Descent among the Wayú. Concepts and social meanings

Journal de la société des américanistes 94-1 | 2008 tome 94, n° 1 Descent among the Wayú. Concepts and social meanings Alessandro Mancuso Electronic version URL: http://journals.openedition.org/jsa/9143 DOI: 10.4000/jsa.9143 ISSN: 1957-7842 Publisher Société des américanistes Printed version Date of publication: 15 July 2008 Number of pages: 99-126 ISSN: 0037-9174 Electronic reference Alessandro Mancuso, « Descent among the Wayú. Concepts and social meanings », Journal de la société des américanistes [Online], 94-1 | 2008, Online since 10 June 2013, connection on 30 April 2019. URL : http://journals.openedition.org/jsa/9143 ; DOI : 10.4000/jsa.9143 © Société des Américanistes DESCENT AMONG THE WAYÛ. CONCEPTS AND S OCIAL MEANINGS Alessandro MANCUSO * Ta king thecontemporary rethinking of the descent notion in Lowland South Amcrican ethnography as a starting point, the a rticle provides an analysis of matrilineal desccnt among the Wayù. Using new ethnographica l data, special attention is paid to indigcno us concepts a nd to the way matrilinca l descent art iculates with other principles of social classifica tio n. By virtue of the rolc that matrilineal descent plays in defiuing tcrritoriality and in feuds, the Wayù offer a very interesting case for ret hink ing the theoretical a nd comparative debate about the iudigenous societies of Lowland South America a nd fo r rellecting on the complexity of the interactions between structure and history in this a rca. [Key words: Wayù lndians, desccnt a nd kinship, Lowland South America.) La filiation chez les Waytl. Notions indigènes et significations sociales. Reconsidérant le débat contemporain sur la notion de filiation ( descent) dans l'ethnographie des Basses Terres sud-américaines, l'a rticle fait une a nalyse de la filiation matrilinéaire chez les \Vayù et de son articula tion avec les autres principes indigènes de classification sociale. Les concepts wayù de la filia tion matrilinéaire sont présentés en tena nt compte de nouvelles données ethnographiques. Une attention particulière est portée sur le rô le de la filiation matrilinéa ire dans la définition de la territorialité et dans les vengeances. Pa r leurs caractérist iques, les Wayù offre nt un cas très intéressant pour le débat théorique et comparatif sur les sociétés indigènes des Basses Terres sud-américaines et pour la réflexion sur la complexité des interactions ent re la structure et l'histoire dans cette aire. [Mots-clés: Jndiens \Vayù, descendance et parenté, Basses Terres sud-américaines.] Concept os y significados sociales de la descendencia entre los 1raytl. Retomando el debate contemporâneo acerca del concepto de descendencia en la etnogra fia de las Tierras Bajas sura mericanas, cl art iculo presenta un anâlisis de la mat rilineali<lad entre los wayù y de la for ma en que ésta se a rticula con los otros principios de clasificaci6 n social. A través de los datos et nogrùficos aqui presentados, se propone una nucva interpretaci6u de las catcgorias indigenas de matrilinea lidad , que es aplicada al repla nteamiento • U niversity of Palerrno/University of Rome« La Sapienza », Via Franco Faccio 2, Palcrmo 90144 [[email protected]]. Jo11mal de la Société drs i1111érica11istrs, 2008, 94- 1, pp. 99- 126. © Société des Américanistes. 99 JOURNAL Dll LA SOCIÉTÉ DES AMÉRJCANISTES Vol. 94- 1, 2008 de la cuesti6n de la rclaci6n entre desccndencia matrilineal, tcrritorialidad y vcnganza en esta sociedad indigcna . Por dichos rasgos, los wayù brindan un caso cuya consideraci6n resulta ùtil tanto para el desarrollo futuro del debatc te6rico y comparativo accrca de las socicdadcs indigenas de las Tierras Bajas suramcrica nas, como para el cstudio de las intcrncciones entre estructu ra e historia en esta area. [Palabras claves : wayù, descendencia y parentcsco, Tierras Bajas suramericanas.] INTllODUCTION As it is well kn own, during the mid-Seventies the development of Lowland South America (from hereon LSA) ethnology was strongly m arked by a rejection of the so-called « descent »or « lineage theory ». Critics argued that LSA indigenous societies «are structured in tenns of the symbolic idioms (names, essences, etc.) that relate to the construction of the persan and the fabrication of the body» (Rivière 1993, p. 510). That is to say that in this area « kinship » is not a matter related to « group » constitution and «corporation », but to « corporality ». As far a s LSA Amerindians areconcerned with principles and categories of social classification, these do not take th e « reified » form of « g roup », but a more symbolic appearance. Furthermore, modes and processes of exchange and incorporation take the place of legal statuses phrased through an idiom of ownership and of rights and dulies (Seeger et al. 1979). However, it is worth noting that what Rivière (1993) called the« Amerindianization of descent and affinity », has so far resultcd in much more new theorizing about « affinity » than about « desceut ». lt was the former notion, once disaggregated from its confinement to the kinship context, that has become crucial for understanding al/ features of al/ LSA indigenous models of social relationship. Viveiros de Castro's «grand unified theory » of Amazonian sociality is one of the most recent developments of this trend. Here, it is argued that in Amazonia « potential » affinity, intended as a cosmological and ontological « generic value», must be considered « the generic given, the virtual background out of which a particula rized figure of consanguineally dominated kinship sociality must be made to appear » (Viveiros de Castro 2001 , p. 26). According to this view, descent, equated with « consanguinity », is simply seen as the last stage of a dynamic process of extraction/construction of de-differentiated identities/ collectives from a cosmological background of « potential affinity ». Even if Viveiros de Castro is clear in stating that « the idea of affinity as a dominant principle » (ibid. , p. 22) particularly suits for explaining the modes of sociality proper to those societics with alliance-based local groups, he thinks that « the situation does not change much when we consider those Amazonian regimes that fcature village or dcscent group exogamy » (ibid., p. 24) 1• 100 Mancuso OESCENT A l\IONG THE \V A YÛ However, this theoretical view leaves open the question of explaining how and in which conditions affinity seems to give way to descent relatedness. Looking at this problem, Hornborg (1988, p. 238) makes the hypothesis that, in the patterns of social orga nization in iudigcnous LSA, « unilocal residence is a li kely point of o rigin for various... codificat ions of unilateral affiliation ... of which descent groups a re merely sporadic expressions ». In a later work, he develops the a rgument: uniliueality [...] seems an aspect of supraloca l integration. Tt is when the tangible boundaries set by local group endogamy dissolve that cultural construed, classificatory boundaries gain in importance. Whereas the Dravidian kin-aftlne dichotomy is egocenlric, « interna i » lo the loca l group, and temporally transient (applicable only to the three med ia] generations), unilineality is a soeiocentric reification of kin-aftlne boundaries, const ruing marriage as« externat ». (Hornborg 1998, p. 178) Jf we follow H ornborg, the issue of how development of « unilateral affiliation » and« descent » is linked with changes in native notions of territoriality and history becomes a relevant one. From this point of view, it is well known how the Lévi-Straussian no tion of « House » has been considered in last yea rs a good starting point for rethinking the issues of« descent »,«corporation »and group identity's historical consciousness in some indigenous societies of this area. Particularly, both Hugh-Jones (1995) and Lea (1992; 1995; 2001) conclude that bot h among the Tukano a nd the Mêbengokre a t least, cultural representations of descent a rc not o nly a n important principle of articulation of spatial a nd tempora l relationships, but thcy also concern modes of owuing and transmitting symbolic items and prerogatives, which are associated with sharing a like-soul component and belonging to the sa me« House » 2 . Conunenting on these works, Rivière (1993) suggested that, o nce we admit the possibility of dissociating the notion of « corporate group » (as a « moral person » who owns al lcast some components defi ning the statuses of its members), from unilineal descent (as a sufficient and necessary criterion for membership), it becomes easy to considcr the« Ha nses» of Mêbengokre, as described by Lea, as kinds of« corporations» 3 . Nevertheless, the way Lca conceives the relationships between « descent » and «corporation » in Mêbengokre « houses » sccms to go against Viveiros de Castro's a rgument (2001, pp. 33-34, with notes) that throughout the region such components and processes involve « potentia l affi nity » as their unique precondition, whereas « substantial identifications», associated with kinship (and « descent »), only figure as a« consequence ». From a nother point of view, it can be askcd if the focus on the frequeut embeddedness of descent categorics in native « House » idioms means that there is no other room for use of the first ones in looking at LSA models of sociality. Last but not least, I wish to point to the existence of some important conceptual diflèrences between the use of 101 JOURNAL DE LA SOC IÉTÉ DES AMÉRICANISTES Vol. 94- 1, 2008 « descent » in recent LSA ethnology and the way of redefining this no tio n in some influent attempts to rethink it after the definitive dismissal of« Lineage Theory ». ln faet, a lso today it is genera lly assumed that a minimal standa rd definitio n of descent to maintain is « a series of filiatio n links repeated generati on after generation » (Parkin 1997, p. 15) 4 • By tlùs definition, a pcrson is desccnded from another if a nd only if, miuimally, hc or she is a child's child of the ot her. The minimal « couuno n descent » relation is the genca logical relation bctween two persons who have an anccstor (minimally, a parent's parent), in common . (Scheft1er 2001, p. 15) But, if the word « descent » is to be meant in this sense, it becomes evident that a set composed of a person, his (ber) children and grandchildren, cannot be considered a set of people rclated by conunon descent. Also for this reason, both Verdo n (1980; 199 1) a nd Scheftler (200 1) have questioned the use of terrns like « descent » a nd « lineality » with reference to rules both of group membership and of acquisition and/or transmission of statuses and assets. For Verdon, the term « descent » sho uld be used only when talking about one of the possible « elements of aggregatiou » amo ng «simple» (that is composed of « individuals ») groups into a composite (that is conformed by « groups ») group. D evelo ping some of Scheftler's more ancien! ideas (1966; 1973), he suggests talking of « g ro ups » for desig nating those sets of individuals that are defined by o ne o r more criteria of membership or entitlement to join together in a specific activit y. The notion of « group » sho uld be analytically separated from both «corporation »or« moral person »(as a set of individuals that is defined by one o r more criteria of membership or entitlement in order to demarcate these individuals' conunon ownership of an« estate »), a nd from «social category » (the simple native recognition of membership criteria for pertaining to a distinct set of people). According to Verdon's terminology, when we are talking of a genealogical criterion for belonging to a« group »,a« corporation »or a« social category »of individuals, it would then be better to refer to it as agnatic, uterine or cognatic « kinship », as distinguished in turn from the simple criterion of « filiation ». As a consequence, « descent canuot apply to problems of succession. The same conclusion also applies to inheritance, in which individuals are selected, but groups are no t aggregated »(Verdon 1980, p. 146) Scheftler's mo re recent view (2001) is similar. He argues that there is little o r no sense in talking, for example, of a« matrilineal »or « patrilineal » descent mie of group aftiliation or of tra nsmission of stat uses and assets. For Scheftler, in a il such cases, we should rather ta lk of rules of patri- o r m atri- filiation, and main tain the term « descent group » only for those groups in which unifiliatio n is both the necessary and suflicient condition fo r inclusion. On the cont rary, when descent is o nly a necessary or sufficient membership criterio n it would be improper to talk of « descent groups » because these groups are no t only or not 102 DESCENT AMONG THE W A YÙ Mancuso always defined exclusively by descent. Furthermore, when non unilineal filiation is the necessary and sufticient condition for membership, we cannot speak of « groups », but of« socially meaningless categories, for there cou Id be no rights and dulies entailed by inclusion in them »(ibid. , p. 31). Taking these theoretical stances in mind and going back to LSA area, it can be noted that at least the idea that clescent is one of the criteria by which some components clefining affiliation to « Houses » are transmitted, turns problematic. In what follows, I shall try to cliscuss some of the issues sketched above whilst analyzing how forms of clescent relateclness are involved in the models of social organization of the Wayù population. In fact, 1 wish to sustain that the Wayù case highlights the neecl to build a bridge between current ethnographical and comparative theorizing on LSA indigenous moclels of sociality, and contemporary theoretical attempts to systematically rethink the issues which « descent theory » claimed to answer. THE W A YÛ AND THElR SOCIALlTY Among LSA indigenous peoples, the Wayù 5 (known also as Guajiros), an Arawakan language speaking people living in the Guajira peninsula 6 , at the northern extreme of South America, have been considcred (Wilbert 1970; Jackson 1975; Picon 1983; Descola 2001) qui te peculiar for their precocious adoption of cattle-raising since the first centuries of Spanish colonization. Also Wayù's contemporary demographic dimensions made them eccentric when compared with the other indigenous groups of the LSA cultural area. According to the bi-national census of 1993, about 300,000 Wayù people live between Colombia and Venezuela. In spite of the massive migration to rapidly expanding urban centres surrounding the Guajira (among which Maracaibo is the most important) from the first half of the 2oth Century, a considerable part of them still live in the peninsula's semi-arid enviromnent. Ecological conditions only permit very limited forms of seasonal agriculture, though not in ail the region. Hunting and gathering forms of subsistence have long since lost importance because of growth of human and livestock population. In coastal areas, fishing is practised, but it is equated by Wayù with poverty (Guerra 1990), and, for this reason, opposed to cattle raising, which, in spi te of its persistent crisis from the first half of the 2oth Century, it is still a very important subsistence activity for the majority of the indigenous population living in the peninsula. From the colonial period, both the adoption of livestock raising and the Wayù's historical involvement in the commerce and contraband networks between the coast and the insular Caribbean seem to have caused the 103 Vol. 94- 1, 2008 JOURNAL DE LA SOCIÉTÉ DES AMÉRICANISTES developmen t of some features of social organiza tio n which a re uncommon a mo ng LSA indigeno us peoples. In fact, as no ted by Descola (200 1, p. 110), whereas th roug ho ut the indigenous LSA area, « the principle of substituting o bjects fo r persons is conspicuously absent », the Wayù are o ne of the two exceptio na l cases (th e Mapuche being the o ther o ne) where we meet « bridewealth »and compensation payments to resolve disputes, including th ose arising over homicide 7 . In spi te of this, feuds ( !ttkawa) are still widespread in the peninsul a to day, especially in its inner areas, where State cont ro l is very limited . As shown by Saler ( 1985), among the Wayù the a mo unt of negotiated go ods (livestock, which was in the past the more relevant item, is now increasingly substituted by cash) is, both in bridewealth and in dispute payments, hig hJy va riable, depending o n the parties' socia l status (oj11t11 «value»), and in turn redefining it. Through the display of materi al wealth in these socia l transactions between persons a nd groups, a strong emphasis 011 hiera rchica l differences is put o n ail the ma in fields of social relations. But at the same time, reciprocit y o bligatio ns hold a very import ant place. In fac t, large networks of people, mai nly but not always exclusively related thro ugh a kinshi p tie, participatc both in contributing to and benefiting from bridewcalth and dispute payment, with the implicit mut ual understandin g that the o ne who brings «collaboratio n » ( 011111111•a11•a) to day will receive some contributio n tomorrow, when asking il in return. = T IIE EllUW KU NOTION In mythical na rrations about the o rigin of Wayù society, the cultural hero M areiwa subdivides the first Wayù into several sets of people and calls what a re now some wild animais to assigna name to each gro uping, so that they can have a clear way to distinguish between themselves. After firs t efforts fa il because of Ma reiwa's rejection of some gro tesque proposais of name attribut ion, fina lly o ne bird, in almost ail versio ns the bird U tta (Hypnelus bicinatus o r Hyp11el11s mfico llis, called in local Spanish «Pico gordo »), gives a proper na me to each set. From that mo ment, that name will permit a Wayù to identify his/her set as his/her eirmku 8 . T he literai mea ning of this term is « A.esh », and, more generally, « substance», « texture», « compo undness » (Jusay u a nd Olza Z ubiri 1988, p. 88). When asked about the procreative process, Wayù people say that the wom an's cont ributio n is « stro nger » (katsiiinka) in providing the child's eirmk u. T he reason is that o nc's person eirmku is essentially a product of her menstru al blood's condensati on which takes place after contact with man's semen (awasain) . T his contact makes menstrua l blood ( asli!t) turn into pulped flesh ( aslmla) , as milk to which curdled mil k is added turns into cheese. 104 Mancuso DESCENT Al\IONG TH E \V A YU Sometimes, this theory of the procreative process is indicated as th e basis for the fac t that the eirruku na me a person acqu ires at birth is that of his/her mother. But more frequently, Wayù people say sim ply that only women « mul tiply », « enlarge » ( all'iti11111aja) one eirr11k11. H owever, the myt hical accoun t of the o ri gin of Wayù social o rder hig hlights how the tenn eirr11k11 is a lso currently used fo r referring to the set of a il those people who bear a same collective na me. Most versions o f the myth mentio n between 20 a nd 30 na mes, corresponding to the number of «clan » names reported by the most recent binational census. From the most ancien! reports, which go back to the second half of r9th Century, the names men tio ned are generally, with few exceptions, the same ones. F urthermore, il must be noted that some names (Uliana, Epieyu, Jpua na, Pushaina, Epinayu, etc.) are borne by o ne o r more tho usand person, while other o nes are borne only by a few dozen people. M ythical narrations a lso tell how a specific area, located in the Upper G uajira, and an iconi c sign (ayawase, literally «identificatio n », also calledjeerii, « iron ») to be used as a brancl fo r its members' cal lie 9 , were assigned by Mareiwa to each named eirrnku. Some versions a lso mentio n one or more a nimal species which came to be « associated » with the members of each group 10 • However, it must be stressed thal, a lso in these mythical na rrat io ns, the na mes of the eirrnku gro ups generally do not have a nything to do wit h the na mes of the animal species o r of the site associatecl with each of them. ln fact, a linguistic ana lysis of the probable etymology of eirr11k11 names, which is possible for most of them 11 shows that only in a fcw cases (significa ntly th ose of the names borne by few people), this etymo logy corresponds to the name of an animal species o r a place. Rather, in the majority of cases, it seem s to rcfer to a behavioural characteristic. In any case, Wayù toclay generally do not att ribute to these na mes any mea ning o ther than that of« pro per names » designat ing the eirrnku groups. But, o n the other ha nd, though ever ybody agrees that nowadays bearing o ne of the more common eirrnku na mes is in no way sullicient fo r being recognized of high status, people usually say that in the past, and to a lesser ex tent still now, some eirrnku na mes are a ssociated with the prestigious econo mic o r military conditio n of their bearers, while o ther ones a rc, for the same reasons, « despisecl » a nd « shameful » for those who bcar them. Often a correlatio n is established between the Jess diffused eirrnk 11 names a nd a« poor » status. Furtherm o re, several cases of people who, in a relatively recent past, « changed » ( awanaja) their own « shameful » eirr11k11 name with a no ther mo re common and prestigious o ne (often aclopted from the gro up o n which they were econo mically o r military dependent) were reported to me. Besicles that, in some cases we meet with groups who are sa id to bear, besides a n eirmku name shared with o ther groups, a second eirmku name propcr to them. M ost of the limes, this fea ture is explained as a n instance of eirmku grou ps that , though d istinguished by thcir ancestral origin (a point we will to return latcr), are 105 JOURNAL DE LA SOCIÉTÉ DES AMÉRICANISTES paatawasu, that is « fta nked », « paired » Vol. 94-1 , 2008 12 . Tho ugh this expression (which can be used also speaking of the relationship between eil'l'uku groupings that apparently have a totally distinct name, for instance Wouli yu and Uliyu, lpuana and Sapuana) explicitly rcfc rs to a privilcged condition of close friendly relationship (« as th ough they were brothers »), it is actually charged with ambiguity because it is frequently considered to mask an asymmetrical subordinate relationship of domestic, working or milita ry service, expressed through an idiom of« nurturing and sustenance » ( epija). In fact, at o ther times these cases of bea ring another proper eirmku name besicles the shared one, are interpreted as a sign of the hierarchical difference of status between the proper « owners » of the shared name, who are« valuablc » ( koj11ts/1i), a nd the othcr people, who are considered a111o'ju/a, that is « lacking », « defected ». THE APÜS/11 NOTlON Ali cthnographers (Watson 1967; Wilbert 1970; Goulet 1981; Saler 1988) agreed on how both the native theory o f the bodily constitution of the person and the emphasis on the subdivision by eirruku in the representations of ideal social ordering are congruent with the prominence that utcrine kinship has in many of the most important spheres of Wayù social identity and practice. Particula rly, Goulet (1981) argues that a strict connection exists betwcen Wayù kinship categories and native theo ries of procreation. He found that kasa a11ai11, « something related to », is the Wayù term for« relative », which applies to anyone linked to a person by genealogical relatio nships of consanguinity o r afiinit y. Among their kasa a11ai11, people disting uish between their apiishi and their 011p(ly11. According to Goulet, the proper meaning of people bcing apiishi is that of sha ring the eirmk u, that is to be relatives« through the flesh ». So, apiishi cornes to desig nate every uterine kin . The father and agnatic kin a re considered relatives« by blood », referring to the assim ilation of the father's contribution to procreatio n (his semen, as said above) to a« ma rked » form of « blood », but they are no t grouped under a specific kin term. Finally, a perso n's 011p(ly11 are his/her father's apiishi, that is his/her father's uterine kin (Goulet 198 1, pp. 163-164) 13 . One problem with Goulel's interpretalion of « ulerine kin , gcnealogically related throug h the eirmku »as being the« intrinsic » meaning of apiishi, cornes from the possibility, in some contexts, of using this tenn for referring also to people who are not genealogically related, o r, though being so, are not « through the flesh ». Goulet himself noticed thal apiislii can rcfer to people who bear the same eirmku namc, even when (as we shall sec soon) it is stressecl they are not genealogically related. He argues that such cases could represent an« extensio n » of the term's « prima ry » meaning. Nevertheless, this kind of explanation does no t account for cases in which apiishi also refers to people in no way related 106 D ESCENT AMONG THE \V A YÙ Ma ncuso « thro ugh the eirmk11 ». So the term ca n be used fo r exa mple, to refe r to all of a pcrson's relatives (often including the father o r a man's children) involved in organiziug his/her burial ceremo ny (Goulet 198 1, pp. 240, 243). Ali thcse aspects of the tcrm 's use bring little sup po rt to the thesis that the « primary » meaning of apiislii is « to be gcnealogically relatcd through the ftesh »; to interpret these uses, in fact Goulet himself ( 198 1, p. 170) turned to a different order of cxpla natio n, which looks at the generic mea ning of apiislii: « member of a category »or« part of a whole » 14 . This is, significantly, the first meaning of the term found in the Wayuunaiki-Spanish dictioua ry of Jusay u and Olza Z ubiri (1988, p. 51 ), where it precedes the o ther meanings given, that are: « member of a clan »; « uterine relative »; « relative in general ». Looking at this evidcnce, it seems to me that, if we want to poin t to an « intrinsic » meaning of the word apiishi, it is rather « Io be lied, linked, connected ». This conclusio n is suppo rted by linguistic analysis too, according to which the word apiishi is formed fro m the root apii, « lace, string fo r binding »(sec the glossary reported in Guerra 2001 , p. 37). This gencric meaning of the term sta nds in good accordance with its current Spanish translation, « familia, familiar », fro m bilingua l peo ple, and with the possibility of graduating the « intensit y » of being apiishi through the frequent adding of qualifying markers, as in the expressions apiislii mai (« very much »), ap1ïslii pejejat («close»), apiislii mwi11je (« attached, clung »), or apiislii ll'attajat (« far, dista nt ») 15 . This docs no t cont radict the fact that in the first instance Wayù identify their apiishi as their uterine kin. As we will see, what they are po inting to by saying this, is that these people, pa rticularly those with who m they share associatio n with a« terri tor y» (ll'o11111ai11 « our land ») o r j oiu in a feud, are those wh om o ne is mo re strongly « bound to, part of ». = TH E CONTROVERSV AilOUT T HE « DESCENT » ORIENTAT ION OF WAYÛ KINSHIP SYSTEM = Wayù kinship terminology rcsembles a Crow type, with FZCh P and MBC h = C h, and a set of specific terms for a ffines 16 . This appears to be congruent with the emphasis o n uterine kinship, but it is well known how Héritier (198 1) rejects the idea of a neccssary connectio n, instead arg uing that a frequent association exists between C row-Omaha terminologies, an emerging cognatic cha racter of kinship gro upings, and semi-complcx forms of rnarriage a lliance. Undo ubtedly, such a thcory permits to account for some important features of Wayù mod cls of kinship and alliance. For example, besicles the absence of any positive marriage mie, stated by ail cthnographers (Watson 1967; G o ulet 1981 ; Saler 1988), 1 found it is considered « good » to marry « far » ( wattasii) wi th someo ne who is a 11atajat, th at is someone who is no t considered onc's own 107 JOURNAL DP. LA SOCIÉTÉ DES AMÉRICANISTES Vol. 94-1, 2008 relative. In fact, such marriage is saie! to make possible Io ex tend one's own social and po litical networks. On the contrary, to marry with relatives results in pat1111ajirrasii (« to hug one another »)and apajirrasii (« to grasp one a nother »). Nevertheless, Héritier's notion of semi-complex systems of alliance cloes not seem Io fit well for interpreting other fcatures of Wayù social moclels. In fact , beyoncl paren ts and chilclren, explicit prohibitions only apply to closes! uteri ne relatives (inclucling matrilatera l parallel cousins, M ZDD, and o ne man sister's cla ug hter). Besicles that, both Goulet ( 1981), Saler (1988) and l too collectecl a consistent proportion of marriages among co usins (including, though rarcly, matrilateral parallel first ones), which a re always ex plainecl throug h refcrencc to circumstantial factors, ranging from strategies of maintaining livestock and territorial presencc concentratecl, to love feelings or lack of other partners. The relationship between the cognatic and the matrilineal aspects of Wayù society is in a certain sense involved too in the very debated question of the kiuship (« ego-focusecl »)or descent (« ancestor focused »)orientation of Wayù concepts of social ordcring by gcnealogica l relatio nships. In the course of this debate (Watson 1967; Goulet 198 1; Saler 1988), which was phrased accorcling to Schefller's (1973) ana lytical distinctio ns and developed before the demise of classic « lineage theory » 17, there was also a constant refcrence to the question of if and when some groups of uterine kin cou Id be seen as« lineages », which actas « corporatio ns». Starting from the first issue, if we put together what is said in mythical narrations about the origin of eirrulw names and groupings and in Wayù theory of procreation and na mes transmissio n, it couic! seem that bea ring the same eirr11k11 name entails the sharing of the same matrilineal ancestry. Moreover, people who bear the same eirruk11 name address each o ther (and, in some contexts, refer to each o ther) using kinship terms even if they cannot indicate how they are genealogically related 18 • For this reasou , almost ail ethnographers since Simons (1885) concluclecl that bearing the same eirruk11 name is what identifies Wayù « mat riclans », defined as groups of « putative» uterine descent fro m the same a ncestors. This view has becn suppo rtecl by Watson (1967), who provided an a nalysis of Wayù social structure acco rding to the conceptual fra mework of classical « lineage theory ». According to him, Wayù matriclans, called eirr11k11 in wayuunaiki and "castes" in Spanish, would actually be dispersed and not « corporate », since territorial concentratio n and corporateness, particularly in feuds, are associated with more "restricted " uterinc descent groupings, which couic! be ana lytica lly termecl as « lineagcs » (see also Wilberl 1970; Saler 1988). Moreover, these matriclans do not correspond to an exogamie unit, which is found only al the « lineage » level of inclusio n. Nevertheless, as Goulet ( 1981) noticed, although thcy considcr the acquisition of the eirr11k11 name by matrifiliatio n to be a consequence of the mothcr 108 Mancuso DESCP.NT AMONG nm WAYÛ alone providing one's « flesh », the Wayù do not believe that ail those people who bear that na me share a common matrilineal ancestry. 1ndeed, among such people it is common to find persons who do no t consider themselves to be genealogically related at ail 19. Furthermore, Goulet (1981 , p. 167) argues that apiishi is an egocentrical kinship category. ln his view, when Wayù speak of the apüshi who join together in a feud , what is significant for !hem is no t these people's sharing a mat riliueal ancestry, but their close uterine genealogical relatedness with the victim o r the aggressor as well as sharing the same « territory » (Goulet 198 1, pp. 223-224). Analogously, Wayù definitio n of territo riality by reference to sets of uterine relatives, implies that it is only by « matrifiliation » that a perso n cornes to share a rightful claim of restricting access to it (ibid. , pp. 129, 135, 167) 20 . For these reasons, relying o n Scheftler's distinctions, Goulet (198 1, pp. 39, 139, 141) concludes that Wayù do no t have any kind of desccnt categories or gro ups, but o nly « kinship » o nes. Saler, who did his field research in the same yea rs as Goulet, disag rees with him about this point. He argues (Saler 1988, pp. 78-87) that the recognition of a principle of m at rilineal descent is implicd not o nly by myt hical accounts of eirmku o rigin, but a lso by current identification of a group of apiishi thro ug h reference to the territorial origin of its ancesto rs. However, he does not provide an interpretation of the meaning of native kinship terms and concepts alternat ive to the one put forward by G o ulet, preferring to focus o n the issue of at what extent it is correct to consider the g ro ups of apiis/1i as « lineages ». Saler's conclusio n is that it is better to say that a variable« approximati o n » of the fo rmer to the latter exists, wh ich depends on the specific group and the context under consideration (greater « approxi mation » occurring when a « descent ideo logy » is « o perating » in jo ining uterine relatives duri11g fe uds). So, he finally comes to admit that in a lot o f cases, the model of ego-centred kinship appears to interpret the nature of the social networks of uterine kin better than « descent ». His wholc a rgument is thus left with a certain ambi guity, insofa r Saler sometimes seems to refer to the « native's point of view », whi le ot her limes he is clear in stating that the primacy of ego-centred kinshi p definitcly holds o nly when we are dealing with the ethnographer's observation of real social practice. Saler's approach to the issue of « dcscent » in Wayù social o rganization ultimately incurs Verdo n's criticism ( 1980) of the uncertain theo retical ground by which ail « classica l » theories of desccnt view « desccnt gro ups » as sets characterized by « ontological va riabilit y »and different « degrees o f g roupness » 2 1• Besides that, it !caves unsolved the q uestio n of how native no tio ns as eirruk u a nd apiishi are linked with « descent » and o ther principles of socia l classificatio n. 109 JOURNAL DE LA socrÉTÉ Vol. 94-1, 2008 DES AMÉRICANISTES ANOTH ER VŒW ABOUT WAVÛ CATEGORfES OF « OESCENT »: MATHILlNEAL ANCES'rRV, SEGi\IENTATION AND THE PLACE OF HlSTOIHCAL MEi\IORY It seems to me that most of the problcms left unsolved in the controversy about Wayù principles of relatedness can actually be overcome not only by adopting an analytical view of the native concepts of « descent » which goes beyond the tenets of classical « lincage theo ry », but also taking into account some terms and statements which have bcen overl ooked or not at ail becn reported in Wayù et hnography. To begin with, though the Wayù do no t bclieve that bearing the sa me eirmku name nccessarily involvcs sharing the same matrilineal ancestry, it is wrong to deduce from this fact that they do n' t group people o n this basis. What Wayù say, rather, is that common uterine d cscent can be claimed only if, as well as this name, people share the sa me ekf, a term whose mcaning is « origi n », but also « head » or « base». As Saler already suggested, this « origin » is usua lly idcntified through the refercnce to the namc of the site whcre their first utcrine female a ncestors (oushii « gra ndmother ») of the group arc thought to have« emerged » ( ojuita), « risen out» ( ell'eta) from the undergro und. Indeed , many people go so far to say that they belong to the same eirmku only if their « origin » is the same, and what happens is rather that distinct eirmku share the same name. ln this scnse, they particularly point to the beforc mentioned cases of gro ups which share a common name, but are distinct for another collective name which only o ne of them is associated with. There are some Wayù notions of group segmentatio n which show the instable relationship between pertai ning to an eirr11k11, bearing the sa me collective na me and descending from the same « o rigin » 22 . Also when it is no t stated that o nly those who share the same o rigin arc of the same eirr11k11, people usually refer to those bearing the same eirmk u name, but who a re « of a <liftèrent origin » ( katata11•as11 shiki), as belonging to o ne of the« many divisio ns » ( suliijalepala) of that eirruku. Aliijale, the tenu used in this context, mea ns « divisio n, departmcnt, partition» (Jusayu and Olza Z ubiri 198 1, p. 89) of something 23 , but this subdivision is not represented as a result of a generative proccss. On the contrary, this idea may be prescnt when the Way ù speak of the distinct «segments» (shiipa) of an eirruku. When this occurs, what is implied is not o nly that people, though sharing the same origin, bclong to matrilineal lines traced from differcnt « grandmo thers », but a lso that the members of these lines act separatcly when someone is involved in a fe ud . Shiipa actually mea ns «segment, part, piece of somcthing »(Goulet 198 1, pp. 170- 17 1) but also « coordinated a nd simultaneous actio n », « the continuation o r horizon of something » (Jusayu 1977, p. 403; Olza Z ubiri a nd Jusayu 1978, pp. 350-35 1; Jusayu and Olza Zu biri 1988, pp. 89-90) . This term th us seems to imply a dimension of continuity as well = 110 Mancuso DESCENT AMONG THE \VA YÛ as coordin at ion, shared by the segment's members, which is absent in the meaning of alùja/e, « partition ». A similar situation is found when rega rdiug the relationships between « sharing the same o rigin » and having the same brand sign. Wayù aftirm that brand ma rks of people who bear the same eirrnku name, but who are of a diffe rent o rigin, look totally different. Neverheless, they admit that differences, also significant ones, in the fonn of these sig ns ofteu exist within subgroups of people who share matrilineal ancestry. ln these cases, these differences are explained as the result of subsequent modifications (eirrata) brought to the « rea l » (shi11111i11) brand sign of their eirrnku. These modifications, consisting in adding, prolonging or curving one or more of the composing lines (which Wayù referas a« head »or « leg » or « arm », of the « original » sign), a re said to occur when a group of utcrine relatives go to live far from other members of their eirm k u, or when they want to escape from their enemies dming a feud 24 . ln fact, when a serious dispute arises, to have one's own cattle brandecl with the same ma rk as that of o ne's « enemy » (mï'1111111va), is sutricient for being identified with them and so beconùng a potential target of retaliation 25 . Furthermore, people refer to their common uterine ancestry by poiutiug to the publicly recognized terminological relationship of the siblinghood of some of their respective identifiable uterine « grandmothers » (that is uterine female ancestors of two or mo re ascending generations, most conunonly no mo re than four from an adult ego). People who are connected in such way are called po11shiiwas11, which means: « their uterine grandmo thers are joined », or pawa/as1ï 11011s/11ï, which mea ns: « their gra ndmo thers are in a real o r terrninological relationship of siblinghood ( awala) » 26 • It must be noted that, according to Wayù tenninology, the term awafa, « sibling », generically applies no t only to full o r uterine brothers aud sisters and matrilateral parallel cousins, but also to agnatic semi-siblings and patrilateral parallel cousins (even if there is a specific term, asa1111a, to designate specifically these kin). For this reason, it could seem that the range of the possible uses of awafa makes the daim two people have their « grandmothers joined »as siblings, a dubious way to ascertain common matriliueal ancestry among them. But for the Wayù, this remains a minor source of ambiguity, insofa r as they always reduce it consistently throug h connecting the terminological relationship of siblinghood between their grandmothers with the previously mentioned identity of their common territorial origin ( eki), or, at least, « provenience », in historical reconstructio ns. T he involvement of these women's uterine descendants in the same past fe uds often constitutcs the main subject of these historical memories. The constant reference to the specific histo ry of a group of uterine descendants from the same « origin » provides an importa nt element to rethink the issue of how the ra nge of uterine kin who join in a feud is dcfined. In fact, such an issue has been al the centre of ethnography regarding Wayù social orga nization, 111 JOURNAL DE LA SOCIÉTfl DES AMÉRICANISTES Vol. 94-1 , 2008 often intersecting with the question of whether and in what way Wayù have « lineagcs ». There is an agreement among ail ethnographers on the fact that, at lcast at the level of the proper native rule, when a feud arises, neither a man 's sons nor the father of either vict im or murderer should be involved, but only their uterine kin (Watson 1967; Goulet 198 1; Saler 1988). Even so, reporting actual cases, people admit that, particularly nowadays, these rules are subject to some margins of variation. So, even non-uterine close kin as well as the kerraii (that is spouses of the fenrnle members) of the uterine group involved may « intervene » (asoukta), if they arc « very affectionate » (ajirrasu). However, these cases are considered exceptions, and these men's uteriue relatives tend to discourage such an intervention, becausc it risks ma king ail of thcm a potential target of revenge 27 . Furthermore, in the case of an homicide, even if a dispute is resolved through a compensation payment, father and oupayu of the victim on ly receive a minor part of the payment (which is called siiwiiirra, « for the tears » or is/1011p1111a, « for the blood shed»), while most of the amount (t he part which is significantly called «for the eirmku »or siijutu, «for the value») is due to his uterine kin 28 . The main point of ethnographical controvcrsy has been about the range of uterine kin who join in the course of a feud . Bo th Watson ( 1967) and, in a <liftèrent way, Saler (1988) deal with this subject by looking at the genealogical depth of such groups, concluding that it varies according the degree of economical and political coordination under a « chief ». As seen before, Goulet, for whom it is wrong to considcr these groups as based on « descent », maintains that involvement in a feud is defined only by the actual close uteri ne genealogica l relatedness with the first victirn or aggressor. The limit of these uterine networks is in turn detcrmined by the actual sharing of a same territory. The whole issue can be reassessed in better terms once we ta ke into account how the sharing of an ancestral origin ( eki) and of the same uterine historical grandmothers is related with involvement in feuds. From this point of view, people aftinn that those who do not share the same eki as that of a particular persan, are in no way involved with him in his feud 29 . On the contrary, those who share his « origin » may be involved. Nevertheless, in the majority of cases on which 1 collected information, a feud involves only those uteri ne kin who are actually associated with the same « territory » and bone cemetery, while it does not involve the other people with whom « grandmothers were joined »,but whosc present territorial association is another one. Whereas on the one hand this seems to confirm, al the level of real practicc, the state of things described by Goulet, on the other hand it must be stressed how this fact is not usually explai ned as the result of a natural tendency « to separatc » (akatajirrasii) which occurs when the descendants from the sa mc « origin » « territorially disperse» ( all'a/akawasii) in the course of time. Rather, in speakiug about this tapie, people point to the history of how in the past one group 11 2 Manc uso DESCENT AMONG THE \V A YÛ « withdrew itself » ( akatalaj111111sii) from a feud in which some one else of their uterine relatives was more directly involved. This withdrawal is sometimes said to have occurred because o f the« fathers » having paid compensati o n to keep their sons out of a fcud o riginated by some of their uterine kin, but more frequently it is explained as the independent decisio n of a group of uterine apiishi. So, people often tell and complain too, of a past feud in which their uterine ancestors had been « united », and of more recent feuds in which the descendants of these ancestors came to act « separately » 30 . lt is probably looking at such features of the relationships between relatedness by uterine ancestry, territoriality and involvement in feuds that Saler (1988, p. 86, my translation) acutely remarks that for Wayù, the political identity of a uteriue descendant group (he says: «of a lineage ») is the« affirmation of a historical particularism » much more than a result of some «structura l principles » of genealogical reckoning and « segmentary oppositio n ». TEni\IS AND i\IETAPHOHS FOR THE « DESCENT PROCESS » Relating people th ro ugh pointing to their commo n ancestral« origi n »and to the siblinghood of their uterine « grandmothers »corresponds mo re properly to what Lea (2001) proposes to call « ascent », instead of« descent ». Regarding these relationships of « ascent », it is interesting to no te how Wayù often talk of the total set of their « grandmothers » and « maternai uncles » (alaiilayu) as siip11/err11a, a term which literally means « those who go forward ».This seems to suggest the ancestors a re Jike people who precede, a re ahead of us, in occupying physica l and social space. Wayù som etimes add that ancestors, once they die, leave behind ( apiita) their living« descendants ». Provided that it is no t correct to interpret, as Goulet did, the « primary meaning » of apiishi as « uterine kin, relatives through the eirmku », it can be asked if the Wayù have specific terms to desig nate a relationship which should be properly interpreted as« uterine descent ». We can start answering this question by noting that there is a specific collective term, aikeyu, which bilingual Wayù commonly translate in Spanish « descendientes, descendencia ». This term is often used in myt hical accounts fo r pointing to the relationship between Juya, « Rain » - a ma le« supernatu ral » figure whose central place in Wayù cosmology was shown by Perrin ( 1976) - a nd the entire Wayù people. In « ordin ary » discourse, a pcrson's aikeyu are ail of his/her aliiin, a term which applies both to his and her proper gra ndchildren (ChCh), as well as to those of his and her siblings, and to thosc of the following gencrations. People say that an ac/1011 - a term that according to the Crow features of Wayù terminology refers not only to o ne's chi Id, but also to a woman's sister's child (wZCh) as well as 011e's materna i uncle's child (MBCh) - , and a ma n's proper asipu (mZCh) 11 3 JOURNAL DE LA SOCIÉTÉ DES AMÉRICANISTES Vol. 94-1, 2008 - are not o nc' s aikey11, because to have o ne's own descent implies a generational continuity of at least two generatio ns. Besicles th at, there is a specific tcrm, 011lill'o11, which is used collectively to designate the uterine descendants of one person. In existing Wayù ethnography we find very few mentions of this term, which nevertheless is reported in the two Wayuunaiki-Spanish dictionaries by Jusayu (who is Wayù) and Olza Z ubiri. ln Jusayu (1977, p. 527) the given mean ing of 011lill'o11 is « children or grandchildren who a re left at death, maternai descendants», white in later Jusayu and Olza Zubiri ( 1988, p. 154) we fine! « woman's uterine grandchildren » 3 1• Even though the two translat ions do not thoroughly coincide, both show that the meaning of the term implies an uterine li nk. The Wayù with whom l have discusscd the issue state lhat 011/i11•011 does not apply either to a woman' s chi Id or to her sister's. T his is congruent with the fact that the term used for these relatives, ac/1011, a lso covers MBCh, who is nota uterinc descendant. Moreover, Wayù say that, as in the case of aikey11, a woman's child cannot be considered her 011/ill'o11, because to have 0111i11'011 implies a more extended temporal continuity of her uterinc descent (ma ny say: of her eirruk11). For these reasons, only « grandmothers » have 011/ill'o11, the first of whom a re her daughters' children. Vicws about who are one man's 011/iwo11 are more swinging. ln fact , some people assert that o nly women have 011lill'o11, as only they « provide » uterine descent, and « multiply »t he eirruk11. On the contrary, other people say thal, as for women, a man's 011/iwo11 are his uterine kin of two o r more subseq uent generations. They often include in this catcgory a man's proper uterine nephews and nieces too. People who support this view cxplain lhat, as far as they « cont inue » his eirruk11, a man's 011/iwo11 are a il the« children » ( acho11 ) of his éiyetse. This lasl term, as reported by Goulet ( 1981, p. 164), refers to a man's entire uterine fema le kin of his same gencralion or one lower (mZ, mM ZD, mZD, etc.). In fact, old Wayù peo ple say that a il these 011/i111011 of a man (mZCh, mMZDCh, mZDCh, etc.) are covered by the sa me kin term, asip11, tho ugh they point to the cunent widespread « bad use » of the aforementioned term aliii11 for designating mZDCh 32 . T he more restrictive interpretation thal only women have 011//111011 could find support in the derivation of the lerm from 011/ia «plant, crop » that is proposcd by a lo t of Wayù people. In fact, an analogy is made between a n eirruk11 and a plant 33 . People often compare a n eirr11k11 to a plant whose «base» (eki, a term, as we have scen, which a lso means « head » and« origin ») is a« grandmot her », white the 011/iwo11, as Wilbert (1970, p. 321) had already reported, are like the shoo ts of the new branches which develop from its stem 34 . At other times the ana logy is pul in a different way. lnsleacl of represenling an eirr11k11 as a single plant whose branches are the 011/iwo11, it is saicl that these latter arc like the procluce of the seeds ( aii11) of a plant's fruits ( ac/1011irr11a) , so origi nating in a new exemplar of lhat piani 35 • 114 Mancuso DESCllNT AMONG TI-fi W AYU However, at other limes people propose an alternative etymological explanation of 011/iwou where its literai m eaning would be « footprints of one's steps », 011/i meaning « foot», « step ».This interpretation leaves more room for adnlitting that men also have 011/iwou - people who closely fo llow them both in tenns of genealogical continuity of thei r eirruku, and, practically, in their spatial movernents and feuds. In both cases the use of these metaphors shows how the idea of a process of uterine descent is closely linked to the meaning of the word 011/iwou. UTERINE KlNSHlP AND TRA NSi\llSSION OF ASSETS Actually, for the Wayù, living near their own uterine close kin is a much so ught-afler ideal, and indeed, as 110ted by Goulet (1981) and Saler ( 1988), it is frequent for a group of uterine siblings and cousins of both sexes to live close to one another in nearby dwellings 36 . However, post-marital residence is highly dependent o n circumstances and can change du ring time. From the moment marriage payments begin, but are not yet totally fulfilled, a man sho uld limit himself to « visiting »the woman nightly in the dwelling where she lives. When payments are completed he can decide whether Io go Io live with her elsewhere or, when he lives in a difièrent village from his wife, whether to bring her to live there. A man who gave « bridewealth » (paii11a) for his wife has the right to receive the same for his daughtcrs, and compensation payments ( awataja) too if one of his children is hurt by another person. If he dies, the same claims ca n be placed by his close uterine kin, who can also take the widow as wife of one of them, though lcviratic marriage ( eisala a11111in) is now on the wane. In a ny case, even if a close uterine kin goes to live far from the other apiishi, closeness is periodically reasserted through frequent visits and seasonal residence in the same place, and definitively restorecl after death. In fact , people feel a strong moral obligatio n to join a relative's remains in the same site where his/her close uterine kin are alrcady buried a nd in whose neighbo urhood some of them are still living. Among the Wayù two burials are made: the first one takes place immediately after the death; the second o ne is o rganized some yea rs later. The dead person's remains are then exhumed and definitely buried in his/her uterine group's « bone cemetery » (jipupa/a) , even if the corpse was buried before in another site (ca lled ash11/apala, « ftesh cemetery »and often traduced by bilinguals as ce111e11terio de paso). This d ynamical and flexible intcrplay between residence patterns, links of uterine kinsllip and the importance of burial sites as markers of group identity, is related with the definition of territoria lit y. There is an agreement between ail ethnog raphers that rightful claims on the land and its resources are acquired by the first to start ongoing exploitation, and are transmitted through uterine kinship ties alone. The presence of a bone cemetery is what permits the close 11 5 JOURN AL DE LA SOCIÉTÉ DES AMTIRICANISTES Vol. 94-1 , 2008 living uterine kin of the people who are buried there to cla im the neighbouring area as their « land »or« tcrritory ». Of course, tlùs set of uterine kin lets other people live in its territory and use the resources within it, but these persons are only tcmpornrily « permitted » to do so, by virtue of having social relationships with one or more of those who are the real owners 37 . ln spi te of ail the ethnograplùcal reports about the strict normative character of the association of territoriality with uterine kinship, until relatively recent times the former seemed to involve restricting access to water sources, good pastureland and small garden plots located near the settlement established by an uterine group, without implying an exclusive ownership of a clearly bounded area. This feat ure has profoundly changed in the course of the 2oth Century. ln fact, population increase (also of non Wayù people) in the peninsula; development of exploi tation of huge minerai resources; acq uisition of tourism potential; urban expansion; and, in the last ten years, the so-called tm11sfere11cias (financial resources destincd by the Colombian State for indigenous people living within a legally protected native area), have ail endowed land with an incrcasing intrinsic economic value. As a conscquence, at least in the Colombian part of the peninsula, an exponential explosion of land disputes among Wayù people has followecl these processes, often becoming the cause of feuds. 1n such a context, the common practice of letting a man's chilclren reside in his uterine kin's « territory » even after his death, according to the will he expressed before clying, has become particularly charged with ambiguit y, as is shown by the fact that these disputes often involve a deacl man's uterine kin on one sicle, and hissons' uterine kin on the other. However, even when taking account of these changes, a nd, more generally, the inAuence of the no11 incligenous society (particularly that involved by inter-ethn ie marriages) in the long run, uterine kinship appears to be more relevant in definition of territoriality than in what concerns familiar transmission of persona! property, whose patterns show a considerable variability 38 . Regarding cattle (which is individ ually owned by both men and women), both Watson (1968) and Saler ( 1988) substantially agree that its transmission is actually towards both uterine and non-uterine kiu, though they differ on points such as: considering transmission to no11-uterine kiu as a by-product of acculturation; distinguishing bctween « formai» and « informai » transmission; and the extent to which transmission patterns are at variance with the amount of wealth transferred. Nevertheless, there is historical evidence (Picon 1983; Barrera 2000), clating from xvmth Century, of a greater prominence in the past of transmission to uterine kin, which is confirmecl by most Wayù people with whom 1 have discussecl the issue. F urthermore, 1 have founcl several cases of feucls which involved on one sicle a man's uterine relatives, and on the other sicle, his sons and their uterine kin, which had arisen because he hacl transferrecl most of his animais Io his sons before his cleath. Of course, that is not to say that in the past animais were given 116 l\'lancuso DESCENT AMONG THE \V A YÛ away exclusively to one's own uterine kin . Rather, the point 1 wish to hig hlight here is that, in a ny case, at least Way ù patterns of inheritance of persona! property can in no way be interpreted through reference to the idea of a « corporate belo nging »of the uterine descent group 39 . CONCLUDING REi\lARKS ln spite of Verdon and Schefller's arguments cited in the introductory sectio n, the Wayù case brings evidence that it can be quite in consonance with at least some LSA indigeno us group's conceptualizations to tenu a set of people sha ring a commo n point of genealogica l origin, and including mo re than two generations, as a descent « category » or« group », inasmuch the pertinent feature of inclusion is considered that of sharing a common origin, wh ile shari ng an « ancestor » (minimally a parent 's parent , as previously defined) might no t be implicated. Besides this, it suggests that what we have to study are the specific meanings that these fo rms of relatedncss assume in particular contexts, instcad of tracing, in whatever way, a d istinction betwcen « groups »and « categories »in which these latter always corne to constitute a residual no tio n whose significance is never fully explained. From this point of view, the issue of how modes of transmission of assets and statuses are va riably correlated with the presence of descent affi liatio n may still be cousidered a theoretically releva nt one. Regarding this last point , I have tried to show how Wayù concepts of uterine descent are config ured and shape their ways of conceiving and practising relatedness without needing to be « phrased » into a Ho use idiom. Among the Way ù, historical memories about territo rial o rigin and spreading of uterine descent groups, as well as about cohesio n a nd divisio ns among their members in past feud s, appcar two important contexts in which clescent relatedness is defincd and « works » through. Besicles this, people makc a constant reference to the bone ccmetery of their uterine descent group to po int both to the histo rical and demographical limits of this group's iclentity and to its« ownership »of the la nd wherein that cemetery is situated. For this reason, as already noted by Gou let (198 1), when the second burial takes place, decisio ns concerning the site where to bury the bo nes of a clead uterine kin, a re an important moment by which both ut erine descent a nd territorial identity are reaffirmed and/o r redefined. The massive prescnce of guests who a re no t uterine kin of the dead person turns burial ceremo nies into a mo ment of public recognition of such idcnt ity in fro nt of the who le Wayù society. At the sa me time, white in feuds g roup affiliatio n by uterine kinship involves (at least idea lly) excl ud ing the relevance of a ll o ther kincls of o ther relatio nships, in bu rial ceremon ies we also fi nd an emphasis on the extendecl networks of social reciproci ty as a necessary conditio n for the uterine group' s perpetuation. For this rcason, Wayù burial ceremo nies mig ht in some way be compared with those 11 7 JOURNAL Oil LA SOCIÉTÉ DES ａｬ｜ ｦｬ ｾ ｉｕ ｃａｎ ｉ ｓ ｔｅｓ＠ Vol. 94- 1, 2008 rituals associated elsewhere with the definition of « House idcntity », as in the case of Gé naming ceremo nics o r of the Tukano complcmentarity between H e and « Foodsgiving » ceremonies (Hugh-Jones 1995). Nevertheless, among the Wayù, what dcfinitely permits to identify a set of uterine kin as a distinct unit it is no t the reference to a« House » associated with some specific elements, symbolic items and prerogatives proper to it, but - evcn mo re than genealogy - the a ncestral and/or actual connection with a specific land and cemetery. It is such connection which cames to be« enactecl » in the ceremonia l as well as in the feud context. As we have seen, with the possible exception of land , a n association of the collective identity of the descent groups with specific belongings proper to each group is instead gcnerally absent or weak, even in what concerns eirruk u and persona! names a nd the use of ma rking o ne's cattle with a brand. Perhaps, a correlation m ig ht exist between this absence of the« House »as an iclio m of the descent gro up identity and hierarchy, and the persistent strength of uterine kinship as exclusive criteria of collective identity in Wayù feuds. On the o ther hand, the dcvelopment, to use Descola's terminology, of a mode of exchange based on« hetero-substitutio n » (Descola 200 1, p. 110), and cxpressed by the social uses of cattle a nd o ther kind of material wealth, cou Id a lso account for this particularity, and explain tao why the reference Io the idea of a« corporate belonging » of the descent group seems to have little relevance even in familiar practices of inheritance of persona! property. For ail these reasons, Wayù may oflè r a very interesting case no t only for rccasting descent in the current et hnographie and comparative theorizing about LSA indigenous models of social ity and rclated ness but a lso for stimulating reflection abo ut the interactions between structure, agency and histo ry in this area. * • Manuscrit reçu en février 2007 et a ccepté pour publication en janvier 2008. NOTF.S l than kfully acknowlcdge the support rcceived from the Wenner-G ren Foundation during the years 2004-2005. l gratcfully thank a il those people who hclped me during my stay in Guajira, limiting myself to mentioning An tonino Colajanni, Wilder G ucrra and Rosa Rcdondo. 1. It is wo rth noting that even Rivière ( 1993, p. 5 14) seems to subscribe to this view, at lcast a s far as he considers that «the existence of third tenn [potential atlinity as a category of social classification) as a mode of artic ulation in a conce ntric dualistic structure is perfectly consistent with the generativc process that we call " desccnt" ». 2. The old questio n of how distinguishing between « phratries »,« clans» and« lineages » uncxpectedly reappcars in thcse ethnographical a nalyses, if only to argue that tbere is little sense in asking Io which lcvel of inclusion nat ive idio ms of« H o use » refer, a lthough al the samc tirne suggesting that there is a good degrcc of corrcspondcncc with the« clan » notion. 11 8 Mancuso DESCENT AMONG THE \V AYU 3. This conclusion is unexpected if one thinks that most carly criticisms of the« African mode! » in LSA ethnology assumed the cquation betwccn « group »and « corporation », thus arguing that iu LSA societics il was quite inappropriate Io talk of « groups » al all (Overing Kaplan 1977; MayburyLewis 1979; Murphy 1979). 4. Filiation in turn is deflned as« the relationship o f child to parent per se» (SchcITTer 200 1, p. 17), though admitting that how titis rclationship is cstablished can vary cross-culturally. l d o not enter into the complex issue, debated, among others, by Schneider ( 1984) and Jngold (200 1), of definitions of fili ation and descent which arc groundcd on principlesof gcnerationdiffercnt from procrcativerclationships, bccause this issue, although stimulating, only marginally regards the Wayù case presentcd here. 5. Tuse here the orthography « Wayù » rather titan « Wayuu », which is more widcsprcad in actual Colombia. \Vayù tcnns rcported in this text have been gcnerally transcribcd under their root fo rrn, but in somccases 1cite a prefixed or suflixed form, which is that most currently used. Two dinèrent systems of orthographical notation for the writing of Wayuunaik i actuallyexist, but hcrc I decided no t Io adopt either. Nc\•ertheless, in some cases 1 have indicated prcsence of a stress by a tilde, the glottal by an apostrophe, and long vowels by their duplication. 6. The peninsula is politically divided bctwcen Colombia and Venezuela. I carried out a total of 24 months o f fl eldwork in the Colombian Guajira bctwcen the years 2000 and 2005. 7. He also suggests a correlation exists betwccn the development of these « modes of exchangc » based on « hetero-substitution », and the historical proccss of these two groups having adoptcd forms o f rclationships with animais bascd on rcaring. S. See the versions publishcd by G utiérrez de Pineda (1963), Paz lpuana (1973), Perrin (1976; 1979), Chacin (2003). 9. Perrin (1986) provides evidcncc which should support the thcsis that Way(1 brand signs wcrc adopted from th ose one uscd by Spanish colouists du ring the 17tb Century. Gucrra (1987, and persona) conununication) howcvcr suggests that the likeness bctwcen Spanish and Wayù brand signs does not exclude the possibility of an indigcnous origin. He points to formai aflinities of Wayù brand signs, which arc also fo und dcpictcd on various rocks of Alta G uajira and used as persona! tattoos, with some Amazonian petrogliphs. The question, which had bcen already raiscd by Lévi-Strauss and Delmont ( 1963), remains open until a dcfinitive date for Guajira rock painting is establishcd. 10. Old Wayù people who have not attended school explain that in ancie nt limes animais were \Vayù, but they assumed their prcsent form when « land changcd », o r for having catcn raw meat. However, it is no t made clcar whether association between a single eirr11k11 and an animal species alrcady existed beforc these latter assumcd their present visible fonn . Narrations concerning the origin of a speciflc association are rare. \Vhen they exist (1 fo und for example some versions concerning the association bctween the dog and the Jayaliyu eirmku) it is told that once a femal e of this spccies (significantly almost always a domcstic one), who had taken on human fo nn , had scxual intercourse with a man of a determined eirm ku, hence the current association of this spccics with that group. However, as already notcd by Simons (1 885), the meaning the contemporary Wayù attribute to the association betwcen eirr11k11 groups and animal spccics is little more than an emblematic one. 11. 1 made the trial and discovered that this etymology is almost always congruent with what is statcd in the version of the myth on the origin of cirrnku published in Spanish by Paz lpuana (1 973). ln this version, which reports 36 eirrnk11 names, the mention of each name is followed by an indication of the charncteristic proper to the people pertaining to the corrcsponding cirruku. To give a few cxamples which concern the most common eirr11k11 names: the namc Epieyu, which is glossed « those who come from their own house », is clearly rclated to kepia, « to have and Io live in a house »; Epinayu, « thosc who pouud ha rd on their road », is actually co mposed fro m the verb epina, « Io pound »; Uliana, « the oncs with thesilent walk », is linkcd with 011/i, «foot, foo tstep »; lipuana « those who live on the rocks» is linked with ipa, « rock »; Pushaina, « the one with the scething bloocl », is form ed from ashft, « blood », in all of these cases being -yu and -na sumxes for the plural. 12. As an exmnple, one can cite the case of the Epieyu, who arc oft en divided in « rea l » Epieyn, Epieyu Woluwoliyu, Shooliyu, Alapainayu, Wunujunaja. Il must also be noted that other timcs, some 119 JOURNAL DE LA SOCIÉTÉ DES M léRICANISTES Vol. 94- 1, 2008 of these second names (Alapaina and Wunujunaja) are rather referred Io groups whose « primary » eirruk11 name is Uliana, suggesling that these cases cannol be interpreted only as a resull of the segmentation of a larger group. Anyway, 1 nevcr met with a concept, as il is pre.sented in the myt h published by Paz lpuana (1973), of a systematic aggregation of ail the eim1k11 (identificd by their na me) in phratries. 13. ln the case of 011p11y11, Goulet (198 1, p. 152) suggests that the main basis for grouping these people into a spceial terminological category lies in the fa ct thal in Wayù model of social organization they share a condition of« potential successors » to some cornponents of the fat hcr's status. 14. The Wayù word for « wholc, ail » mpiishiill'a, scems clcarly to be formed from apiislti. 15. Viveiros de Castro ( 1998) shows how the existence of what he calls «gradients of gcnealogical or sociopolitical distance», is a common feature of many LSA indigcnous kinship terminologies, whieh is particularly relevant for understand ing Amazonian Dravidia n systems. 16. Following l ounsbury's typology ( 1964), \Vayù terminology rcsembles to Crow type 1 «semibifurcated », but it prcsents some atypical featurcs that cannot be analyzed herc. Detailcd descriptions of Wayù lcrminology are provided by \Vilbert (1970), Goulet ( 1981 ) and Saler ( 1988). 17. Allhough published only in 1988, the essay of Saler, wherc he traces the history of the controversy and proposes his proper point of the view on the whole issue, was actually wrillen in 1979. 18. This also occurs, but only in addressing a person, with people whose cim1k11 na me is the same as that of one's own father and grandfather. 19. Not even association with an animal species is considercd relevant to the malter, evcn in those cases of people who bear the sa me eirmk11 na me, but daim to be associatecl with differcnt animais. Ooth Simons (1885) and Perrin (1976) decluctivcly interpreted these latter cases as being the result of a process of subdivision of larger « clans ». 20. This explains how, over the course of lime a« territory » may frcquently come to be associated with a different uterine kin group from that with whom il was associated in the past, which would not be possible if territoriality were dcfincd and acquired by sharing matrilineal ancestry with previous owners. ln fact, as seen before, several conditions may bring people to live in a dinèrent area from that where their uterine ki n have territorial rights. When prcvious claims on this land by a diffeœnt uterine kin group do nol exist or arc already extinguished (as allested by the defi nitive abandonmenl of an alrcadyexistent cemetery), these peoplecan not only establ ish their own claim to il, but also choose this site to bury his/her boncs and thosc of their closer uterine relatives. ln this way, they come Io have a separate « tcrritory » from that of thcir other utcrine relatives. For thcse reasons, e1'cn the aforcmentioncd use of kinship terms in addressing and often also in rcfcrring to someone who bcars the sa me eirr11k11 name, must be simply seen as a forrn of « courtesy», which is not of social rclevance, excepting fcw limited conlexts (for exmuple when ask ing or giving concession for pasturing animais during transhumance migration). 21. Verdon distinguishes three thcoretical models of linking together the descent not ion and the group notion: the jurai, the cultural and the ideological one. ln ail of thcm, « the groups that descent will for m either as a rule of group mcmbership, as a rule of behaviour during a process performed by many individuals, or as an idcology, will conscqucntly be "ontologieally variable"» (Verdon 1980, p. 38). Verdon sces the reduct ion of groups to interpersonal bchaviour as the rool of a theoretica l « malaise» which renders comparati1'e and accurate sociological analysis impossible. For titis reason, he proposes his 011•n « operational » definition of « descenl »and « group », which 1 sketched abovc. 22. This instable relationshi p is probably a resull of the part icular historical and demographical processes which \Vayù population wcnl through during the centuries which fo llowed the Europcans' arrivai. Howcver, this rcmains a conjecture, due to the Jack of relevant information in historical sources. 23. So, .rn/iijelc 11111111iki means: « division of a discourse »; s11/iijelc 111iiclti is << division of a house ». 24. In both situations, an alternative option is the adoption of the brand mark of one's own fathcr's eirr11k11 or of that of the eirmku of people in whose lerritory one is « hosted », but this rcpresents a temporary solution which nevertheless can become definitive when a low slatus group of aptïshi decides Io change ils cirmk11 name too. 120 Mancuso DESCENT AMONG THE WAYÛ 25. This can involve raiding caille as a preventive stralegy to debilitalc the owners, eithcr for inducing them to negotiate a compensation payment rapidly, or as a first step for undcrtakinga revcnge (pasalmra) against them. 26. At this respect, the Wayu arc quite similar Io another matrilineal group o f LSA, the Cancla, who « expla in longhouse gcncalogical relationship, al lcast these days, recko ning from ego " up" and "across" through a live or rccently dead "sisters" (i.e. parallcl cousins) not " up" to and "down" from a common anceslor, link by link » (Crockcr 1979, p. 240). 27. A lternativcly, an utcrine kin group ca n express ils dissociation fro m those membcrs o f it who fi ghl in a fend alongside their non uterine kin, by giving p reventive paymcnl compensation to the enemy uterine group o f thcse. Conversely, if thcsc membcrs are injurcd or killed during the feud, their uterine ki n can daim compensation from the uterine group in whose support they in tervencd . 28. This amount is malerially delivercd Io one elderly prcstigio us man among them who111 they consider as their « chicf » (almïla, a term which also 111ca ns « old person » a nd « maternai uncle », though actually he might not be the victim's p roper MB, lml a more d istant ulerine kin). 29. This excludes clearly defincd «allies » ( e111e'j1111a) who provide rcciproca l suppo rt in lheir respective feuds, or the so-callcd « accompanying people» (a111ajac/1i), who serve a s« soldiers »for the gro u p which lhey are lied lo by prcvious subaltern economic or po litical rclalionships. 30. \Ve can find an exam plc of this trend in the story, fro m the 19th Century till the present tirne, of the fcuds conccrning two groups of ute rinc kin, the Uliana and the Jayaliyu, which is narratcd by Nemesio Montiel in his historical nove! E 'irrukuirra ( Linajes) (Mont ici 2002). The two groups arc respectivcly the ulcrine group and the fat her's utcrine group of the author, who is Wayll. 3 1. In his list of kin terms used fo r addrcss, Goulet ( 198 1, p. 175) reports 011/hrn a s an a lternative term which covers a ll those people who arcconsidcrcd aikeyu, but, in agreement with what il is asserted in Jusayu a nd Olza Z ubiri's dict ionarics, ail peo ple Tworkcd with denicd that the terrn can rcfer to any type o f no n-uterine kin. 32. According Io Goulet ( 198 1, p. 175), «ail c/assifica1ory 111a1ema/ 1111cles mu/ sorara/ 11ephe11-s and nieces eau bealso lermed as gra11dfa1hers and gram/sons». H is informants said that in fact MM B can be termed as both almïla (« maternai uncle » and a/ushi (« grandfather »).and converscly mZDCh can be termed both as asipii, (mZch) and a/iii11, (Ch Ch) (ibid., pp. 182-1 83). 33. The use of vegetal metaphors for reprcsenting kinship relatcdness is rcported fo r other LSA indigenous groups, both whereconcepts of unilinea l dc.scent exisl, as a mong the Canela (Crockcr 1979) o r the Tukano (Hugh-Jones 1979; S. Hugh-Joncs 1995) and where it does not, as amo ng the Achuar (Taylor 2001). 34. Indccd, \Vilbert ( 1970) reports all'ûliaaj1111a, considering it as a sort of connotative term, whose translation wo uld be« the shoot that sprouts a t the base of the stem », for asipii. Probably, this is an erroneous transcription of 011/i1m11. Alte rnativcly, if the derivation of 011/in-011 is from 011/ia, « plant, crop », it may be another substa nt ive for m com poscd from the same root, as in the verb mrii/irra: « to become green again », reportcd by Jusayu and Olza Zubiri ( 1988, p. 76). who also mention the word (ibid., P- 154) 011/e: « firs t fru its to ripcn ». 35. l t is interesting to compa re these statements 1 have collccted with what it is said in the myth o f eirrnku origin published by Paz lpuana ( 1973, p. 197): « the core o f the family shall consist of five mcmbers, closely represcnted by your fi ve fingcrs. Tajapu, the hand, shall rcprcsent the common o rigin of you r lribe {eirmk11}. Soushu tajapu shall correspond to the mate rnai grandmother, reprc.sented by your thumb. Shii tajapu shall correspond to the mo ther, o r your index fingcr. Sii' laii'la tajapu shall correspond to the maternai unck , or your middle fingcr. Siichon tajapu shall correspond to the son, or your ring linger. Siilüin tajapu sha ll correspond to the grand son, or your litt le fin gcr. T hus the intimate circle of your fa mily shall be: the grandmother, the mothcr, the maternai u nclc, the son (who is also the nephew) and the grandson) » (cnglish version in \Vilbert el al. (cds) 1986, p. 107). The termino logical iden lity bctween the terms dcsignating the fivc fingers and the kin terins for « grandmother », « mother »,« maternai uncle »,«son », « grandson », is partially confirmed by Perrin (1982, p. 23). lt must be underl ined how the mcmbers of the« core of the family » mentio ncd in this myth correspond 121 JOURNAL DB LA SOCIÉTÉ DES AMÉRICANISTES Vol. 94- 1, 2008 to a gencalogical depth of five generations. Furthcrmorc, it can be obscrved how the term taslii (F =FB = FllS = FBDS) - which , unlike thoseoncs mentioned, does not refer to any kind of uterine kin - is not included. 36. ln the peninsula, mosl Wayil settlements look quite scattered, with each house sometimes at a distance of scveral hundred meters from the next. A fcw people live in each dwelling, gcncrally related by close links of consanguinity and/or aninity. Surrounding one or a few of such dwellings, there is a kitchen, a polcd nnd roofcd spnce for receiving visitors and holding meetings with more distnnt relatives, and fenced ranches fo r goals, sheep and, whcn owncd, cows and horscs. 37. Marriage is of course one way through which free allownnce is gained. Poliginy, which is still very common (also with women relnted to ench other by close kinship links) is highly vnlued not only for bcinga status-marker, but also for this reason. ln fact, both transhunrnnce migrations ( 001101ra) nnd stratcgies of shnring out one's own fl ock and herd betwccn diffcrent pasturing nrcas controlled by relatives (not only uterinc ones) were (and partly still remain) very common for mnnaging scasonnl or prolonged periods of drought as well ns other risk factors (for cxample disenses, robbery or raids, this last practice occurring when people are involved in a dispute and/or n fcud) which could determine the loss of animais. Other means for hnving nccess to land nnd its rcsourccs are those of giving the owners n payment, cnlled ale1ro11, n term which literally means « for the stomach » (that is, for « compensating » fccding of both people and livestock); or being tied to them in a subordinate rclationship of domestic, working or militnry service. 38. Forms of transmission of livestock to consanguine relatives occur not only at a person's death but nlso during lifc. Thesc for ms include gifts received from birth, and frequcntly fo r a woman, the transfer of animais from one or both parents to her when she goes to live elsewherc with her husband. Th is constitutes n form of endowment to which cthnogrnphers, with the exception of Watson Frnnke (1987) very rnrely pnid attention, when nnalysing the transactions which take pince at marriage. 39. Goulet (198 1) alrendy strcssed this point in his criticism of the analysis of Wayil social organization provided by Watson ( 1967). BmLIOGRAPHY BARRERA MoNROY Eduardo 2000 Mestizaje, co111ercio y resistencia. la Guajira d11ra11te la seg1111da 111itad del siglo xv111, Tnstituto Colombiano de Antropologia y Historia, Bogota. CHACtN Hilario 2003 Lîrica y 11arrati1·a desde ww l'ision ll'ay1111, Editorial Antillas, Barrnnquilla. CROCKER William H. 1979 « Canela Kinship and the Question of Matrilincality », in Maxine L. Margolis a nd William E. Carter (cds), Brazil: anthropological perspectfres, Columbia University Press, New York, pp. 225-249. DESCOLA Philippe 2001 «The genres of gender: local models and global paradigms in the compa rison of Amazonia and Mela nesia »,in Thomas A. Gregor and Donald Tuzin (eds), Ge11der i11 A111azo11ia and Melanesia, University of California Press, Berkeley/Los Angeles, pp. 91- 114. Gouurr Jean Guy 198 1 El 1111frerso social y religioso g11ajiro, Bibliotcca Corpozulia, Maracaibo. 122 Mancuso DESCENT AMONG THE \V A YU GUERRA Weildler 1987 « Aarash, la piedra de las marcas », Uunpara, XXV ( 105), pp. 16-2 1. 1990 « Apalanchi: una vision de la pesca entre los wayuu », in Gerardo Ardila (ed.), La Guajira. De la 111e111oria al porrenir, Univcrsidad Nacional/Fondo Fen, Bogota, pp. 163- 189. 200 l La disputa y la palabra: la ley en la sociedad 1rnyuu, l/M Edit ores, Bogota. GUTIÉRREZ DE P1N1mA Virginia « Organizaci6n social: el clan (casta), cl matrimonio »,i n lndios y bla11cos en 1963 La Guajira, Ediciones Tercer Munda, Bogota, pp. 89-113. HÉRITJIJR Françoise 1981 L'exercice de la parenté, Le Seuil, Paris. HORNDORG Alf 1988 Dualism and hierarcliy in Lowkmd South America: trajectories of indigeno11s social organization, Almqvist and Wiksell, Uppsa la Studies in Cultural Anthropology 9, Stockholm. 1998 « Serial rcdundancy in Amazonian social structure: is thcre a method for poststructuralist comparison? », in Maurice Godelier, Thomas R. Trautmann and Franklin E. Tjon Sie Fat (eds), Transformations of ki11ship, Smithsonian Institution Press, Washington/London, pp. 168-1 86. HUG H-JONES Christine 1979 From the Milk Rfre1: Spatial and temporal processes i11 Nortlill'est A111azo11, Cambridge University Press, Cambridge. HUGH-JONES Stephen 1995 « Tnside-out and back to front: the androgynous house in Northwest Amazonia », in Janet Carstcn and Stephen Hugh-Jones (eds), About the House. Lél>i-Stra11ss and beyond, Cambridge University Press, Cambridge, pp. 226252. INGOLD Tim 200 1 « Ancestry, generation, substance, memory, la nd», in Tim Ingold, The perception of the enviro11111e11t, Routledgc and Kegan, London. JACKSON Jean 1975 « Rcccnt ethnography of indigenous Northern Lowland South America», A111111al Re1•iew of Anthropology, 4, pp. 307-340. JusA YU Miguel Angel Diccionario de la lengua guajira. G11ajiro-Caste/lano, UCAB, Caracas. 1977 JusAYU Miguel Angel and Jesùs ÜLZA Zunm1 1981 Diccionario de la lengua g11ajira. Caste/la110-G11ajiro, UCA B/Corpozulia, Caracas y Maracaibo. Diccionario sistematico de la le11g11a guajira, UCAB, Caracas. 1988 123 JOURNAL DE LA SOCIÉTÉ DES AMÉRICANISTES Vol. 94-1, 2008 LEA Vanessa « Mêbengokre (Kayap6) onomastics: a facet of Houses as total social fac ts 1992 in Central Br:izil », Man , 27 (1), pp. 129-1 53. « The houscs of the Mêbcngo krc (Kayap6) of Central llrazil: a new door Io 1995 their social orga nization », in Janet Ca rsten and Stephen Hugh-Jones (eds), About the flouse. L é1•i-Stra11ss and beyond, Ca mbridge University Press, Cambridge, pp. 206-225. « The composition of Mêbengokre (Kayap6) Household in Central Dra2001 zil », in Laura Rival and Neils Whitehead (eds), Beyond the 1•isible and the material. The A111erindianization of society in the 1rork of Peter Rfrière, Oxford University Press, Oxford, pp. 157-176. LÉVI-STRAUSS Claude and Nicole BELMONT 1963 « Marques de propriété dans deux tribus sud-américaines», L 'Ho111111e, Ill (3), pp. 102-108. LOUNSDU RY Floyd G. 1964 « A for mai account of Crow and Omaha types kinshi p terminologies», in Ward H. Goodenough (ed.), E xplorations in cultural anthropology, McGraw-Hill, New York. MAYDURY-LEWIS David 1979 « Introduction » and «Conclusion», in David Maybury-Lewis (cd.), Dialectical societies. The Gé and Bororo societies of Central Brazil, Harvard University Press, Cambridge. MoNTIEL FERNANDEZ Nemesio 2002 E'irmkuirra ( Linajes), UCAB, Maracaibo. MU RPHY Robert « Lineage and lineality in Lowland South America », in Maxine L. Margolis 1979 and William E. Carter (eds), Brazil: anthropological ｰ ･ ｲ ｾ ｰ ･｣ ｴｦｲ ･ｳ Ｌ＠ Columbia University Press, NewYork, pp. 217-224. ÜLZA Zumm Jesus y Miguel Angel JusAYU 1978 Gra111atica de la le11g11a guajira, UCAB, Caracas. ÜVERING KAPLA N Joan na 1977 « Orientation for paper Topics » and « Comments », in Joan na Overing Kaplan (cd.), Social ti111e and social space in Lowland South America, Actes du XLII Congrès des Américanistes, vol. 2, Société des Américanistes, Paris. PARKIN Robert 1997 Kins/Jip: m1 introduction Io basic concepts, Dlackwell, Oxford. PAZ l PUANA Ramon 1\1itos, ley endas y cuentos guajiros, Instituto Agrario Nacional, Caracas. 1973 PERRIN Michel 1976 L e chemin des illdiens m orts. A1y thes et symboles goajiro, Payot, Paris. 124 Mancuso 1979 1982 1986 DESCENT AMONG THE WAYÛ Siikuaitpa 1Vay1111. Los guajiros: la palabra y el 1•Mr, Fundaci6u La Salle de las Ciencias Naturales, Caracas. Antropologos y 111édicos frente al arte Guajiro de curar, Biblioteca Corpozulia, Caracas. « "Savage" points of view on writing », in Edmundo Maga ùa and Peler Mason (eds), Mytlt and tlte imaginmy i11 tlte Nell' IVorld, Foris Pub., Dordrecht Providence, pp. 211 -231. PICON Fra nçois-René Pasteurs du Nou1·eau Monde. Adoption de f'élemge citez les Indiens Guajiros, 1983 Édit ions de la Maison des Sciences de l'Homme, Paris. RI VIÈRE Peter 1993 «The Amerindia nizalion of dcscent and aAini ty », L'Ho111111e, 33 (l 26- 128), pp. 507-516. SALER Benson « Principios de compensaci6n y cl valor de las personas en la Sociedad 1985 Guajira », Mo11talba11, 17, pp. 53-65. « Los wayù (guajiro) », in Jacques Lizot (ed.), Aborige11es de Venezuela, 1988 vol. III, Fundaci6n La Salle/Monle Avila Editores, Caracas, pp. 25-1 45. SCHEFFLER Harold 1966 « Ancestor worship in anthropology », Current Anthropology, 7, pp. 54 1551. 1973 « Kinship, descent and alliance», in John J. Honigmann (ed.), Hall(fbook of social and cultural anthropology, Randy Mc Nally & Co, Chicago. 200 1 Filiation and a.Olliation, \Vestview Press, Boulder, Colorado. SCHNEIDER David M. 1984 A critique of the tlteory of kinship, Chicago University Press, Chicago. SEEGER Anthony, Roberto DA MATIA and Ed uardo V1vmROS DE CASTRO 1979 «A construçào da pessoa nas sociedades indigenas brasileirns », Boleti111 do Museu Nacional de Antropologia, 32, pp. 1-20. SIMONS F. A. A. «An exploration of the Goajiro Peninsula», Proceedings of the Royal 1885 Geographic Society, 7, pp. 781-796. TAYLOR Anne Christine 200 1 «\Vives, pets and affines: marriage among the Jivaro», in Laura Rival and Neils Whitehead (eds), Beyond tlte l'isible and tlte 111aterial. 1ï1e A111erindianization of society in the ll'ork of Peter RMère, Oxford University Press, Oxford, pp. 45-56. VERDON Michel 1980 « Descent: an operational view »,Man, 15, pp. 129-150. 1991 Contre la culture. Fondements d'une anthropologie sociale opémtiom1elle, Éditions des Archives contemporaines, Paris. 125 JOURNAL DE LA SOC l{ITÉ DES AM ÉRICANISTES Vol. 94-1, 2008 YI VEIROS DR CASTRO Edua rdo 1998 « Dravid ian and related kinship systems», i11 Maurice Godelier, Thomas R. Trautmann and Fran klin E. Tjon Sie Fat (eds), Tra11sfor111atio11s of ki11ship, Smithsonian Institution Press, Washington/London, pp. 332-385. 200 1 « GUT feelings about Amazonia: potential aHinity and the construction of sociality », i11 Laura Rival and Niels Whitehead (eds), Beyo11d the 1•isible and the 111aterial. The A111eri11dia11izatio11 of society i11 the \l"ork of Peter Rfrière, Oxford University Press, Oxford, pp. 19-44. \V ATSON Lawrence « G uajiro social structure: a reexamin ation », A11tropol6gica, 20, pp. 3-36. 1967 1968 « The inheritance of livestock in Guajiro Society», A11tropol6gica, 25, pp. 317. WATSON F RANKE Maria Barbara « \Vomen and property in Guajiro Society», Etlmos, 12, pp. 229-245. 1987 \VILBERT Johannes 1970 « G uaji ro kinship and theeirruku cycle», i11 Walter Goldschmidt and Harry Hijier (eds), The social a11thropology of l atin America. Es.m ys i11 l1011or of Ralph l eo11 Beals, UC LA, Studies 14, Los Angeles, pp. 306-357. \VILBERT Johannes, Karin SIMONEAU and Michel PERRI N (cds) 1986 Folk litera/ure of the Guajiro l11dia11s, UCLA, Los Angeles. 126