlap or sharing active in a malagan community, for these principles are codified in the names of carvings and associated rules of use. 3 Thus, such principles can plausibly be called the social reality of the meanings expressed in the carvings. This would be the iconographic contribution of the carvings to the meaning of the whole ritual event.

This emphasis on incomplete knowledge and minimal consensus goes contrary to two recent and important ethnologic studies of art. Lewis-Williams' study (1981) of San rock art and Hanson's study (1983) of Maori carving claim that belief and the forms of art are homologous, so that one is completely expressed in the other. Art in this view is another way of stating what people believe. As Lewis-Williams puts it, art helps people see what they believe. Brouwer (1980) and Fergie (1983) make similar claims for malagan. But this approach faces the grave difficulty of explaining why such an oblique form as the art object should be - 242 chosen to express these beliefs. It also assumes a Durkheimian coherence, or fit, between art and social life, that often seems untrue to the actual social life we see. That is, if art is obviously expressive and obviously social, it is still not obvious why it should express social meanings in the same way as other institutions, such as kinship relations, clan organisation, or legal codes. For at the very least art faces constraints derived from its media; these limit what can be “said” (Boas 1951). It is also more restricted in the universe of its producers: very few New Irelanders carve malagan, for example. This may allow a more salient role for individual variation in carvings and meanings, and thus a less direct social structuring of the art. Also, while the Durkheimian claim of co-extensiveness between art and society is attractive, it quickly founders on the difficulty of isolating what aspects of society and what aspects of the art are the critical ones. The fit between the two is often forced or else simply made too easy as a result of selectivity on the part of the analyst. Skorupski (1976) and Gilbert Lewis (1980) have made a powerful critique of the Durkheimian “symbolist” position on just these grounds.

The approach taken here, by contrast, does not claim such a rigid social determination of art. It only claims that a minimal social consensus informs expression, which in turn allows a wide range of local and individual variations. Given this, the task becomes one of specifying this social consensus, evident, I claim, in the organisation of form in the carvings.

What, then, can be said about this organisation of form? Following Lewis-Williams, our first task must be “algebraic”, a specification of the range of forms that appear in the carvings. For this task, database programmes offer a good means of sorting through disparate information to reveal regularities in the arrangement of form, name, use, locale and material basis. These regularities give us an inkling of the regional and temporal organisation of the system, which may not be discernible within a single village perspective. The results of such a study are clearly critical to what Lewis-Williams calls the “semantic” component of the study of art. But I am currently only at the very beginning of this effort.

Rather than the big picture, then, for the remainder of this paper I shall concentrate on the friezes mentioned earlier. Walik appears to be a major class of carvings, a “big name” division in the Tabar terminology reported by Gunn (1982). Turu or kata appear to be terms describing the form of this type of malangan, the horizontal board with rounded ends. Kolepmu, for example, is another of this type, which can be usefully compared with walik. As general classes, walik and kolepmu have many subtypes. These too are named, though it is unclear how far they differ from the major type. Thus, some subtypes appear to have clear iconographic variations within the frame of the wider class, while others may only be named differently, perhaps to link the particular carving to some commemorated deceased person. An example of the former is the variation between one-hearth and two-hearth walik types; an example of the latter may be represented in the various names Powdermaker (1971 [1933]:316) lists for kolebmur (kolepmu), 4 or the variations in the representation of the human figures in these waliks, i.e., use of some personal feature (birthmark, hair style, etc.), scene - 243 (way of death), or name to link the image to a particular deceased person (see Krämer 1925:38, 81). Between two named classes, however, there are obvious iconographic variations. As walik and kolepmu are close in appearance, it is instructive to compare them for these differences.

Figures 1a and b are kulepmu malagans, while Figures 2a-e are waliks; all are taken from Krämer (1925:71, 76). Though both share a similar division of the decorative field, with a central circular figure, rounded ends, and an elaboratedly worked field between the two, walik and kolepmu are distinct in significant ways. The circular centre in kulepmu is truly a hole bored through the board; the circles in walik are worked out with an elaborate infilling. Kulepmu also lack the human figures within the board or attached to the top evident in walik. While kolepmu were used in particular rites in which men stuck their head through the hole or threw coconuts or spears through it, walik has no such tradition. Instead, walik is capable of commemorating groups of dead, all of whom appear as figures in the carving. These differences must be contrasted with a major iconographic similarity; carvings in both classes may show birds whose heads are pointed towards the central circle.

A glance at the two sets of figures shows that Parkinson's discovery probably involves a kolepmu type; Figure 1b fits his description of the carving fairly well, though the snake is not evident in Kramer's reproduction. Thus, it is quite possible for the general circular centre figure in turu or kata carvings to represent two different objects, water ponds in kolepmu and hearths in walik. While Parkinson indirectly discovered the referent of the circle, Krämer was given the term mbane or ‘hearth’ in Amba to describe the central circle, as mentioned earlier. Powder-maker offers confirming evidence for the latter: in Lesu the centre of walik is call-

a, b. Kulepmu malagans identified by Krämer (1925:71, 76). Figure 1b appears to be very close to the one identified by Parkinson, which prompted his discussion of meaning and representation in malagan.
- 244

ed balji, “hearth fire” (Powdermaker 1971 [1933]:132). Gunn found similar information in Tabar. Here the form itself is named (mateling) and denotes ‘fireplace, eyes of fire’ (Gunn 1982:79).

Now, given these data, we need to go beyond referents (which by themselves are not very revealing) to see what sense relations motivate the appearance of these forms. I see at least three profitable lines of inquiry, which I shall outline here. These are (1) to discover the internal motivation of the forms, that is, to show how the particular circle forms can be associated with fire in the one case, and water in the other, from other evidence in the carvings; (2) to analyse the ritual motivation of these forms from the uses of the carvings, i.e., the inserting of the head in kolepmu and the attaching of human figures in walik; and (3) to explore the further significance of the circle form in the two classes of carving by finding intermediate malagan classes that seem to bridge the two and in this way shed light on both. For the first task I shall concentrate on walik, for the second on kolepmu, and for the third on the auvere malagan mentioned by Walden (1940:28).

The internal motivation of the circle form of walik is not hard to establish. It is a reduced version of the form evident in the wowara malagan, which is called “sun”. A dance mouthpiece, here shown as Figure 3, shows a bird-like animal carrying such a form and immediately connects the piece to the myth in which the kingfisher steals fire, carries it on its back, and gives it to Moroa, the god associated with the sun (Heintze 1969:60ff.). Thus, it is not surprising to see the form elsewhere; it appears in the bodies of birds represented in many carvings. Figure 3 also shows an elliptical form protruding from the centre of the circle, where the four prongs meet. Inspection of Figure 2a-e shows that this form at the centre of the circle is an eye, and this impression is confirmed by additional data related to display of the carvings. Inserting the eyes in malagan sculptures gives the sculptures power and heat, so that one who fights when the sculptures are displayed angers the eye of the sculpture and risks its wrath (Walden 1940:13). Thus, eye and fire are indeed linked, both conceptually and iconographically.

I have by no means covered all the forms that appear in walik, and knowledge of these is probably critical in understanding why a hearth should be placed in the middle of these sculptures. This issue in turn raises a larger one. Given that the form is embedded in a set of meanings that link fire, sun, eye and power, why should it appear in the body of these fish, with birds dipping their beaks in it and men adjacent to it? When we pose this question we are really asking about the overall meaning of the carving. This question is fraught with difficulty, as I tried to show above. But evidence from a number of domains allows us to state with some confidence that walik is a scene, a picture of the transformation of the dead. The human figures are clearly pictures of commemorated dead. Krämer (1925:81) mentions many examples to show that the carvings “decidedly aim at an embodiment (Verkörperung) of the deceased”. The sharks, fish, birds and snakes are agents in the transformation of the dead. They carry the dead away, or form repositories of the dead, or battle over humanity depending on a set of myths collected from various areas of New Ireland. Thus, for example, kata, the term

- 245
a-e. Walik malagans (ualik) identified by Krämer (1925:plates 38, 39). These all appeared in a display house in Lesu, though some originated in neighbouring Amba. a-c show one central mbane or ‘hearth’ form; d and e show two such forms. An eye protrudes from within the hearth. Human figures appear in all the waliks shown: heads or whole bodies are attached in a, b, d, and e; in c the human figures are within the carving, with the heads pointing towards the lateral ends of the sculpture.
- 246

designating frieze, also means ‘shark’, a masele (pidgin, masalai), or clan emblem (Peekel 1927:16), i.e., a totemic figure often associated with a particular place, thus linking clans to particular animals as ancestral figures and to the places in which such figures are said to reside. Likewise, the houses in which the sculptures are displayed in one area are called a ling or ‘caves’ (Peekel 1927:44), entry ways into the dwelling places of the dead that are also masele places and literal homes to sharks. Figure 2c shows human figures within the body of the fish, while 2a, b, d and e show figures carried by the fish. The sculptures are scenes of the new condition of the dead in more ways than we know.

The position of the hearth in these sculptures draws its significance from this fact. We can assert with some confidence that fire has a place in the world of the masele and carried some meaning about the condition of the dead. At the very least it points to the cremation of the dead at one time common in many parts of New Ireland. As corroborating evidence here, we can point to the carrying of the sculptures in canoes as they were brought into display houses, or to their display in canoes. This finds its rationale in an alternative mode of burial: disposal of the corpse at sea. Though the data are not yet available to answer the question, it would be interesting to know, then, if walik-hearth figures were common to areas practising cremation, and if other sets of figures were linked to sea burial.

A dance-mouthpiece reproduced in Krämer (1925:plate 68, bottom; see also Heintze 1969:fig. 17). The hearth form appears on the back of this bird-lizard form. Objects such as these were carried in the mouths of dancers. Often the dancers would insert such pieces into the mouths of standing sculptures. Many mouthpieces repeat motifs evident as one part of larger sculptures.
- 247

The motivation of the hearth form may become more clear if we turn to our second line of inquiry and the kulepmu malagan. A kolepmu is distinguished, as I have said, by the open hole in its centre. Modelled skulls were placed in the hole, or boys and men placed their heads through it as part of the preparation of the sculpture for display. Alternatively, objects were thrown through the hole; Peekel mentions lances, poles and stones. About the significance of these rites there is little specific information. But their import can be seen in a more general use of the sculptures.

While malagan sculptures represent a scene with ritual significance, they are also themselves handled within ritual sequences. The carvings merge with their contexts of execution. By this I mean that the scene displayed in a carving is also represented in the acts of spectators, so that the carving is only one part of a larger drama or depiction of the transformation of the dead and the fate of the living. Thus, maskers appear along with the carvings, and these represent particular dead; Walden and Groves mention how these figures were called by the names of the deceased and met with wailing. Also, dances dramatised the acts displayed in certain carvings, for example, the acts of birds (Peekel 1928). The cremation of a dead person already points ahead to the transformation scene displayed in carvings, and Walden mentions that the deceased was placed in a “nest” on the platform where he would be burned. Finally, the display of the corpse with poles to prop up the chin looks forward to the standing carvings that show these poles as snakes or infilled extensions (Biro [Bodrogi 1967]). Thus, in these rituals the carvings were often an actor. This can be demonstrated more concretely as well. The masks worn by dancers might be hung up in the display house. Or dance mouthpieces might be used by dancers and then inserted in the mouth of a carved figure. Or, to return to our current example, rather than a skull being placed in the hole of a kolepmu, a man might also stick his head through it.

Thus, the scene depicted in a carving extends outward into its context of execution. And so we see a basis for the similarity of form between walik and kolepmu. The fire of cremation depicted in walik is co-ordinate with the water-hole or entry way into the world of the deceased depicted in kolepmu. In both cases, carving and ritual depict the passage between life and death. Confirming evidence again comes from Powdermaker, who reports that the centre hole of kulepmu (kolebmur) is called mut or ‘door’ (Powdermaker 1933:21), an appropriate sign of passage. Thus, it is no accident that the two classes of sculpture share the same form. Within the semantics or sense of the sculptures, hearth and water-pond are related; a central circle can stand for both. Analysis of walik and kolepmu shows not two referents that conflict but rather two referents linked in sense.

In fact, the third path of analysis mentioned above confirms this finding. The auvere malagan is truly an intermediate form between these two classes of malagan and offers revealing data on both. According to Walden, this malagan has a hole in the middle that is called ‘fire’, which is poked through with a spear made of plant material during a ceremony (Walden 1940:28). It thus shares the two features that otherwise separate kulepmu and walik: it has the hole, as in kolepmu, but this hole is called fire, as in walik. Moreover, this fire is treated as if - 248 it were the pond in kolepmu, that is, an object is inserted through it. Auvere also shares the fish heads at each end and the birds whose heads are held over the central circle.

Here, then, is further proof of the link between fire and water-pond and their representation as a circle in the centre of a carving. This intermediate malagan is a powerful demonstration that malagan forms are linked in sense relations. These relations span different classes of carving and allow a single form to represent a group of referents linked in sense. Thus, all three avenues of investigation confirm a recognised field of meaning, where the names, uses and referents of the form fall into a coherent pattern. My claim is that a New Irelander might only be familiar with kolepmu, say, and thus might only be able to point to one referent of the circle. But the ability of the circle to represent a water-pond none the less depends on a larger set of sense relations, which I have tried to outline here. This field of sense relations can be reconstructed from diverse data, and it reveals the minimal social consensus about meaning mentioned above. In this way, a principle of the organisation of form in the carvings can be said to be an expression of this consensus and a measure of the social reality, rather than individual psychological reality, of the meanings expressed in the carvings.

How far this social consensus extends, either spatially or temporally at a given time, is a more difficult question. I have argued that the social reality of the meanings expressed in the carvings exists on an implicit, systematic level, so that particular realisations of carvings at a given time or place draw on meanings often set down at another time or in another place, and with perhaps other meanings than those evident to a carver and his community. Therefore, in this paper a definite localisation of this consensus in time or space has been neglected in favour of a less direct but perhaps more suggestive approach, namely, analysis of the broad coherence of form and meaning in a series of carvings. However, coherence at this internal level would be supported if an external measure of such coherence could be devised. This would amount to a measure of a community's participation in the conceptual world of the carvings. The necessary data for this measure would include a detailed record of all the carvings commissioned and realised in a community within a given time, with particular concern for the transmission of knowledge associated with these events, i.e., what clans gain rights to execute carvings; what masks, dances, songs or myths are brought from one community to another; what affinal links or land claims are made public; etc. Given these data, one could conceivably map a set of carvings and particular forms on a grid of time and space, showing where iconographic and conceptual transitions or variations occur in the malagan region. This would require data that are still not available, yet Lewis early on suggested just this sort of inquiry into the “life history” of the carvings as the only way to establish the true contours of the malagan system (Lewis 1966:175).

In this paper I have, of course, only scratched the surface of the malagan system. I have tried mainly to be programmatic, indicating a new approach to the problem of form in the carvings. What I have presented here does not allow a conclusion about the system as a whole; that must remain a task for the larger - 249 study mentioned earlier. Still, even this first sounding in the systematics of New Ireland art allows us to hazard a conclusion about form and representation in malagan.

The big question is why New Irelanders choose to depict certain principles in a plastic art form. Brouwer has given one rationale. In her study of the Pinatkin malagan series labadama, she concludes that control of the movement of the maskers and control of space in the carvings are allied; both are ways by which New Irelanders depict certain aspects of the nonmaterial world in order to control this world of spirits and otherwise uncontrollable forces (Brouwer 1980:445f.).

Brouwer's claim is very general. It applies best to the masks she observed and applies to the malagans described here only by stretching the meaning of “control” beyond its accepted sense. For example, it is hard to see how control of space in frieze malagans implies control over the total scene they depict. One might with equal plausibility claim that depicting the scene “ungates”, to use Gombrich's (1960) term, a wide range of emotions and ideas, for little can be known about the channelling or control of subjective response to the carvings. But I am uncomfortable with Brouwer's claim for another reason. For at least one New Ireland myth explicitly concerned with representation goes contrary to Brouwer's emphasis on control.

In a myth collected by W. W. Cox and published in Man in 1913 we find a creator-giant named Sikodo who tells his nephew to make a hat boroi ‘a representation of smooth stone said to resemble a pig’. The nephew, Padamalana, makes this image but out of sand, dark sand on one side, light sand on the other. People come to see the image and pay him. Later the image is decorated with fish and seaweed before Padamalana destroys it.

This is an origin myth about malagan, though it is interesting to note that commemoration of a deceased is not mentioned. Like many similar tales, it involves a transfer of knowledge about making the image from spirits to men. Most interesting here is the focus on what makes an image. Stones are supposed to represent a pig, but sand is used to stand for the stones. And the stone-sand pig is decorated with fish and seaweed. What is the referent of this image, and what is used to represent it? This myth would seem to show multiple layers of representation, a sophisticated notion of representation. The myth speaks not of control over images, but rather of the image-like quality of all things. Pigs can be represented by stones, but stones can be replaced by sand, and the image of pig can be merged with fish.

I conclude that a far more complex idea of representation informs malagan. I cannot go into this here except to say that malagan sculptures show a clear recognition of the image as an image, so that the image is not simply a stand-in for something represented. Thus, the image is not restricted to representation of a deceased. Rather, the sculptures explore the power and ambiguity of images, which depends on recognition that the images are and are not the objects they portray. Thus, rather than too easily assume our own separation of image and object represented, with the Romantic notion that the one enables control over the other, I suggest that we explore the concept of representation that actually in- - 250 forms malagan. The analysis of sense relations and the distribution of forms across classes of carvings, which I have outlined here, is designed to carry us some distance in this effort.

Concern for sense relations in iconographic form also allows us to help settle a long-standing debate in studies of malagan. Starting with the earliest inquiries about the sculptures, observers have asked whether the lack of exegesis and fortuitous pattern of realisation that mark the system is a recent development or a systemic component. Even this short inquiry shows that a wide-ranging network of meaning governs the particular forms that appear in the carvings. It seems safe to assume that this network of meaning must have emerged from within a more tightly knit carving system than the one evident in recent decades. Thus, I think it safe to say that the current fragmentary system is a recent development. It remains to be seen whether the increasing spatial mobility in New Ireland will lead to a more cohesive system or to further devolution.


The author would like to thank the participants of the informal session, “New Ireland and Environs: History and Ethnography”, organised by Dorothy Billings, at the Association for Social Anthropology in Oceania meetings, March 9, 1985, in which a version of this paper was read. Thanks also go to Phillip Lewis, who offered welcome criticism in the preparation of the essay. Ruthe Karlin prepared the artwork.

- 251
  • Boas, F., 1951 [1927]. Primitive Art. New York, Dover.
  • Bodrogi, T., 1967. Malagans in North New Ireland. L. Biro's unpublished notes. Acta Ethnographica, 16:61-77.
  • Brouwer, E., 1980. A Malagan to Cover the Grave: Funerary Ceremonies in Mandak. Unpublished Ph.D. thesis, University of Queensland.
  • Bühler, A., 1933. Die Totenfeste in Neu Irland. Verhandl. Schweizerischen Naturf. Gessellschaft, 114:243-70.
  • Cox, W., 1913. New Ireland Myths. Man, 13:195-9.
  • Fergie, D., 1983. Malanga: The Reproduction of Cosmic Conceptions. Paper read before Anthropology Graduate Forum, University of Chicago, March 1984.
  • Gombrich, E., 1960. Art and Illusion. Princeton, Bollingen.
  • —— 1963.Meditations on a Hobby Horse. London, Phaidon.
  • Gunn, M., 1982. Tabar Malagan: A Report on Fieldwork. Museum and Art Galleries of the Northern Territories. Darwin, Australia. Typescript.
  • Hanson, A., 1983. When the Map is the Territory: Art in Maori Culture, in D. Washburn (ed.), Structure and Cognition in Art. Cambridge, Cambridge University Press.
  • Heintze, D., 1969. Ikonographische Studien zur Malaggane-Kunst Neuirlands. Tübingen, Schwitalla Himmelsthür.
  • Krämer, A., 1925. Die Malanggane von Tombara. Munich, Georg Muller.
  • Lewis, G., 1980. Day of Shining Red. Cambridge, Cambridge University Press.
  • Lewis, P., 1966. The Social Context of Art in Northern New Ireland. Fieldiana: Anthropology, Vol.58. Chicago, Field Museum of Natural History.
  • —— 1973. Changing Memorial Ceremonial in Northern New Ireland. Journal of the Polynesian Society, 82:141-53.
  • Lewis-Williams, D., 1981. Believing and Seeing: Symbolic Meanings in Southern San Rock Art. New York, Academic Press.
  • Parkinson, R., 1907. Dreissig Jahre in der Südsee. Stuttgart.
  • Peekel, G., 1926-1927. Die Ahnenbilder von Nord-Neu-Mecklenburg. Anthropos, 21:806-24, 22:16-44.
  • —— 1928. Lang-Manu. Festschrift Publication D'Hommage Offerte au P. W. Schmidt. Wien.
  • Powdermaker, H., 1971 [1933]. Life in Lesu. New York, W. W. Norton.
  • Skorupski, J., 1976. Symbol and Theory. Cambridge, Cambridge University Press.
- 252
  • Walden, E. and H. Neverman, 1940. Totenfeiern und Malagane von Nord-Neu-Mecklenburg. Zeitschrift fur Ethnologie, 72:11-38.
  • Wollheim, R., 1980. Art and Its Objects. Cambridge, Cambridge University Press.
1   Variations in the orthography of the term derive both from regional dialect differences across northern New Ireland and from alternative transliterations. Thus, in Tabar Gunn (1982) speaks of malagan, and in Lesu Lewis (1966, 1973) and Powdermaker (1971 [1933]) speak of malanggan; others refer to the carvings as malangan. Krämer (1925) refers to malanggane, and Bühler (1933) refers to them as mulligan.
2   This emphasis on minimal consensus and the potential for more rigorous quantification of shared knowledge has been developed by Sankoff in two papers concerning the Buang of Morobe Province, Papua New Guinea: “Cognitive Variability and New Guinea Social Organization: The Buang Dgwa” and “The Quantitative Analysis of Sharing and Variability in a Cognitive Model”. Both are reprinted in G. Sankoff, The Social Life of Languages (Philadelphia, University of Pennsylvania, 1980).
3   But see Gunn's Tabar informant's remark: "Salle said that if they said [the carving] was a lunet then it was a lunet, although other people might think it was a marada". Gumm concludes, "apparently it is not easy to tell from what genera a malagan sculpture will belong" (Gunn 1982:62). Loss of distinct names for carvings is no doubt a sign of change in the consensus upon which the art rests. But is it something new or a regular feature of the system? I take up the question below.
4   Powdermaker gives 10 "individual malanggans" within the "large group" of kolebmur (Powdermaker 1971 [1933]:315-16). Of these 10, I have managed to track three down in other sources (araraun, lekiu and balanei); each shows an iconographic organisation distinct enough ot have been recognised by a number of ethnographers. One other (agasmuggawuh) may be a variant of another named type that appears within the malagan classes, i.e., a gas(a malagan and spirit figure). The six remaining names (kuwawuwu, kuleipanga, ambeli, ammaris, amarindan, avutimisi) may be “personal” malagan, to use Peekel's distinction, but this is an argument from an absence of proof to the contrary, and is thus invalid. In fact, more than likely “personal” and “historic” or “mythic” malagan, to use Peekel's distinction again, shade into one another, with a malagan originally commemorating an individual produced enough times with enough regularity of form for it to become “historic” or “mythic”, and hence iconographically distinct. Evidence would also point to a traditional malagan type being combined with a representation of the individual commemorated at a particular feast.