a Mangarrayi representational system: environment and cultural symbolization in northern Australia

a Mangarrayi representational system: environment and cultural symbolization in northern Australia

FRANCESCA MER LAN-University o f Sydney

This paper describes an unusual sign system used by the Mangarrayi, an aboriginal group of the western Roper River, Northern Territory. This sign system represents the conjunction and integration of a number of analytically distinguishable systems of cultural symbolizations. Each involves a complex of social categories and beliefs, and an understanding of these i s necessary in order to appreciate the organization of the sign system, i t s meanings, and its functions. The Mangarrayi are a people whose traditional country i s shown approximately on Figure 1. The Mangarrayi language belongs within a genetic subgroup that includes Alawa, Warndarrang, Marra, and (now extinct) Yugul. Mangarrayi is not mutually intelligible with any of these languages. Many Mangarrayi, especially older people, are proficient in at least one other language. Most speak more than one, either within the same linguistic subgroup or, like Yangman and Jawony, genetically distant but geographically close. Especially since the establishment of pastoral stations in the Roper River region, many Mangarrayi have lived in close association with Yangman people in aboriginal station camps; the approximate extent of traditional Yangman country i s also shown in Figure 1. The numbers of Mangarrayi and Yangman were greatly reduced, both by disease and by a concerted effort on the part of pastoralists t o exterminate "wild" aborigines in the latter part of the 19th century and the first decades of the 20th century (see Merlan [1978] for Mangarrayi accounts of this period). Only a handful of people are left who identify themselves as Yangman, though through intermarriage with Mangarrayi some identify themselves as Yangman-Mangarrayi.

I t is common enough for aborigines in many parts o f Australia to regard environmental features as visible signs o f world-forming activities of totemic creator figures. The relationships between individuals, social groups, and the creator figures are cast in complex ideologies that express the ties between men and locality. In the Northern Territory along the Roper River, an aboriginal group, the Mangarrayi, also use vast numbers of environmental objects, mainly trees, to represent directly individuals linked to creator figures who made the localities where the objects stand. Thus, a l l three components o f totemhnantland relations, which are so central a part of aboriginal culture, exist or are represented in the landscape. This paper explains how the environment i s utilized to represent this system of relations in a way and to an extent that has not been previously reported in the Australianist literature, illustrating the cultural constitution o f the "natural" environment. [Australia, totemism, environment, semiotics]

Copyright @ 1982 by the American Ethnological Society 0094-0496/82/010145-22$2.70/1

Mangarrayi environment and symbolization

145

Figure 1. Locations of mentioned language groups in the Mangarrayi area

The majority of people who call themselves Mangarrayi and Yangman-Mangarrayitoday live at JembereAboriginal Community, an excision area within Elsey pastoral station. Elsey Station has existed since the early 1880s (though the homestead has been moved several times), and this and other local stations have been principal centers of European influence on the aborigines.’ The aborigines moved away from their former camp at Elsey Station in 1974,after the implementation of award wages (in 1968)and a changing political climate made possible the first steps toward the establishment of aboriginal communities independent of station control. The present population of Jembere fluctuates seasonally between about 45 and 65 people, depending on work opportunities, ceremonial activity, and other factors. Jemberei s within traditional Mangarrayi country. Throughout their contact history, a sizable proportion of Mangarrayi people have lived within their traditional country; others have lived either seasonally or permanently at other aboriginal communities within Mangarrayi country (Coondooloo and Moroak Stations), at the periphery of Mangarrayi country, or outside it (at Hodgson Downs and Roper Valley Stations, Ngukurr and Bamyili settlements, or the town of Katherine). Even today the Mangarrayi have an intimate knowledge of many parts of their country and a great attachment to it. Much of this country i s well watered and well provided with trees. The Mangarrayi pride themselves on their association with the riverine country along the Roper and other bodies of water within their country. Numerous prominent tree species grow along the rivers, lagoons, and billabongs; many of these species are used in the representational system described here.’ Large tree species, such as ironwood and smoke tree, also occur in drier or only seasonally wet areas. Many Europeans have had the experience of being shown named land areas by aborigines and have been told about the mythological creator(s) who traveled through them or emerged from them and endowed the areas with meaning, often leaving signs of their activities and presence in features of the landscape. Aborigines also recount who stands in intrinsic relationships to land areas, relationships that we imperfectly label “ownership” in English. I had spent several months with Mangarrayi people and had been shown various localities ”created,” or imbued with significance, by certain mythological figures. I had been told of the travels of some of these creator figures, the generic Mangarrayi term for

146

amerlcan ethnologist

which i s WARRWlYAN (dreaming). Some WARRWIYAN are far-traveled, coming into Mangarrayi country from distant places; their travels constitute a jigu (road), the sum of localities at which the WARRWWAN ”came out,” “made” features of the landscape, interacted with other WARRWIYAN, or otherwise left evidence of themselves and their actions. Some WARRWWAN “came out” at one place only and are sometimes called barnamyirrag (unique locality; barnam [camp, named locality] and -yirrag [unique]). The WARRWlYAN are animals and plants of all kinds; some are human in form, and some are incorporeal entities such as dib (sickness) and jab (wild wind). This kind of cosmogonic belief, according to which totemic entities left signs of themselves that endowed a preexisting landscape with meanings for people, is common t o many aboriginal groups in Australia. The visible signs left by the totemic entities, whatever their form, are also called warrwiyan. This usage of the word will be written in l o ~ e r c a s e . ~ One day, as I was sitting with several older women at a sand bank created by wanggij (child) dreaming, one woman asked me if I had seen m y gagag (MMB). Assuming that she was referring to a certain man who stands in that relationship t o me, I replied that I had not, that he had left the camp early in the morning. With a significant look the woman turned and pointed to a large Melaleuca leucadendron, a tall paperbark species that grows in great numbers on the banks of the Roper, and said “That i s your gagag,” naming the man to whom I thought she was referring. Pausing to see if I understood, she turned and pointed to another paperbark on the opposite bank and said “There stands your MMB,” naming a now-deceased man, full brother of the first as I knew from his genealogy. In quick succession she pointed to a number of other trees and named them. All the names corresponded with those of living or dead persons, and I recognized most of them from m y genealogies. I understood that although not all those she named belonged to a single patriline, all belonged t o one or another patriline falling within a particular semimoiety. Everyone looked rather sly as I realized that I was learning something very important about what these people see as they visit and walk through “country.” The many questions that occurred t o me were not all answered immediately, and some remain; but henceforth, as we visited barnam (localities), older people walked with me and showed me who was “standing up.” Since mainly trees are used to represent people, the Mangarrayi generally speak of the warrwiyan as ”standing up,” using verb particle and inflectable auxiliary construction jirr -jaygi- (to stand). But i t turned out that sometimes rocks, sandbanks, and other features are used to represent people, in which case the warrwiyan are said to ni (sit). In some instances m y companions could not specify detailed genealogical relationships for all people represented but could only remember names and perhaps a few other details. In some localities aborigines named up to 30 people, and there was remarkable agreement on the identity of particular natural objects, whether several people together showed me a locality or the same locality was visited more than once with different individuals. They were no longer restrained about talking to the people represented and explained to me directly and indirectly some of the meanings and functions of these signs, which to us are merely natural objects. Stanner (1965:227) has written that anyone who walks with aborigines through the Australian bush “moves, not in a landscape, but in a humanized realm saturated with significations.” The following description does not, of course, exhaust the complex of meanings, constructs, and associations that lie behind this particular sign system. It does, however, show that the natural environment i s utilized to express a system of meanings in a particular way and to an extent that has not been reported previously in the literature. I was able to find out about this remarkable system only through the kindness of m y Mangarrayi teachers and friends.

Mangarrayi environment and symbolization

147

sketch of social units A brief outline of social units that I recognize for the Mangarrayi may be a useful introduction to more detailed consideration of the relation of these units to the representational system. There are two fundamental kinds of units in terms of which the Mangarrayi cast the relations of people to land, as reflected in the representation of people at localities. These are semimoieties and patrilines. A system of (unnamed) semimoieties constitutes an uninterrupted patrilineal form of organization (Reay 1962:95).As the (perhaps misleading, see below) term “semimoiety” suggests, there are four of these, which provide a means for conceptualizing and organizing links between fathers and their children. Each person is prescriptively in the same semimoiety as his F/FZ, FF/FFZ, FFFIFFFZ, and so on; also, a man’s children and his son’s children are in the same semimoiety as himself, while a woman’s children and her son’s children are in her husband’s semimoiety. So strong i s the feeling that children should belong to their father’s semimoiety that potential irregularities in the social classification of children of nonpreferred marriages are usually corrected by placing the children in the father’s semimoiety, regardless of the social classification of the mother. The Mangarrayi do not use distinctive proper names for semimoieties, unlike some other Roper tribes with semimoiety systems (such as the Marra; see, e.g., Spencer 1914:60-64).Instead, the Mangarrayi refer to semimoieties by pairs of subsection (or “eight-class,” per Shapiro [1979])terms. In i t s own right, the subsection system i s frequently used by the Mangarrayi to talk prescriptively about marriage (see note 4). But in its linkage with the semimoiety system, which is conceptually more central in the social construction of people’s relationships t o land and to other people, each subsection is significant as an alternating generation class of a semimoiety. As is generally the case where subsections are used in Australia, the Mangarrayi tend to explain subsection assignment of children as being determined by that of the mother. However, as mentioned above, the principle of matrifiliation that prescriptively governs subsection assignment i s overridden where “following the mother” would prevent a child from belonging to the father’s semimoiety. The patriline i s the minimal landholding group. Membership in these kin groupings is immediately determined by patrifiliation, but a person’s identity within the patriline i s in fact built upon the sharing of names and other links with second generation agnative kin (FF for boys, FFZ for girls). There is no Mangarrayi term for patrilineal grouping at this level, as there is among neighboring Arnhem peoples. The Mangarrayi tend to express relationships to land in terms of semimoieties, a mode of speaking that disguises the existence of patrilines within the overarching semimoiety categories. Above, I spoke of the term “semimoiety” as perhaps misleading in relation to Mangarrayi social organization. It is so if it is taken to suggest that existence of clearly recognizable moiety organization. The Mangarrayi do indeed recognize links between two semimoieties, one’s own and that containing kinsmen of the types MM/MMB. This pair is opposed in a variety of ways to the other two semimoieties. For certain purposes (e.g., ritual cooperation) the two linked semimoieties are considered to be interdependent. Such semimoiety pairings may appear to be structurally comparable to patrimoieties among some neighboring Arnhem groups. However, I believe the links between pairs of semimoieties among the Mangarrayi consist essentially of kin-based ties between individuals and small patrilines, rather than reflecting membership in social groups at the patrimoiety level. In short, while semimoiety systems such as those used by the Mangarrayi seem to have a builtin potential for increasing institutionalization of relations between semimoieties into higher-level patrilineal groups, I believe it would be forced to find here a clearly delineated patrimoiety organization such as is found in some other parts of Australia.

148

amerlcan ethnologlst

The Mangarrayi formerly participated, at least to some extent, in a system of matrilineal social classification similar to the ngurlu classification described by Stanner (1979:13, 37, 50) for groups t o the south and west of the Mangarrayi, including the Mudbura, Curindji, Wardaman, Djamindjung, Warumungu, and Djingili. The Mangarrayi refer to this system as wururr and are aware that the Yangman observed it. Today, however, Mangarrayi knowledge of this system i s fragmentary, and it is possible that even in the past it was not an integral part of Mangarrayi social classification. It seems to have no bearing on the representational system.

relations of the WARRWWAN to land and people The totemic creator figures, the WARRWIYAN, brought men and placed them in the form of trees and other objects at localities created by the WARRWIYAN. The objects that represent people are, however, not simply part of some "natural" environment: the Mangarrayi c a l l them by the same term, warrwiyan, as the creator figures themselves. (See Merlan [I9801 for the inappropriateness of a distinction between "natural" and "cultural" in regard to Mangarrayi totemic entities.) When the Mangarrayi speak of a tree that represents a person, they refer to it as warrwiyan, landi (warrwiyan, [in the form of a] tree), or they name the particular tree species. Hence, the objects are indeed signs of the presence and activities of the creative forces. Not a// signs left by the creator figures represent people, however; at each locality there may be one or more prominent features (a tree, rock, bend in the river) explicitly said to have been fashioned by the creator figure but not associated with any individual. Such objects are also referred to as warrwiyan. But the particular class of objects with which this paper deals are those which represent people for the Mangarrayi. Each WARRWIYAN is known to belong to a particular subsection. For example, wijwij (opossum) is of the subsection ngarrijbalan, and garawi (male plains kangaroo) is of the subsection gamarra. It is important to note, however, that especially animal and anthropomorphic WARRWlYAN tend to be clearly identified as members of one subsection within a semirnoiety, while many plants and other objects are identified as belonging t o both subsections within a semimoiety (see Merlan 1980). This evidently reflects the greater salience of animals and humans as potential actors. In any event, the Mangarrayi use of a subsection nomenclature thinly disguises an underlying system of semimoieties according to which they actually organize many aspects of their thought on ritual or totemic affiliation.' Although the semimoieties are unnamed, there is a generic term, lirrag, which we can translate as semimoiety and for which the Mangarrayi supply as equivalent expression the dyadic kin term barda-yi (father and child). People refer to a semimoiety by combining the two subsection terms that comprise the father-child couple (see Figure 2). The Mangarrayi recognize and can use the patrimoiety names Yirrija and Duwa, widely reported from Arnhem Land, though they think of them as borrowed or introduced (see Figure 2). There are, however, recognized links between the two semimoieties (or kinsmen within them) that structurally approximate that patrimoiety of some Arnhem groups. The term darlnyin i s applied to specific individuals (usually close or actual kinsmen) of the MM/MMB types and reciprocals, who should be helpful and supportive in many contexts including ceremonial ones. This term, which connotes a cooperative and friendly relation, may also be applied at the most general level to the entire semimoiety linked to one's own. Usage of the term darlnyin i s paralleled by institutionalized joking between kinsmen of the MM/MMB types and those of the reciprocal categories. The Mangarrayi say of the members of these two linked semimoieties wow? ja-wurla-wa-n, "they are company, in league."5 They also say of relations between members of the two linked semimoieties yawarl ja-wurla-yag, "they cadge," meaning that concerns relating to land and ritual are

Mangarrayl environment and symbolization

149

bongoriny

- - - - - -/oml/n

---.- - - -gongilo - burrolo

ngorr//balon burlanybolyorriny

- - -- -

-

-garnorm

(a) Patrilineal semimoieties as pairs of subsections, showing correct or first-choice marriage between the subsections.

burlany

(bl Matrilateral moieties, showing cycles of subsection affiliation.

Figure 2. Subsections showing patrilineal semimoieties (a) and matrilateral moieties (b)

shared, so that persons in the two semimoieties are involved in a constant give and take in relation to mutual obligations. A link between the other two semimoieties is also implied by the widest usage of another term, barrai. This connotes property (ritual or totemic, including land) of the other semimoieties: mother’s semimoiety and her linked darlnyin semimoiety. Used most narrowly, barraj may be applied to a mother’s semimoiety only, and sometimes to property of one’s actual mother. The term barrai is usually glossed nanyi (mother).6Of a certain land area and/or the associated creator WARRWIYAN, or any object at the locality, a person may say barrai-nganju (my barrai) or ngarla-nanyi(my mother [subject feminine noun class prefix ngarla-I). Also, ritual objects symbolizing totemic figures are referred to as barrail nanyi and darlnyinlgagag. At the level of greatest generality, each WARRWlYAN is spoken of as belonging t o a particular semimoiety, and each land area with which the WARRWWAN is associated is also spoken of as belonging to that semimoiety as a whole. While people/WARRWIYAN relationships are spoken of in this way, the WARRWWAN are also placed in relation to environmental features and zones. For example, wiiwii (opossum) i s often described as belonging to bulula (“jungle,” i.e., bullrush and lagoon country), within which many opossum sites are located. By association, people affiliated with opossum are also said to belong to, or be of, bulula country. Every named land area i s associated with at least one totemic creator. Where there are multiple associations, usually only one i s the principal figure with whom the shaping of the locality is associated. Other WARRWIYAN with whom the principal one interacted at the locality are sometimes of the same semimoiety, sometimes of others. The WARRWIYAN, as they traveled or emerged, left “people” of their own semimoiety. A tally of people represented at sites made it clear that the fundamental landholding groups are actually patrilines that fall within a particular semimoiety, though not usually all of them. The Mangarrayi talk about relationships of people to land areas at several levels: at the level of individual relationships to land and creator figures; at the level of the patriline, by tracing inheritance of rights to land and ritual resources through socially recognized agnatic connections; and at the level of entire semirnoieties. The latter i s actually the most frequent and contextually unmarked way of talking about people/land relationships. There are reasons why this way of speaking should be the most widely used. For one thing, the semimoiety constitutes a higher level of generality that can encompass the notion of more than one patriline having a joint relationship to land areas or WARRWIYAN. This situation occurs quite often, especially when long and important creator tracks run across portions of Mangarrayi country and sometimes extend into areas that Mangarrayi recognize as the territories of peoples of other language groups-Yangman, Alawa, and so on. Because of

150

american ethnoioglst

this higher level of generality, the use of the semimoiety category allows for the possibility-which the Mangarrayi recognize but verbalize only in terms of culturally accepted modes of totemic affiliation and their acquisition-that people of descent groups without previously known or recognized affiliation with particular land areas and WARRWIYAN may advance claims of relationship to them. This includes individuals who marry into the Mangarrayi community. The potential for change within the system of man/totem/land relationships is an essential part of a complex that is characterized by an ideology of absence of change since the establishment of the cosmological order (Stanner 1965). Thus, when an aborigine enters a locality, often centered around some striking physical feature (a bend in the river, a tree, a rock outcrop), besides knowing what the WARRWIYAN did or said there, he sees the signs of the intrinsic relation between men and the totemic figure in the form of warrwiyan the people brought there. Around a billabong or along the river he may see many trees and rocks, but he does not refer to them all as warrwiyan, nor do all large trees represent people. Some trees are jalala (green, young) and are mun landi (just trees), part of the environment. Mangarrayi see those that represent people as having always existed-or, to use Stanner’s (1956:52)more apt phrase, as existing in the “everywhen,” in a dimension that cross-cuts our ordinary experience of the passage of time. This dimension is not limited to the past, and the activities that take place within it have not come to an end. Individuals and descent groups are affiliated with more than one WARRWlYAN (though certain totemic affiliations are more salient t o individuals and descent groups in some ways and in certain contexts than in others). Hence, the same ”people” may be represented by warrwiyan at localities connected with various different WARRWIYAN, as well as at more than one locality associated with the same WARRWIYAN. For example, a certain man (now deceased) of burlany subsection i s represented by trees at localities created by channel-bill cuckoo and by olive python; he i s found at more than one locality of each creator figure. In addition, not all people who have an intrinsic relation to a WARRWlYAN are represented at each locality it created. Mangarrayi say, for example, that only two brothers are represented at a particular opossum locality, whereas at another opossum site these and other people stand, including some members of other patrilines. There i s no doubt from the way Mangarrayi speak of such things that other members of the two brothers’ patriline have the same affiliation with the opossum WARRWIYAN; they are simply not represented at that particular place because nobody brought them there. They are linked to the WARRWlYAN without respect to particular place; they are represented at some opossum places. In this sense an individual may, at any given locality, effectively represent the affiliation of an entire patriline t o the place and to the WARRWIYAN. My reason for identifying patrilines as the unit that individuals can represent depends upon some of the Mangarrayi concepts relating individuals to descent groups.

individuals and descent groups For the Mangarrayi, each person represents the warlayjnyin (Pidgin English “shade”) of an actual, or sometimes close, classificatory FF (if the individual i s male) or FFZ (if female). The shade comes as a wanggij (child), seeks out its father (whom it recognizes), and asks t o be shown its mother (whom it cannot recognize until shown) so that it can enter her body. Warlayjnyin also refers to a shadow, and as a person’s shadow may be said to follow him, it seems to have connotations of coming after or following.’ When a person dies, an entity that the Mangarrayi call mirirr, and translate into Pidgin English as ”spirit,” i s ritually sent away “forever.” They say the mirirr, by the singing of appropriate funerary songs, i s first made to go in a direction called dinyjalin, which connotes

Mangarrayi environment and symbolization

151

gayarra (west), then mawurr, which connotes gayanyja (east), and finally it goes east forever. Although the Mangarrayi do not, and will not, neatly describe the connotations of mirirr as opposed to warlayjnyin, the way in which they talk about it suggests that mirirr is some aspect of the spirit of individuals, and this is disposed of (“sent”) during the mortuary rites. But the shade, which underlies the oneness of FF and 55,appears to express the importance of the patriline. The returning wanggij in some cases is described as incorporeal, but sometimes it may come in the form of an aquatic species, like a fish, since the Mangarrayi believe that children emerge from water. Although the wanggij i s said to look for its father, it may first be recognized by someone else. (In the instances I was told of, this was always by a close patrilineal relative; one woman caught her brother’s child as a barramundi, as is now evidenced by the fact that the hook, which tore its mouth, left a scar that the child bears to this day. The barramundi species is not a totem associated with the child’s semimoiety.) Otherwise, people say that gurwaran (clever men) are able many times to divine the identity of the shade as it seeks i t s father. The main point is that there is an assumed link between the child and someone, usually an actual or close relative of the FF class. Although people have individual physical and mental characteristics, each i s animated by a life-force that preexists within the patrilineal line of descent. Much i s integrated into the concept of “name” (see Stanner 1956:53).The generic word for “name” i s ni, but warlayjnyin (“shade”) is a commonly used synonym for it. Children traditionally were not given names at birth (though newborns were given English names). When able to toddle or old enough so that i t s survival seemed relatively certain, a child was given a first name, one of those borne by the FF or FFZ to which it was linked. Although it may seem inconsistent to us, the Mangarrayi say that a child can be given this first name while the FF or FFZ i s s t i l l alive. In fact, it i s not uncommon for older people to say which grandchildren will assume their names. Thus, sharing of the warlayjnyin does not depend on the grandparent‘s being deceased, though in many instances the grandparent is dead by the time of the child’s first naming. In the past, a small ceremony marked the giving of the first name. In cases where the grandparent was still alive, this involved the child’s superimposing his footprint upon the ground on top of that of the grandparent. If the grandparent were dead, the name could be given by a patrilineal relative, usually a close The first name assumed is that which the Mangarrayi describe as mun barnam (only country). This is usually the name by which the bearer is most commonly known in his daily life; it i s also the name by which the warrwiyan representing the person are called. There i s little prohibition against referring to individuals and calling them by this name, and sometimes shortened forms are used as nicknames (janggurl [short], as the Mangarrayi say). When the name has been newly given to someone, the warrwiyan are again associated with a living person, whereas for a period (assuming the grandparent had died) there may have been no living individual of that name. Hence, at localities a single name may evoke memories of the former bearer of the name and the present bearer, and, possibly, the grandchild who i s to assume the name may be mentioned if consensus has been reached on this point; that i s the extent of the actual generational depth with which most warrwiyan can be associated. That is, the warrwiyan constitute a continuous system of signs through which changing generations of referents are cycled. Besides the rather public kind of name mentioned, each person assumes a much less commonly used name, usually taken from a song-cycle associated with a WARRWlYAN and his sernimoiety. People are usually reluctant to use and hear this name in public, but sometimes take pride in confiding what it i s and in the associations it evokes. This name was also formerly borne by a person of the FF class, but it need not be the same FF as the

152

amerlcan ethnologlst

one who bore the more public name. There seems to be some latitude in the nature of the association between the person who assumes the private name and the FF. People told me of several cases of individuals who had taken the names of FFs with whom they had close social and ceremonial ties, but not close genealogical ones. This name, I was told, i s usually assumed in a ceremonial context and apparently may change at different times throughout a person's life. One man told me that on the death of this actual brother he assumed a name the latter had borne for a time. It i s not too uncommon for people (sometimes even close kin) to seemingly not know, or at least be unsure of, the private names of their relations; quite commonly, the names of certain more distant or deceased kinsmen are unknown to many people. Both men and women know the names of some people of the same sex, and of the opposite sex. This kind of name seems t o express not so much peoples' determined positions within corporate descent groups, as subsequently formed ties that they as adults choose to emphasize. People say the name i s usually assumed during critical periods, particularly during ceremony, at which changes of status and significant links to others may be publicly displayed and validated. Another kind of appellation is based upon the link between a person and the particular FF or FFZ whose warlayjnyin is shared. This is formed by adding to the toponym designating the place of death of the grandparent the suffix -wuyi (--buy; following nasals and stops). The equivalent Yangman suffix i s -buwa. This kind of grandparental necronym i s generally seldom used in address, more often in reference and in specific contexts that highlight the grandparent as a link. Thus, ngarla-Manyjurngan-buy; might be used with reference to a woman whose FFZ died at Manyjurngan, a place. Although her semimoiety is not totemically affiliated with this place, it is a locality of some ceremonial importance, and continuing recognition of a relationship to it i s possible through use of the necronym. A few individuals are commonly referred to and called by a form of the necronym from which the suffix i s dropped, and the toponym perhaps shortened or i t s form slightly altered. In this form the appellation may be used commonly, even as a nickname. This kind of necronym seems to now be somewhat less important than formerly. As people have become more sedentary, deaths tend to occur in the hospital or in the main camp. Many individuals have one or more nicknames descriptive of some (usually physical) characteristic, either congenital or the result of some accident or circumstance, such as obesity. Some individuals are commonly and publicly known and called by the name of a totemic species; the mode of their affiliation to the particular totem is "conceptional" (see below). In sum, the most public name is generally that by which warrwiyan and the people to which they refer are known. In trying to devise a descriptive label for this kind of name (as no special term exists for it in Mangarrayi), some aborigines described it in Pidgin English as a "country name," or said in Mangarrayi that it was mun barnam (only country). The importance of the link between persons of the same descent group i s reflected in special uses and forms of the kin term murimuri (FF, FFZ). (Gender of the referent is distinguished by noun class prefixes.) The term for the reciprocal category is marranju (SS, SD), but Mangarrayi, like many languages in the area, has a set of "dyadic" kin terms. These special formations, based entirely or in part on simple kinship terms, express meanings of the sort "X and Y," or "X-Y relationship," such as "father and child" (see Heath and Merlan in press). The dyadic term denoting the relationship "FF and SS" i s muri-wa-yi, the plural dyadic term-with plural denotata FFs and SSs-is muriri-wa-yi." Although various sibling terms exist and may be the basis for dyadic terms, one way of referring to "siblings" is by use of the dyadic terms based on murimuri. The plural dyadic term may also be used to designate those of the same descent group, including F, 5, FF, SS, and others of the patriline. Other uses or forms of murimuri emphasize the component of meaning "same descent group" overriding generational difference.

Mangarrayi environment and symbolization

153

Although people and their descent groups are affiliated with a variety of WARRWIYAN, certain affiliations are particularly significant to individuals, and others to individuals as members of patrilines. Following Stanner’s (1965:227)categorization, totemic affiliations particularly significant to individuals may be termed “conceptional” and those especially significant to entire patrilines “descent-affiliative.” For the Mangarrayi, the former i s usually a special subcase of the latter, in the same sense that the “conception” area i s within the set of the patriline’s land/totem affiliations. As mentioned, wanganggij (children, reduplicative plural) emerge from water at particular localities; whence they emerge i s known first to close patrilineal relatives.” The WARRWIYAN creator figure of the place of emergence i s the totemic figure with which individuals identify themselves, saying “I am X” (opossum, wild wind, sugar bag, and so on). The link with this WARRWIYAN i s often evidenced by a physical trait, as a person who emerged from a black cockatoo site bears reddish markings which are the red coloration of the bird’s tail (see Strehlow [1978:23] for similar evidence of conception links among the Aranda). Thus WARRWIYAN, who gave rise to both human beings and totemic species, link them indivisibly. Some individuals are known by the name of their “conception totem.” The place of emergence is in theory associated with a WARRWIYAN to which the individual’s patriline already possesses recognized affiliations, except that the same imagery of emerging from a particular place within Mangarrayi country often describes the establishing of totemic ties by those who marry into the Mangarrayi community, thus making this equivalent to conception filiation. A person’s particular conception totem may be the same as his father’s or FF’s, but it need not be. Often among groups of siblings, at least one has the same conception totem as the father. Each patriline has at least one, and sometimes two or three, especially important totemic affiliation, and figures symbolizing the totem(s) should be depicted upon the lorrgon (hollow burial log) formerly used in the final deposition of bones. These are particularly salient “descent-affiliative” totems. They use, for example, jabunu, the name of a stick formerly set in the ground during rainmaking ceremonies, to denote a particular group of hereditary jayway (rainmakers). The name does not apply to any other patrilines, even those of the same semimoiety which are jointly linked with them to rainmaking sites. Hence the patriline i s the largest group to which an inclusive totemic label is applied, though other patrilines may claim t o be “company” with it with respect to WARRWIYAN. A t least at one site (if not more) associated with the descent-affiliative totems, all or virtually all members of the patriline are represented in the one locality.

mingirringgi and junggayi: “owners” and “managers” The intrinsic totem/land/people relationships of the aborigines are complex. The people and patrilines represented at localities by warrwiyan are collectively termed mingirringgi in Mangarrayi. The nearest English equivalent that has gained currency in the anthropological literature i s “owners.” Speaking most generally, Mangarrayi sometimes say that all persons of a particular semimoiety are mingirringgi of a creator and i t s localities and that those of the darlnyin (MMIMMB) semimoiety are ”company” with them (see note 5 for the Mangarrayi expression). Indeed, there are a number of localities (bearing a single name) through which two major WARRWIYAN passed, crossing each other’s paths, one affiliated with one semimoiety and the other with the linked darlnyin semimoiety. In such cases mingirringgi of both semimoieties are represented at the one locality, usually spatially separated. For example, in one place wijwij (opossum)of bangariny-ngarrijbalansemimoiety and gurrwandan (olive python) of burlany-balyarriny semimoiety passed each other; some mingirringgi of the former are represented on top of a slope about 90 meters from the river bank, and some

154

amerlcan ethnologlst

mingirringgi of the latter immediately on either side of the bank. Although the ownership of localities is generally expressed in terms of entire semimoieties, examination of those along a single WARRWIYAN’s path clarifies which patrilines within the semimoiety are represented. Thus, the mingirringgi relation to land and totems is regulated by an ideology of patrilineal descent. The ownerlmanager dichotomy (see Maddock 1972:36ff.) operates strongly among the Mangarrayi. The term for ”manager,” as in many languages in the area, is junggayi. Specific junggayi relations to land areas and their owners are determined through the mother: a person i s junggayi for his mother and property of her semimoiety (though again, speaking a t the greatest level of generality, the Mangarrayi say that one is junggayi for persons and property of mother’s and the linked semimoiety which may both be referred to as barraj or nanyi). In this way, a gangila person may be said to be junggayi for bangariny, his mother, and ngarrijbalan, the subsection which includes MF and MBC. In ceremony the junggayi organize and manage those who stand in the mingirringgi relation to totems celebrated in the ritual performances. The ownerlmanager dichotomy also has important implications for people’s relations to land areas and their resources in daily life. Those who are mingirringgi are in various ways restricted in their relations to their own land areas. A t their own sites, mingirringgi are not supposed t o gather wood, draw water, or light fires; persons in the junggayi relation to the mingirringgi and the localities should perform these jobs. Mingirringgi may exploit vegetable and flesh resources at their own localities, but only the junggayi may, in various ways, petition directly for bounty at the sites. Here the warrwiyan play an important role. Trees and other objects representing the mingirringgi are present at the sites, and as aborigines approach, anyone may greet the warrwiyan, generally by announcing who has come and his relation to persons represented. The junggayi, however, often say that they have come hungry, that they require a plentiful catch, hunt, or harvest; they flatter warrwiyan by saying that they are barraj-yirrag (unique mother’s property), that is without equal in plenty and potential. The junggayi often stroke or pat trees representing people and may rub them with mud as part of the procedure of asking for plenty and to ensure fertility of the locality. Thus, the warrwiyan serve as a direct link to the potential of localities at which they stand. I have discussed totemic food taboos elsewhere (Merlan 1980). but the observance of food restrictions is particularly relevant to the mingirringgiljunggayi relation. The Mangarrayi are a people who have been referred to in the anthropological literature as practicing subsection totemism (see, for example, Spencer [1914:205-2071 for an early description of what he called “subclass” [ = subsection] totemism among the Mangarrayi and Yangman). Actually, as explained above, each totemic entity is considered to be linked with a semimoiety comprising a father-child subsection pair. Although only certain patrilines within the semimoiety may be mingirringgi with respect to the particular totemic entity, all persons within the semimoiety category are expected to observe food restrictions with respect to edible species of their semimoiety. This usually involves (in the case of animals and fish) prohibitions against eating only certain parts of the animal. For example, people of gangila-jarnijin are prohibited from eating goanna tail, the part that i s considered a special delicacy. but they may eat other parts. Members of the owning semimoiety are usually enjoined from eating large exemplars of their own vegetable species, but they may eat small(er1ones. Any person, however, may hunt or gather any species. This means that people of each semimoiety have a long l i s t of things with respect to which they must observe food restrictions. If individuals are caught in an infraction of food restrictions, they must pay their junggayi. Formerly, payment was largely in edible foods but also in nonutilitarian and symbolic items such as locks of hair. Today, tobacco, money, cloth, and some other items of manufacture are also given. People say that when hunger forced some-

Manganayl environment and symbolization

155

one to infringe a prohibition, he could later pay his managers and explain what had happened. WARRWlYAN in their travels often distributed edible species at their own localities; some of these species belong to the same semimoiety as the WARRWIYAN-by association, as it were. For example, jab (wild wind) of balyarriny-burlany semimoiety distributed yurrulanyan (white currant [Securinega melanthesoides]) at a number of localities; this species is also of balyarriny-burlany semimoiety because it was brought by jab to certain localities. Owing t o its abundance at these jab places, jab mingirringgi find themselves restricted with respect to a major seasonal resource at several of their own localities; but food restrictions are partial (for most flesh foods) or relative (for plant foods). Inevitably, warrwiyan in the form of trees fall down, die, or are destroyed by fire. Those who stand in the junggayi relation t o localities and their owners are responsible for keeping an eye on the state of repair of natural resources and warrwiyan. Mingirringgi cannot carry out conservation measures in their own places; the junggayi are supposed to be aware of what needs to be done, and to do it, and the mingirringgi are supposed to compensate them for their work. For example, if a soakage becomes blocked with debris, one or more junggay; should clear it. Mangarrayi consider it improper and even dangerous for mingirringgi to do this. If a warrwiyan falls or i s destroyed, the person represented (if living) is responsible for paying his (one or more) junggayi, who in theory should have been aware of the situation and informed him of it. If the warrwiyan represents no living individual, one or more other members of the patriline must arrange to pay the junggayi. From accounts of payments made, it seems that it i s common for payment t o be made to only one close junggay;, and rarely to more than three. The junggayi in turn may distribute some of the payment to others. Many payments I recorded were to actual MBCs. Rather than using the kin term for such relatives, Mangarrayi often refer to them as wuwari, which connotes "guardian, protector." I visited various sites where warrwiyan had fallen, and payments either had been made or were s t i l l being demanded or negotiated, but I surmised that the process of replacement (i.e., the presumed eventual transfer of the name to another tree or object) is not immediate and probably requires subtle discussion and reaching of tacit agreement before being advanced as public knowledge. During World War I I , when an army camp existed near the town of Mataranka, trees were destroyed over a fairly large area. Aborigines told me that payments exchanged hands for a long time afterwards (but they say that army personnel were never told this and never knew of the situation). Now aborigines say that nobody is represented there, though the site is s t i l l associated with i t s WARRWIYAN, and mingirringgi are represented in other localities associated with the WARRWIYA N. It i s clear that important themes underlying the ownerlmanager relation are protection and interdependence, economic and otherwise. The theme of separation i s also important: owners who stand in intrinsic relation to localities and totemic species avoid direct contact with them in certain ways, their relations to them mediated by managers. Since Mangarrayi localities extend into more arid areas away from major water bodies, there are places that have no obvious natural objects suitable for representation of people. These localities are, however, associated with particular WARRWIYAN. I visited several such places with aborigines. For example, Garlan is in open, flat country watered only by seasonal rains. This place, created by plains kangaroo on his journey, has as its focus a low outcrop of soft, easily fractured reddish stone, said to be mirnday (leeches). Plains kangaroo was supposed to leave file snakes (considered good eating) there, but in a change of mood which is the source of considerable humor to the aborigines, he left a pile of leeches instead. There are no trees at Carlan and no people are represented there. Plains kangaroo continued to a large lagoon nearby where many people are represented.

156

amerlcan ethnologist

All the sites without warrwiyan that I visited were of this sort: the track of the WARRWlVAN at other points crosses bodies of water or localities with natural features suitable for representation. In any case, the density of warrwiyan i s greatest along the Roper, its tributaries, creeks, lagoons, billabongs, soakages, and rock-wells, though smaller and less imposing tree species at less well watered places, and rocks (where these are present), do represent people. I visited a site several miles away from the main river at which an imposing rock outcrop runs for several hundred yards. A few trees close t o the site represent people, but various parts of the continuous rock outcrop itself represent more. There may be, however, localities (especially some that are the only places associated with a particular WARRWIVAN) that have no warrwiyan. The major totemic paths follow or cross bodies of water at various points.

the representation of people and descent groups Without examining in detail the localities recorded and the linkages of patrilines that are joint owners of various localities and totems, I hope t o convey some idea of the layout of sites, the distribution of warrwiyan at them, and the reasons for considering the patriline the basic unit in relation to totems and land areas. Figure 3 shows a billabong through which jab (wild wind), affiliated with balyarrinyburlany semimoiety, traveled. The people at this site are represented by trees of three species: gulu [E. camaldulensis], murrinyja ( E . microtheca), and mardabula (Terminalia platyphylla). Number 1 represented a burlany man whose son 2 i s dead. One of 2’s sons is a young boy who has his FF’s name, that of 1. Other children of 2 (by at least two different wives) are 3, 4, 5, 6, and 9 (all daughters), and 7 and 8 (now young men, s t i l l unmarried). Number 10 i s a daughter of 1’s last wife, but I did not ascertain positively (never having met

10) whether she i s an actual daughter of 1 or whether he is simply considered her father. Number 11 i s said to be 1’s f u l l (older) brother. He had a son (not represented at this site, I was told) whose son has 11’s name. Number 1 2 i s a man who came t o Mangarrayi country from a distant group but is not considered 1’s brother (see below for further comment on this man). Number 12’s children (by at least two wives) are 13, 14, 15, and 16 (daughters), and 17, 18, and 19 (sons, though 18’s biological father i s supposed to have been Malay). One of 12’s sons (not represented here, but at another place nearby) fathered two sons, one of whom has the name of his FF 12 and the name of 20; the other son has the names of 21 and 22. The assumption of two names i s rare, but apparently i s neither improper nor at variance with the theory of the link between child and ancestor of the FF class. Two other children by 12 (that I know of) are not represented here. The children of 2 tend to cluster on one side of the billabong, and 12’s children cluster on the other side. The last people represented by numbers 20, 21, and 22 are considered full brothers and are not

* Dv-

,I

0

Figure 3. Diagram of jab (wild wind) locality (balyarrin-burlany semimoiety).

Mangarrayi environment and symbolization

157

identified as Yangman. They were not full brothers of 1 , though they are now called his brothers. Number 23 i s also called Yangman, but the man t o which 23 formerly referred had male descendants, one of whom has a son called by the name of 23. The known genealogical relationships among these people are diagramed in Figure 4, with numbers matching those of Figure 3. The warrwiyan of 21 and 22 are known by appellations of the grandparental necronymic sort. Often, when all other names by which people were known have been forgotten, necronyms remain as the only appellation by which they are remembered and by which the warrwiyan representing them and their SSs are called. Some necronyms are thus incorporated into the system of names by which warrwiyan are called. It i s significant that the last appellation t o fade from memory i s one expressing the link to FF and hence to the descent group. The two brothers who now bear the names of 12, 20,21, and 22 have each assumed two country names instead of just one. One has the name of his actual FF 12, who came from an area several hundred miles away sometime, according to all evidence, in the late 1890s or early 1900s. This actual FF married in the Elsey area, had a number of children as shown, and became incorporated into the local system of landltotem affiliations. He did this possibly by virtue of his subsection category in his place of origin and/or by virtue of some of his original totemic affiliations, but also of apparent importance was his friendship with the man 1 of the Elsey area. The newcomer 12 was treated and eventually regarded as brother of 1 , and 12's totemic affiliations with respect to several major WARRWlYAN overlap with those of 1 , as evidenced by the representation of both 12 and 1 and their children in various localities. Often people use the imagery of conception filiation in speaking of 12's arrival in Mangarrayi country: he emerged from a certain place (to which 1 was totemically affiliated). People also sometimes simply say that he was brought by the totemic figure of the place whence he emerged in Mangarrayi country, and that footprints on a rock at that place are tracks that he left as he came with the totemic creator. (The interpretation of markings on rocks as footprints of people as they emerged or were brought along is a common form of environmental imagery.) The second name of the first brother was borne by a Yangman man 20, long since deceased. No very precise genealogy could be obtained for him; he seems to have had no male descendants, but he was not an actual brother of 1 . The second brother bears the name of two Yangman men of his FF class, now also long

I I.

I 1

13.

14,

IS.

Aburfony

burfony

- --

A A 3.

Al.

16.

17.

18.

5.

4.

6.

A 7. I A

8.

bur/ony IS.

60fyorriny 12. 20.

A

A 21. 2 2 .

burfony Figure 4. Genealogical relations of figure 2

158

amerlcan ethnologl8t

A

9.

-q

6

10.

dead. For one of them, 22, the recoverable genealogy indicates only female children; for the other, 21, no children could be named. There are indications that the three Yangman men 20, 21, and 22 may have been actual brothers (at least they are now so considered). If so, the names of 20, 21, and 22, and their title to this locality and others, are effectively incorporated into that of the t w o young men, 12’s actual SSs. Genealogical knowledge tends t o be fairly shallow; usually the grandparental generation i s the most distant to which anything approaching accurate or complete genealogical relationships can be traced, including all the complexities of polygamy and children engendered as the result of casual liaisons. I t is not easy to define patriline clearly, as the complexities recounted here show. A general process of incorporating patrilines into others over time seems t o occur, rather than outward ramification into a set of patrilines that consider themselves linked in a kind of clanlike structure, tracing their descent t o a common apical ancestor. This kind of incorporation of patrilines into one may be attributable t o the various factors that resulted in the population decrease earlier in this century. Diseases, perhaps infertility due t o venereal disease, and changes in diet may be partly responsible for the apparent scarcity of direct descendants of 20, 21, and 22. (Another possibility, of course, i s that direct descendants have, for one reason or another, been discounted, but if so they have been effectively “lost.”) Also, in precontact times the possibility always existed of some patrilines expanding and others decreasing, but i t i s probably reasonable t o think that the establishment of pastoral stations affected the aboriginal population drastically. Yangman people living in the Mangarrayi community at Jembere fully understand and participate in the representational system. At unique localities of WARRWIYAN, the minimal landholding group is composed of members of a single patriline. Such a locality, identified as Yangman, is a large billabong in flat, grassy terrain within Yangman country. The place was created by dederran (masked plover [Vanellus miles]), a species that commonly inhabits wet grasslands and mudflats and i s totemically associated with gamarraburrala semimoiety. There are also other totemic associations (e.g., with a small plains kangaroo), of the same semimoiety. Only one patriline of the semimoiety is recognized as mingirringgi of the locality. The man who showed me his patriline’s site is labeled ego (I) in Figure 5, and other genealogical designations are in relation t o him (see also Figure 6). All people at this site are represented by murrinyia (€. microtheca). This man is about sixty, and his F, FEs, and 6s represented here are all dead. Number 4, ego’s actual father, had one other son that I know of who is not represented here, for reasons not made clear (there are other available trees at this site). This son is, however, represented near his father at other sites associated with other WARRWIYAN. Ego and his next-older brother 3 had the same mother; ego’s older brother 2 was by a different mother. None of ego‘s brothers i s said to have any surviving sons. At least two o f ego’s sisters (8 and 9) have surviving children represented at land areas of their fathers’ patrilines. Ego’s two sons have now assumed the names of ego’s F and FE-; both of them are s t i l l relatively small children. His patriline i s joint owner with several others of localities along several important tracks that run across Mangarrayi country and into the territories of other language groups to the east and south. At other localities, members of this patriline are identified as Yangman. Several brief expeditions into Yangman country with a few individuals showed that people are represented at sites in the Mangarrayi fashion, at least in the northern Yangman areas closest t o the Mangarrayi. Yangman, and also people of some of the other language groups (e.g., Alawa, Jawony), are represented at sites throughout Mangarrayi country. Also represented, and referred to by language-group labels, are some people from more distant, and lesser-known, places who married into the Mangarrayi community. Although this sub-

Mangarrayi environment and symbolization

159

5. FB* gomorr

6. FB-, Sgomorro

7. Z burrolo

8.

-F2.B amall atrram

Blilobong, known oa r r l o t i v r l y plocr f o r

good

aomr turtle aprclea,

not totrma o t gonorro- burrok. -

burrolo

I. A ego burrolo

burrolo

Figure 5. Diagram of dederran (masked plover) locality (garnarra-burrala semimoiety).

ject requires separate and extended discussion, non-Mangarrayi people so representedwho come from geographically nearer and better-known groups often continue to be identified as such-for example, as Alawa, or Yangman. (Such people may also be labeled ”partMangarrayi” depending on the marriages made.) Those, however, who came as single individuals from more distant, lesser-known countries and groups are themselves (or their warrwiyan) identified by their original language-group affiliation (e.g., Ngalakan, Jawony, Ritharrngu, or Calgurdungu, which seems to be used as a generic term for Queensland-area aborigines). Their children, however, if they remain (and marry) in the Mangarrayi area, are generally called Mangarrayi for purposes of present and positive identification (though more precise family histories can of course be given). People from distant groups have married into the Mangarrayi community over the past few decades, and the adult offspring of these marriages consider themselves, and in turn are considered, Mangarrayi. Various patrilines jointly represented at totemic localities are united in their common affiliation to the totem, which serves as an outward sign of their unity. Although united with other patrilines in joint ownership of wanggij (child) and garawi (plains kangaroo) sites, the patriline at the masked plover site regards itself as the sole owning group of that locality, but is not regarded as one of the owning patrilines of certain rainmaking sites, which belong to other patrilines of the same semimoiety. Thus it seems we must regard the patriline as the basic landholding unit, allied with various others with respect to certain totemic properties and affiliations. Intensive study of the representational system might help to reveal patterns in the change in structure (e.g., fission, amalgamation) of patrilines over time. In the examples I have given of warrwiyan in Mangarrayi localities, the progeny of persons in the relevant patrilines have issued from “correct” marriages in subsection terms (see Figure 2), so the question of the subsection assignment of children born of “wrong” marriages has not arisen. As discussed above, the strong patrilineal ideology that regulates affiliation to land and totems would suggest that children’s subsection affiliation must be adjusted where necessary so that they can be within the father’s semimoiety. The three examples of adjustment that follow are typical. First, a woman of burrala subsection (who should have married a ngarrijbalan man as first choice) irregularly made a first union with a man (not a local) of bangariny subsection. The one child born of this union was assigned to the same subsection as the father, bangariny. In other words, in this case the child, by being assigned to the correct subsection in the matricycle, was thereby assigned also to the father’s semimoiety, though to the same subsection as the father. In the second case, a woman of gamarra subsection irregularly married a man of jamijin subsection, that is, an older man who was her classificatory MMB, whereas she should have

160

amerlcan ethnologlst

s+

s-

Figure 6. Genealogical relations of figure 5

married a balyarriny man as first choice. The two children of this union are assigned t o gangila subsection, thus having been assigned to the other subsection within the father’s semimoiety. If they had been assigned to the putative correct subsection within the matricycle, they would have been burlany, thus not in the father’s semimoiety. In the third case, a burrala woman had a child by a balyarriny man to whom she was classificatory ZC; correctly, she should have married a ngarrijbalan man. The child of this union was assigned to burlany subsection, as it would have been had the man’s marriage been correct. Had the child ”followed the mother,” it would have been bangariny subsection, but i t s assignment t o burlany places it in the father’s semimoiety. This method of placing the child in the father’s semimoiety by “throwing away” the mother i s the most frequent kind of adjustment that I recorded. Note that these adjustments show concern with placing the child in the father’s semimoiety specifically, not simply within what would elsewhere correspond to the father’s patrimoiety.

semiotics of the representational system The Mangarrayi landscape i s ”personalized” in a way that has not been reported for other Australian groups. The vast array of natural objects constitutes a network of signs,” conceived of as continuous and inalterable, which have as their denotata the people of successive generational cycles. The signs denote people by virtue of the indissoluble link between totems, land, and people, and this link constitutes the meaning of the signs, which in the Mangarrayi view cannot be further reduced. People do not exist without a place in a continuous line of descent affiliated with totems and their localities. The assuming of a country name fixes this position in social structure and in the system of environmental imagery that exists parallel to it. When a name by which warrwiyan are known i s mentioned, informants try to be helpful by explaining that it is “only country.” I do not think this implies any disparagement of the link between people and locality. The Mangarrayi seem to be trying t o convey the idea that the name has meaning in the context of associations of people as members of corporate groups to country and totems and is not exclusive to the individual who bears or bore the name. The more private name seems to be more evocative of the individual personality, and mention of this name is usually followed by exclamations of shame or pity, whether the bearer is alive or dead. The importance of locality to Australian aborigines i s immense. The social personalities of people, like the totems themselves, are not conceptualized as independent of locality. Just as the totemic activity and presence are always in the context of locality, from which

Mangarrayl envlronment and symbolizatlon

161

the locality itself derives significance, so people as members of patrilines are localized, and the objects by which they are represented are significant as the visible evidence of their ties to locality and totems. The Mangarrayi have adapted features of the environment to the expression by visible signs of the familiar Australian conceptual link between people, locality, and totems. I t seems needlessly literal to ask whether trees and other objects “are” people: both people and the objects are linked with locality by the WARRWlYAN that placed them in the social order. The objects are treated like people in certain ways. Easily perceptible natural objects are used by the Mangarrayi as a system of expression conventionally correlated with a certain content, the interconnected cultural ideas described above. This conventional correlation of expression with content establishes a cultural order in which it is meaningful for a member of the society to talk about links between trees (or other objects) and persons, even to the extent of saying that a certain tree (or any number of trees, and other objects) “is” a certain person. Nothing belongs to the sign system that i s not culturally considered to represent an expression/content correlation of the requisite kind. Thus there are many available trees that do not (at least at any given moment) belong to the system. Each tree and plant species is affiliated with a particular semimoiety, and all members of the species are the totem without respect to time or place. For instance, all gilirrlfxcoecaria parvifolia) belong to gangila-jarnijin semimoiety because lawurr (egret) of gangila-jarnijin smashed jarna (Long Tom), a fish of bangariny-ngarrijbalan, against a gilirr tree at a certain locality affiliated with gangila-jarnijin semimoiety. People may, however, be represented by tree species not affiliated with their semimoiety; thus, objects as signs of something else are freed from the associations that they have as abstract totemic symbols. Also, only certain trees of any species represent people; these are inevitably older ones that have not grown visibly during a single human lifetime. The use of only older trees to represent people i s probably consistent with the aborigines’ view that people represent a recycling of what has always been; representation by immature trees would be too much at variance with what i s known of physical growth. There i s thus an equation of maturity of the signs with absence of change in the sign system. The sight of trees that have fallen does not seem t o make owners sad: the link that is the meaning of the sign is continuous. News of a fallen tree usually occasions discussion of indebtedness to the junggayi, and the interpretation of people’s outward actions should possibly be that the junggayi are to be compensated for having lost something that partly functions to express dependence of mingirringgi upon the junggayi. I think, then, we may speak of the landscape as “personalized” but not ”individualized.” The links that form the meaning of representation by natural objects are primarily those deriving from the corporate and unifying aspects of social personality: the relation of people to descent groups and of descent groups to land and totems.

implications of the representational system The representation of people by warrwiyan serves in various ways to validate and regulate relationships of people to land, to totems, and to each other. Some of the ways in which the sign system is used are detailed below. 1. The warrwiyan are an important part of the cognitive map that people have of localities. When describing localities, people often refer to the distribution and internal organization of warrwiyan. They may say that all women are upriver at a site, the men downriver. Sometimes they mention relations between members of alternate generations, as fathers on one side of a rock outcrop and children on the other-even, all the Mangarrayi downriver at one locality and the Yangman upriver toward the collective Yangman

162

amerlcan ethnologlrt

area. These descriptions correspond in the main, if not in detail, t o what is present at the sites. The distribution of warrwiyan i s yet another important aspect of locality that the Mangarrayi pride themselves on knowing intimately, along with information about resources, configuration of the country, totemic mythology, and the history of their own and others’ associations with places. 2. The warrwiyan provide a means for junggayi to establish direct contact with the potential of sites by petitioning those represented for bounty.

3. The warrwiyan constitute a partial system of genealogical representation. Often some significant spatial relation i s discernible among members of the same patriline, especially at sites jointly held by several patrilines. Although not all localities exhibit neat structuring of this sort, genealogical relationships are obviously a factor in the placement of warrwiyan. There is even a hint that spatial arrangement may reflect social assessment of people. At one locality all members of a patriline except one are arranged around a billabong and represented by trees. The exception is a man represented by a rock who stands at some distance from the others. This man has spent much of his adult life living away from the Mangarrayi community. His claims to ritual knowledge and power exceed those of most members of the Mangarrayi community. Indeed, they credit him at times with special powers, while at others they dismiss his claims as just so much talk. His placement by himself may reflect some assessment of his status as special for one reason or another. 4. The representational system serves as a means of symbolizing and validating the incorporation of people from other places into the system of land and totemic affiliation. The imagery of conception filiation, or at least of having been brought by the WARRWIYAN, i s often applied t o those who have married in. One man who came to Elsey as a young stockman from Jawony country worked a certain area and knew it well. He eventually married a local woman, and now people say that he emerged from a black cockatoo site in Mangarravi territory. His son’s conception totem i s also black cockatoo. I also learned of a woman living in a nearby Roper community who would assume, during ceremony, the name and warrwiyan of a dead woman who was of an important owning patriline with respect t o a secret-sacred mid-Roper Valley ceremony, but whose identity n o one else had assumed.

5. Finally, though the ideology of the representational system is one of immutability, occasionally it happens that a warrwiyan is reported t o be in a place where it was not known to have been before. One morning it was said that a warrwiyan representing a certain man had been found at a locality associated with an important ceremony. This locality was created by a WARRWIYAN of bangariny-ngarrijbalan semimoiety, and the man now represented there, most unusually, is not of this semimoiety but of the linked semimoiety balyarriny-burlany within the same patrimoiety. This amounts to an advancement of a claim t o affiliation with the totem of the site. The people who reported the news are the man’s close patrilineal relatives; once the news was announced, no further discussion of the matter occurred, at least in my hearing. I surmissed that time was required for an assessment of the situation. If the claim is publicly accepted, people may eventually say that the warrwiyan has always been there. It i s common cultural knowledge that a WARRWIYAN, in i t s travels, has moved certain warrwiyan from their original localities, often allegedly t o place them near particular patrilineal relatives. Mangarrayi believe that such movements have occurred in the “everywhen” dimension, and the idea that men could see them i s considered ridiculous. Most of the movements I heard of do not substantially alter the relationships of patrilines t o totems since the move is t o another locality associated with the same WARRWIYAN. But the movement recounted above i s a more radical change. Accounts about these move-

Mangarrayi environment and symbolization

163

ments seem to constitute an extension of the totemic mythology in which men are intimately involved. The idea that the warrwiyan may move provides a basis for asserting that their movements t o new localities support previously unrecognized claims to land areas and totems. It i s common enough for aborigines to use environmental features as visible signs of the activities of totemic creators, as the Mangarrayi do. The difference is that the Mangarrayi also use them to represent the remembered and living people in the vast system of representation I have described. It would not be easy to develop the Mangarrayi system in drier environments, but one might expect it to flourish in the riverine country elsewhere in Arnhem Land and to the south. Perhaps disruption of land/people ties since European contact has destroyed the system in some areas or affected people’s ability or inclination to reveal it to outsiders. Clearly the remaining Yangman, as well as the Mangarrayi, conceive of environmental configurations in this way, though I believe it likely that the Yangman have adopted this kind of representation from the Mangarrayi. It is not, as Stanner (1965:237)comments, remarkable that Australian hunters and gatherers developed a zoomorphic and phytomorphic imagery. The signs in the highly developed and personalized representational system described here, like the creator WARRWlYAN themselves at another level, provide an explicit means for expressing the continuity of men’s relationships to land, totems, and the constituted social order. The fact that the signs are visible and are manipulated provides some insight into the ways by which the need for expression of change in the cultural system is accommodated and change brought about.

notes Acknowledgments. Work with the Mangarrayi was made possible by a linguistics grant from the Australian Institute of Aboriginal Studies, Canberra. I am grateful to Clyde Dunlop for botanical identifications. I would like to thank James Urry for helpful comments on a first draft of this paper and especially Marie Reay for extensive editorial comments. This paper is dedicated to the late Professor W. E. H. Stanner. Application for the original Elsey Station lease was made in 1877, though the property was not declared stocked until 1882. Neighboring Hodgson Downs was established in 1884. These include murrinyja (€. microtheca), a coolibah; rnanyal (Barringtonia acutangula), freshwater mangrove; garlayarr (Melaleuca leucadendron), a paperbark; jalmbalmbuj (Cathormion urnbellaturn), ”bean tree”; gulu (E. carnaldulensis); rnanarliny (Ficus racernosa), cluster fig; garrwag (€. papuana), ghost gum; gamulumulu (L ysiph yllum cunninghami); mardabula (Terminalia platyph ylla); gilirr (Excoecaria parvifolia); gurnja (Strychnos lucida), ”poison tree”; jijwirrij (Nauclea coadunata), Leichhardt tree; and wangarr (Pandanus aquaticus), water pandanus, among others. One anonymous reader of this paper remarked that it is common in aboriginal languages for the same term to be applied to a “natural species” and everything associated with it or made of it, and that the relation that I have attempted to convey by WARRWlYAN versus warrwiyan is just another case of this kind and need not be specially indicated. I therefore wish to make clear why I retain the uppercase versus lowercase device. It is frequently true that the same aboriginal word is used for a “natural” entity as well as for things used that are made of it. For example, Mangarrayi damayi (fire) i s also used to mean firewood. However, there are a number of distinct terms, such as manyrnany (kindling), which show that while fire and firewood may be partly identified lexically, they are not coextensive conceptually. Thus we must remain alert to different senses of the same phonological word, and I have attempted to distinguish two senses of one word by an orthographic advice. Further, the sense distinction expressed by uppercase versus lowercase is not that of ”natural species” versus thing made of it. Uppercase is used in the sense of “totemic creator figure,” and while such figures may loosely be considered prototypes of species (though I must say I have never heard this idea so directly expressed by any aborigine), a clear conceptual distinction is made between “totemic creator” and the existence of instances of that species in the world. (See Merlan [1980]for discussion of lexical distinctions relating to totemic figures and ”natural species” beyond those given here.) These two concepts must be distinguished from a third, written here as warrwiyan, the visible signs left by totemic creator figures. ‘While the semimoiety is the category in terms of which ritual-totemic affiliations are spoken of,

’

164

american ethnologist

the correctness of marriages is spoken of, at the most general level, in terms of subsection affiliations. One subsection is designated as "correct" for another; for example, balyarriny should marry gamarra. This is the subsection that contains (from male ego's point of view) those of the kin type MMBDD and also ZSD. Wow!, verb particle "to be company, allied"; ja- third-person nonpast conjugation marker; -wurla3PI subject pronominal; -wa- auxiliary (visit); -n present tense. The people of the semimoieties corresponding to Yirrija patrimoiety are owners of the Yabudurruwa and managers of the Cunabibi; those of the semimoieties corresponding to Duwa patrimoiety are owners of the Cunabibi and managers of the Yabudurruwa. The mid-Roper Valley secret-sacred Cirrijingarra ceremony, in which the community participates, is owned only by people of bangariny-ngarrijbaln semimoiety; the linked balyarriny-burlany semimoiety is spoken of as darlnyin (see below), and sometimes this, as well as the other two semimoieties, is spoken of as junggayi (managers) with respect to the ceremony and i t s owners. A homophonous verb particle is used to mean "to dream creatively," that is, by dreaming to bring something into being, as rain; the ordinary verb that designates dreaming during sleep also contains this morpheme, -barrany+bu- (where j-ny before the stop of the inseparable auxiliary bu). The Mangarrayi kinship terminology, of the Aranda type, distinguishes four lines of descent, each line within a different semimoiety and ego within the same semimoiety (and same subsection) as FF and FFZ. The Pidgin English gloss "shade" requires some explanation. Those who have extensive knowledge of matters described here are mostly older people who speak a form of English closer t o the pidgin end of a pidgin-creole continuum in the Roper area. The more pidgin (as opposed to creolized) forms of English do not contain the word "shadow." Instead, the single word "shade" is used for "shadow" and "shade" in two senses, that of comparative darkness caused by interception of light, as well as shelter built to cause that effect (also, "bough-shade" i s used in this latter sense). The Mangarrayi word warlayjnyin is also polysemous, used for "shade" in both senses given above (though other terms also exist for various kinds of constructed shelters) and for "shadow." I assume Pidgin English "shade" for warlayjnyin in this context would best be rendered by Standard English "shadow." See below for the use of warlayjnyin as a synonym of ni (name). The Mangarrayi, like most aborigines, observe a period during which the various names of the deceased are not to be mentioned. It seems that depending on circumstances, this period of silence may be as little as three years (and today, sometimes less). The name of a woman was first mentioned in my hearing about a year after her death, when ceremonial purification of close kin had been completed. The period now held to be desirable for the purification by ritual washing is about one year. I got the impression that in the past the naming ritual described was sometimes the occasion for the removing of the speech ban on a name. I cannot confirm this from firsthand evidence since no naming rituals of this kind occurred during my stay. l o Note that the dyadic stem is reduced by haplology compared with referential form murimuri and required additional suffix -wa followed by -yi, which is the morphological exponent of the dyadic notion when used with kin terms; elsewhere it is a derivational suffix that expresses "having X" where X is a noun. l 1 As reported for many other Australian groups, conception i s sometimes heralded by a large or unusual find of some plant, animal, or other resource. The object found is not necessarily affiliated with the child's semimoiety. l 2 This is an anthropological phrase and not one that glosses any single Mangarrayi word or phrase.

'

'

Here I refer the reader to the collected papers of C. S. Peirce (1931-58; citation is by volume and section number). Peirce's (2.228) (simplest) definition of the sign is "something which stands to somebody for something in some respect or capacity." (But see, e g , Peirce 1.339 and 5.484 for some elaboration of the sign and the essential elements of semiosis.) The signs described here are able to "stand for" something because a system of expression (the natural objects) is placed in correlation with cultural content. As Peirce (5.287) notes, a representative function arises "if I take all the things which have certain qualities and physically connect them with another series of things, each to each, they become fit to be signs." See also Saussure (1966:lllff.) for the concept of valeur (value) arising from the systemic correlation of (linguistic) expression with content.

references cited Heath, I . C., and F. Merlan in press Dyadic Kinship Terms. In Oceania Linguistic Monographs. Sydney: Oceania Publications. Maddock, K. 1972 The Australian Aborigines: A Portrait of Their Society. London: Allen Lane, Penguin.

Mangarrayi environment and symbolization

165

Merlan. F. 1978 Making People Quiet in the Pastoral North: Reminiscences of Elsey Station. Aboriginal History 2(1):70-106. 1980 Mangarrayi Semimoiety Totemism. Oceania 51(2):81-97. Peirce, C. 5. 1931-58 Collected Papers. Cambridge: Harvard University Press. Reay, M. 1962 Subsections at Borroloola. Oceania 33(2):90-115. Saussure, F. de 1966 Course in General Linguistics. C. Bally and A. Sechehaye, eds. W. Baskin, transl. and introduction. New York: McGraw-Hill. Shapiro, W. 1979 Social Organization in Aboriginal Australia. Canberra: Australian National University Press. Spencer, B. 1914 Native Tribes of the Northern Territory of Australia. London: Macmillan. Stanner, W. E. H. 1956 The Dreaming. In Australian Signpost: An Anthology for the Canberra Fellowship of Australian Writers. T. A. G. Hungerford, ed. pp, 51-65. Melbourne: F. W. Cheshire. 1965 Religion, Totemism and Symbolism. In Aboriginal Man in Australia. R. M. Berndt and C. H. Berndt, eds. pp. 175-237. Sydney: Angus and Robertson. 1979 Report on Field Work in North-Central and North Australia, 1934-35. Microfiche No. 1. Canberra: Australian Institute of Aboriginal Studies. Strehlow, T. G. H. 1978[1964] Central Australian Religion: Personal Monototemism in a Polytotemic Community. Adelaide: Australian Association for the Study of Religions.

Submitted 16 September 1980 Revised version received 23 February 1981 Accepted 19 March 1981

166

amerlcan ethnologlst