A New Conception of Totemism - iscte-iul
Transcription
A New Conception of Totemism - iscte-iul
A New Conception of Totemism Author(s): Robert H. Lowie Source: American Anthropologist, New Series, Vol. 13, No. 2 (Apr. - Jun., 1911), pp. 189-207 Published by: Blackwell Publishing on behalf of the American Anthropological Association Stable URL: http://www.jstor.org/stable/659643 Accessed: 14/01/2010 10:38 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=black. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Blackwell Publishing and American Anthropological Association are collaborating with JSTOR to digitize, preserve and extend access to American Anthropologist. http://www.jstor.org American Anthropologist NEW SERIES VOL.I3 I9II APRIL-JUNE, NO. 2 A NEVVCONCEPTIONOF TOTEMISM BY ROBERT H. LOWIE recentpaperon totemism significance of Dr Goldenweiser's lies in the fact that it presentsfor the Srst time what rnay view of totemism,"be legitimatelycalled "an 9Kmerican "American"not only becauseit takes into accountthe data of Americanethnography,btlt in the far moreinnportantsense that principleswhicharebecoming it is a view basedon methodological the commonpropertyof all the activeyoungerAmericanstudents of ethnology. Accordingto the traditionalview, totemismis an integral phenomenonwhich is everywhereessentially alike. Thus, in Frazer'slatest work on the subject,Totemismand Exogamy, the thruston the shouldersof thosew-ho burdenof proofis explicnitly questionthe identity of totemicphenomenain differentquarters of the globe and who upholdthe theoryof convergentevolution. In Part I of his paper, "Australiaand British Columbia,"Dr Goldenweiser has anticipatedthis challenge. He selectsthe series of featuresthat are commonlyregardedas distinctiveof totemism, and comparesthe formsthey assumein the two areasconsidered. The result is sufficientlystriking. On superficialconsideration, it appearsthat the Australiantotem groupresdmblesthe clan of British Columbiain the exogamicregulationof marriage. But this resemblanceis not significant;in both cases the exogamous THE Tofemism, An Analytical Study. By A. A. Goldenweiser. Reprinted from the ournal of Simerican Folk^Lore, vol. XXIII, April-June, I9IO, No. LXXXVIII, pp. II5. I 89 AM * ANTH., N. S., I3-I3 I90 AMERICAN ANTHROPOLOGIST [N. S., I3, I9I I characterof the clanis not a primary,but a derivativetrait. Becausethe clansare,in bothareas,partsof the largerphratricunits, and becausethese phratriesare exogamous,the totem clans must be exogamous,even thoughthe cLan,as a clan, may have nothing to dowithexogamy. In otherfeatures,the totemclansof Avlstralia and British Columbiaclearlv diverge. In Australiathe social importanceof the clan dwindlesinto insignificance as compared with that of the phratrwr; in BritishColumbiathe clanis the social unit, par e;rcellence.On the Northwestcoast thereis evidencefor the developmentof the clans from village communities,such as nowhereexistin Australia. Finally,the Americanclansaregracled as to rank a conditionlikewiselackingin Australiantotemism. In the matterof clan names,what similarityexists is again of a superficialkind. In Australiaall clans are named from their totems; in British Columbiaclans frequentlyderivetheir names fromlocalities. But preciselywherethe Americansocialdivisions (phratries)are namedafteranimals,we occasionallyfind that the eponymousanimalis not identicalwith the crestanimal,whichis the one that corresponds, in religiousfunction,to the Australian totem. If phratriesare comparedit is found that those of the Tlingitand Haidabearanimalnames,but that only a few of the Australianphratrynamesaredefinitelyknownto referto animals. The view that the totemiteis a linealdescendantof his totem is clearlydevelopedinAustralia;onthe NorthwestCoast,onthe other hand, there is a fundamentalbelief in hlmaa3descent:the crest animalis onewhichhasin somewaybeenassociatedwiththehuman ancestorof the group. Nevertheless,the authorpointsout, there are mythsin whichthe associationis very close,and, in one group of traditions,the ancestoris the crestanimaltransformed.These instances,insteadof militatingagainstthe author'spointof view, constitutein reality strong evidencein supportof it. For the myths in questionresultfromthe reactionof the guardian-spirit conceptuponthe basicbeliefthat humarlbeingshave humanancestors Now, the guardianspirit conceptis practicallyforeign to Australia. Whatsimilaritythereexistsbetweenthe Australian and the Americanmyths is accordinglyan ideal instanceof convergent evolution. There remainsthe criterionof the taboo A NEW CONCEPTION OF TOTEMISM LowIE] I9I against eating or killing totems. Of this phase of totemic life Australiaremainsthe classicalexample;in British Columbia,on the otherhand,not a singleinstanceof totemictabooshas hitherto been discovered,thoughthereis an abundanceof taboosof nontotemiccharacter. A survey of the currentlyassumedsymptomsof totemismin the two areas discussed thus reveals far-reachingdifferences. It wouldbe artificial,however,to confinethe comparisonwithin the limitsset by conventionaldefinitionsof totemism. If we wish expressedin these to disabuseourselvesof the preconceptions insists,we mustnot neglectto consider definitions,Dr Goldenweiser are empiricallyfound in intimate which those culturalfeatures associationwith the criteriagenerallyrecognizedas totemic. In Australia,two elementshave risen to so commandinga position within the totemic complesthat each has been assumedas the of totemismgenerically. Theseelements essenceandstarting-point are the intichiumaceremoniesconductedfor the multiplicationof of ancestral the totem animals,and the beliefin the reincarnation features analogous America, of Coast Northwest the On spirits. totemic from dissociated wholly are indeedfound, but they are viewing in institutions. A parallelconditionof affairsis revealed America. The the dominanttraits of social life in northwestern taking into without unintelligible is Kwakiutl the social life of spirit; guardian same the sharing accountthe groupsof individuals essentially is among the tribes farthernorth the clan tradition an account of the ancestor'sacquisitionof his guardianspirit, while the circumstancesincident thereto are dramatizedin the dancesof the secret societies. In Australiaguardianspirits are rare,and,wherefound,aregenerallyquitedistinctfromthe totems; factor even whenthe two conceptsdo coincide,the guardian-spirit significance special of trait second A moment. is of relativelyslight in the Americanarea is the relationshipof totemismto art,-the saturationof practicallyall decorativeattempts with totemic a totemic motives,andthe retroactivetendencyto give,secondarily, intimate This origin. in decorative purely designs to interpretation charstyle quasi-realistic the on connectionis largelydependent geowhere acteristicof NorthwestAmericanart. In Australia, I92 AMERICAN ANTHROPOLOGIST [N. S., I3, I9I r art has exertedbut little influence metricalmotivespredominate, on totemiclife. of the totemiccomplexin the two The empiricalconsideration typical regionsdealt with thus establishesthe essentialdiversity compared. Thedominantmotivesof Australian of the phenomena totemismare not the dominantmotivesof NorthwestAmerican totemism,and vice versa. What resemblancesexist are either superficial,or are functionsof traitsnot directlyassociatedwith the criticismmight be made that tototesnism. Here, howesrer, becausetheAmerican is temismin theareasselected not comparable development. "The of stage institutionrepresentsa far later totem,"as the authorputs it in anticipationof this stricture,"has becomeattenuatedto a crest,to a symbol;theliving,fleshandblood relationshipwith the totem animalhas been transferredinto the realm of mythology; and naturallyenough,the taboo on the totem animalhas dwindledaway and finally disappeared."Dr clear. "To a retortof that answeris unmistakably Coldenweiser's character,I wouldanswerthat we may safelyassertthat thereis in an evolutionnot one phaseof humanculture,so far represented the succession where development, of stages successive ary seriesof historicfact of observation by given has been so amplyjustified as to be safely adoptedas a principleof interpretation"(p. 22). The evidencefromNorthwestAmericamust thus be admittedas coordinatewith that fromAustralia. the validityof the theoryof convergent Havingdemonstrated evolutionfor the totemicphenomenaof Australiaand the North PacificCoastof America,the authorproceedsin the secondpart of his paper,"The TotemicComplex,"to consider,one by one, the traditionalelementsof totemism,and to determinetheir mutual relationson the basis of the widest possibleethnographicbasis. takesthe imIn the first sectionof Part II, Dr Goldenweiser of totemism. concept the from exogamy portantstey)of divorcing Nandi of the Assam, A discussionof the data on the nativesof E;astAfrica,severalPlainsIndiantribes,and the Samoan-Fijian culture province,leads to the conclusionthat "clan exogamy, althougha usualconcomitantof the othertotemicfeatures,is not . . hencenot a necessary,concomitantof the latter;and . onstatlt, LOWIE1 A NEW CONCEPTION OF TOTEMISM I93 again,wherethe otherfeaturesareabsent,exogamymay nevertheless occur" (p. 55). The pages immediatelyfollowingembody what is probably the author'smost importantand originalcontributionto the suS ject, a critiqueof the conceptof exogamyitself. Not only may clans exist independentlyof esogamicrules, and exogamicrules independentlyof clans, but even whereclan exogamydoes exist the unionof exogamywiththe clanunitmaybe a secondaryfeature. In the case of the Kurnai,to be sure, specialconditionsseem to warrantHowitt'sconclusionsthat marriagewas originallyregulated by non-localizedtotem clans, that paternaldescenteffected a localizationof the totems, and that consequently,in recent times, locality has appearedas the regulatingfactor. But we this modeof shouldnot by an) meansbe justifiedin transferring reasoningto other areaswhere difFerentconditionsprevail. In 13ritishColumE)ia,for example, it seems probable that the clans at one time occtlpiedseparate villages. Now, whether the clans were exogamousas clans,or becausethey happenedto coincidewith local exogamicdivisions,becomesan open question as soonas the existenceof exogamydissociatedfromclanshas beenestablished. The point comesout even moreclearlywhere kinship exerts an influenceon matrimonialregulations. Among the Todas a purelyobjectiveinvestigationrevealsa numberof exogamousclans,as wellas a seriesof matrimonialrules based on the matteris quitedifferent. considered, kinship. But, subjectivelnr The Toda merelvknowsthat certainrelativesthroughthe father and certainrelativesthroughthe mother all of them included personsfor him. Besides in the termpuliol are not marriageable membersof otherclans,an individual'spuliol groupalso embraces all the peopleof his own clan,but this fact does not seemto have been noticed by Rivers'informantsbeforehe pointeditiout to them. Similarlw,amongthe Blackfoot,membersof a band are forbiddento intermarry,not as band members,but becausethey are consideredblood relatives. In a manuscriptby Sternberg, which the author lays under contributionfor additionalfacts, the Gilyakaredescribedas a peoplewithexogamousgentes. Nowt wheregentes as such are the exogamousunits, two gentes suffice I94 AMERICAN ANTHROPOLOGIST [N. S., I3, I9I I for the regulationof tribalmarriages,the rnenof gensA marrying womenof gens B, and aJiceuersa. This reciprocalrelationshipis preciselt7 what does not obtainin Gilyaksociety. If the men of gens A marrywomenof gens B, the men of gens B are ipso facto debarredfrommarryingwomenof gens A, the two gentes being regardedas gentesof "sons-in-law " and "fathers-in-law " respectively The men of B must thus marrywomenof gens C, and so forth. It is, accordingly,obviousthat the gens as such does not determinemarriage. The thesis is thus establishedthat "whenthe fact of a given socialgroupnot marryingwithin itself is ascertained,the informationacquiredis but partiallycomplete"(p. 59). Theexogamous natureof a group,as objectivelyobserved,mayindeedbe a primary trait;butit mayalsobe a derivativetrait, a necessaryconsequence of otherregulations not linkedwiththe groupas such. Dr Goldenweiseris thusemboldened to inquire,whethertheexogamiccharacter of Australiantotemclansis a primaryor a secondarycharacteristic. Takingup firstthe simplerformof socialorganizationtypifiedby the Dieri, viz., two exogamousphratriessubdividedinto smaller totem groups,he findsthat a givenclan can not be consideredan exogamicunit becausein no case are the exogamicmarriageregulationsfully determinedby clan affiliation. If this conditiondid obtain,nothingwouldpreventmembersof clan a frommarrying membersof clan b of the samephratry. But this is emphatically not what takes place. In reality, clan a "behaves exactly as wouldan individualof phratryA if therewere no clans" (p. 60). Becauseit formspart of a largerunit exogamicin its own right, clan a must ipso Jcgstobe exogamic. The conditionof affairsis strictly parallelto that among the Tlingit and Haida Indians. In Australiathe derivativecharacterof clanexogamyis illustrated amongthe anomalousAranda(Arunta),wheresome clans occur in both phra-tries.Herea man of clan a may marrya woinanof clan a if she belongsto a differentphratrysbut he mustnot marry her if she belongsto the samephratry. The argumentis greatly strengthenedby a considerationof the four-classsystem. Forherethe class is the marriage-regulating unit, and the clanstare in no senseexogamousunits,as each clan LOWIE] A NEW CONCEPTION OF TOTEMISM I95 containstwo sets of memberswithdistinctmatrimonial regulations. In tribeswith the eight-sub-class systema parallelargumentholds: the sub-classis the marriage-regulating unit,andeachclanconsists of "fourmatrimonially heterogeneous units." The questionarises, what, in thesesystems,may be the marriage-regulating functions of the phratry,and of the phratryand class, respectively? Dr Goldenweiser is of opinionthat in the four-classsystem the fact beyonddoubtis the exogamyof the cZass,whilethat of the phratry remainsto be investigated;in the eight-sub-class system the immediatedata indicatethe exogamousnatureof the sub-class,and the matrimonialfunctionsof the class and phratryremainto be investigated. The point to be determinedwould be the native feelingwith regardto theselargerunits, whether,forexample,the phratryof four-classtribescontinuesin the mindsof the natives to constitutea distinctexogamicgroup. The necessityof taking into account the subjectiveattitude of the natives is strikingly illustratedin perhapsthe most suggestivepassageof the entire paper. A purely objectivedescriptionof the regulationsfound among four-classtribes does not by any means necessitatethe current mode of representation.Instead of subdividingtwo phratriesinto two exogamousclasses,it is possibleto unite the intermarryingclasses into endogamousmoietieswith exogamous subdivisions. This has actuallybeendoneby ProfessorKlaatsch, a relativelwr naiveobserverin mattersethnological, in thedescription of Niol-Niolsocialorganization.This travelerhas even recorded native names for the endogamousmoieties. Dr Goldenweiser rightlyinsiststhat, objectively,Klaatsch'smodeof representation is as legitimateas the one ordinarilyemployed. His suspicions arearousedmerelyby the fact that endogamous moietiesrecognized as such by the aborigineshave hithertoescapedthe eyes of other Australianethnographers.Moreover,the classnamesof the NiolNiol apparentlycorrespond to thoseof the Aranda,whosephratryclass organizationseems firmly established. Accordingly,the author does not contendthat Klaatsch'sschemerepresentsthe subjectivefacts,thoughhe admitsthat "on a parwiththedominant phratricorganization theremayalsoexistin theseAustraliantribes a consciousnessof the objectivelyendogamousgroupsconstituted by the pairsof intermarrying classes"(p. 64). I96 ANTHROPOLOGIST AAIERICAN S., I3, [N. I9I I illuminatingdiscussionthe critic can In this extraordinarily "exogamy." findfault only with the author'suse of the term members wordis obviouslyrequiredto designatethe ruleagainst A the rule ofa groupmarryingamongthemselves, in otherwordsS is the etyofthe incest group. The word sanctionedby usage exogamy." " term andself-explanatory unexceptionable mologically however,modelinghis conceptionof the "typical DrGoldenweiser, relation"on the conditionssupposedlyfoundin Ausexogamous tribeswith two phratries,writes:"An exogamousrelation tralian onlywhenboththe groupwithinwhichmarriage isfullyrepresented prescribed, isprohibited,and the one into whichit is permittedor four(among class the views he aregiven" (p. 60). Accordingly, an finds and classtribes) as the exogamousunit par excellence, institution to "puretotemicexogamy"in the Arabana approach only one ofeach totem c]an being permittedto intermarrywith phratry. Werethe matter clan of the complementary particular of a currenttermwould, there-definition onepurelyof nomenclature, howof course,be perfectlylegitimate. In the presentinstance, can not ever,it seems to the critic that the term is not, and sense. modified the in convenientlybe, used with consistency own their in UJherethere are only two social units exogamous followsas a physicalnecessity;the groupinto right,intermarriage by the which marriageis permittedor prescribedis determined obviously is This merestatemDntof the prohibitoryregulations. within not the casewhentherearefour,or six, or fourteengroups, might be each of which marriageis prohibited. To be sure, it Dr Goldensaid that in suchinstancesthe exogamousrelation,in statements where as, inasmuch weiser'ssense,is fully represented, marry may to the contraryare lacking,a memberof group I are to membersof all other groups. But if positiveregulations whether be takeninto account,it certainlyis not the same thing own, his besides existing a man mustmarryinto the only group groups,or whetherhe must marryinto one of a numberof other To whetherhe may marryinto ary of the othergroupsextant. Either the critic it seems that there are only two alternaties. mutual the Then esogamy. of we adoptthe author'sconception classeswith rulesagainstintra-class relationshipof intermarrying LOWIE] A NEW CONCEPTION OF TOTEMISM I97 marriage wouldformthe standardillustrationof exogamy;phratries wouldformally,but, for reasonsjust given, might only formally, exemplifyesogamy;and it wouldbe inadmissibleto speakglibly of four exogamousTsimshianclans (p. 9), of a great numberof exogamousKhasiclans(p. 53), of fourteenexogamousBahimaclans and forty-oneexogamoussepts (p. 74). Or, we cling to the acceptedusageof the term. Thenexogatnymaybe ascribedto anv groupprohibitingmarriageamongits members. In this case, the exogamyof the Kamilaroiclass,as wellas the exogamyof the Aribanaclan,is a derivativefeature, a logicalconsequence of phratric exogamy. In additionto this derivatively(andthereforerelatively unimportant)exogamictrait, the Kamilaroiclassand the Arabana clan have certain positive marriage-regulating functions,whicht however,have nothingto do with exogamy,of whichthe functions are only prohibitory. In the next part of the sectionon "Exogamyand Endogamy," the authorbrieflymentionsthe constanttendencyto extendregulationsof marriage,even wherefairlydefiniteregulationsalreadv exist. An unusuallysuggestiveinstanceis furnishedby the Toda (p. I68). Withinthe (endogamous) Teivaliolmoiety there are a numberof exogamousclans. But the membersof the Kundr clanoutnumber theotherclansto suchanextentthattheexogamous rulecanonlybe followedby the Kundrmarryingmostof the members of the other clans. Thus very few of the latter are left to marryone another,and the conditionof affairsseemsto approach as a limit the widespreaddivisionof a tribe into two exogamous intermarrE ing phratries. The occurrenceof positi2oeobligations forcertainclassesto intermarrv a pointtoo little noticedby other writers is strikinglyillustratedby the Gilyakgroupsof prospective husbandsand wives. Dr Goldenweiser, in discussingthe matrimonialinstitutionsof this people,also callsattentionto the correlated rulesof "psychic intercourse."Thereis restrictionof conversationand intimacy between personswho might come into conflictfromjealousy,and avoidanceobtains,in differentdegrees of stringency,betweenrelativesdebarredfromintermarrying.On the otherhand,thereis greatfreedombetweenprospectivehusbands and wives, and an extraordinary cordialitycharacterizesthe re- I98 AMERICAN ANTHROPOLOGIST [N. S., I3, I9II lations of fathers-in-lawand sons-;n-law.There can be little doubtthat the correlation of the rulesfor sexualandpsychicintercourse,whichthe authorconsidersonly in connectionwith a single tribe,meritsmoreextensiveinvestigation. The next two sections,"TotemicNames"and '-'Descentfrom the Totem,"add little to the argumentof the corresponding dirrisionsof Part I. Someadditionalexamplesare adducedto show thatreponymous totems,xvhileremarkablyfrequent,do not occur universallyS and that the totem is not invariablyregardedas the ancestor-ofthe group. Underthe heading"Taboo,"the author pointsout that quitegenerallyrestrictions of confltlctareassociated with groupsother than totem clans, while there are tribes,such as those of the Iroquoisconfederacy,whose totem clans are not connectedwith taboosagainstkillingor eatingthe eponymousanimal. In reply to the speciousreasoningof many writers,that totemitesabstainfromkillingor eating their totemsbecausethey regardthem as kin, Dr Goldenweiser intimatesthat taboosma+t have a variety of origins. The Omahafurnisha tellingseriesof illustrations. For heremanyof the taboosassociatedwith totem groupsare logicallyunconnectedwith the totems, an(l it seems practicallycertainthat eachof these "fancifulprohibitions" had a distinctorigin. In the pages on "The ReligiousAspect of Totemism,"the author emphasizesthe fact that totemismand animal cult are distinctphenomena. Animalworshipis prominentwheretotemic groupsdo not exist. On the otherhand,worshipof the totem is very rare,andin somecasesthereis a completeabsenceof religious associationswith totems. It is obviousthat underthe circumstancesit wouldbe absurdto regardtotemismas a formof religion, or as a distinctstagein the evolutionof religiousbeliefs. On this point, at least, Dr Goldenweiser findshimselfin agreementwith the views put forwardin Frazer'smost recentpublicationon the subject. The comprehensive surveyof ethnologicalphenomena in PartII thus confirmsthe conclusionsarrivedat from a comparisonof Australianand BritishColumbianconditions. Totemismcan no longerbe consideredas an integralphenomenon.Totemiccom- LOWIE] A NEW CONCEPTION OF TOTEMISM I99 plexes are ''conglomeratesof essentially-independentfeatures" (p. 88). It may be possibleto trace logicallythe development of the severaltraitsfroma singlehypotheticalfactorof fundamental importance,but only throughhistoricalproofscan suchdeductions gain scientificvalue even for limitedareas. Neithera systemof naminggroupsafter totem animals,nor the doctrineof descent fromthe totems,nor a religiousregardfor the totem,in fact, not one of the symptomsordinarilyassumed,is a constantfeatureof totemism;andthereis no evidenceforthe historicalor pswrchological primacyof any one of them. The instancesof other factorsmagicalceremonies in Australia,estheticmotivesin BritishColumbia-rising to prominencewithin the totemic complexillustrate the variabilityof the phenomenonstudied,and lead to the important queries,"If totemissmincludes,roughlyspeaking,everything,is totemismitself anythingin particular) Is thereanything specificin this phenomenon,or has the name 'totemism'simply been appliedto one set of featureshere,to anotherset there,and still elsewhereperhapsto both sets combined?"(p. 89). Dr Goldenweiser repliesthat, in the light of his foregoinganalysis,the specifictrait of totemismcan not be a certaindefinite sum of elements,but only the relationobtainingbetweenthe eZements (p. 92). In a given totemic complex,factorsa, b, c, . . ., are associatedandcorrelatedso as to forma relativelyintegralcombination. The fairly completeintegrationof totemic factorsresults fromthe factthatelementsin themselvessociallyindifferent become associatedwith clearlydefinedsocialgroups,the associationbeing effectedby meansof descent(p. 93). In definingthe relationship of the totemicelements,the authorstartsfroma consideration of the currentview that totemismhas a religiousand a socialaspect. Theoccasionalabsenceof any religiousfactor,notablyamongthe Iroquois,induceshim to eliminatethe term "religious" and to conceivetotemismas the-associationof "objectsand symbolsof emotionalvalue"with definitesocialunits,the latterbeingdefined asunits perpettlatedthroughdescent. Again,totemismis usualty dese.ribed as a static phenomenon.Yet, nothingis rnoreobvious than its variabilityin time. Dr Goldenweiser'sinvestigatiorl, accordingly,culminatesin the -dynamic definition:"Totemism 200 AMERICAN ANTHROPOLOGIST [N. S., I3, I9I I is the processof specificsocializationof objectsand symbolsof emotionalvalue" (p.-97). As an epilogue,the substanceof whichwouldhave more appropriatelyprecededthe definitionof totemism,comesa discussion of "Origins;in Theoryand History." Schmidt's, Frazer's,and Lang'stheoriesare jointlysubjectedto a methodological critique. Insteadof attemptingto understand presentconditionson the basis of theirestablishedantecedents,these theoriesselect a prominent featureof moderntotemismand projectit into the past, assuming it to be the starting-point of the totemicprocess. This,the author contends,is unjustifiable;for what is now of overshadowing signiScanceneednot alwayshavefiguredwith equal conspicuousness. The secondstep made by the theoristsmentioned, namely,the deductionof otherfeaturesfromthe one assumedto be primary, is likewiseillegitimate;for it assumesthe unity of the totemic featuresanda uniformlawof development. Theformer assumption has beenrefutedby the precedinganalysis,whilethe latter seems doubtfulin the light of modernresearch. Finally, the authors criticizederr in neglectingthe influenceof borrowingon the developmentof culturein a givenarea. To bring home this last point, Dr Goldenweiser proceedsto showwhat the courseof totemicdevelopmenthas actuallv beenin the carefullystudied region of British Columbia. While the southernShuswaphave the loose village organization typical of the Salishtribesof the interior,the westernShuswap have a socialsystem obviouslypatternedon that of the coastal tribest andindirectlyderivedfromthem. Amongotherinstances within the same area, the transformation of the irxstitutions of certain Athapascantribes is especiallyremarkable.Such features as potlatches, clan exogamy,and an hereditarynobility,have been obviouslyborrowedfrom neighhoringcoastal tribes; and in so faras the Athapascantribes possessingthese traits differ in the detailsof these institutions,the differencescan sometimes be directly explainedby contactwith correspondingly differingtribes ofthe coast. The actualhistoryof such changescould neverbeforetold by meansof speculationsas to primitivepsychology;it wasascertainedonly by intensivestudy of the influencesto which IdOWIE;] A NEW CONCEPTION OF TOTEMISM 20I each tribe has been subjected(p. I09). In the data alreadyaccumulatedon Australia,Dr Goldenweiser findsevidenceof the farreachinginfluenceof diffusionon culturaldevelopment;and his paperterminatesin the confidentprophecythat future research will revealconditionsof borrowingcomparable to thoseestablished in BritishColumbia. As Englishethnologistsseem to adoptonly with reluctancethe historicalpoint of view advocatedby other students,it may be well to recallTylor'smemorable words:"Most of its phenomena(that is, of humanculture)have growninto shape out of suchta complicationof events, that the laboriouspiecing togetherof theirprerrious historyis the only safe way of stuclying them. It is easy to see how far a theologianor a lawyerwouldgo wrongwho shouldthrow history aside, and attempt to e2iplain, on abstractprinciples,the existenceof the ProtestantChurchor the CodeNapoleon. A Romanesque or an EarlyEnglishcathedral is not to bestudied-as thoughall that thearchitecthad to do was to take stone and mortarand set up a buildingfor a given purpose" (Researchesinto the Early History of Alankind,p. at). The historicalsignificanceof Dr Goldenweiser'sessay will perhapsbecomeclearerfroma parallelbetweenthe (levelopment of ethnologicalthinkingand the evolutionof philosophical thought sn general. Popularphilosophyhas always had the tendencyto assumea necessarybondbetweenthe constituentsof a relativelwr stable complexof observedelements,-to assumethat there is a "thing"whichhas properties,an ego whichhas sensations,feelings, and other manifestationsof consciousness.Valuable as such summariesof experienceare froma practicalpoint of view, they becomeindefensiblefroma higherstandpoint. The ideaswe form of "things"resultfroman association(by contiguitv)of the ideas of its properties. A childlearnsby experiencethat a brownpatch of colorand a certainform of resistanceto the touchare linked together,and lry connectionsof these i-deasdevelops the idea of a table. A "thing"-is thus nothingdistinctfromits properties; it is nothingbut tlie sum-totalof theseproperties;thereis no mvstic unitv in reality apart from the properties. (Cf. Hoffding,PsychoSogie,pp. 2I2, 226, 285.) The ethnologist,like the uncritical philosophers is confrontedat everystep with conjunctionsof feat- woX AA!ERICAN ANTHROPOLOGIST |N. S., I3, I9I I united. >A geometricalpattern ureswhichat firstseemindissolubly isassociatedin th-eprimitivecraftsman'smindwith somedefinite is animalor plant. It is naturalto assumethat the associatuion realistiG at attempt a pr;maryone, that the designis a degenerate representation.Gamesare playedas meansof divinationor processesof sympatheticmagic. - Shouldthey not be conceivedas ceremonialcontrivances?Tales of heroic exploits culminatein the hero'sascensionto the sky. Must not the whole plot be a functionof hiscelestialaffiliations?Socialunitswithanimalnames and food taboosprohibitmarriagewithin the group. To regard names,taboos,and exogamousrtllesas merelymanifestationsof the samefundamentalphenomenonis, at a relativelyearlystage most intelligiblething of inquiry,the (3bviousand psychologically to do. At a morecriticalstage,however,theinstabilityof the complexes notice. Whatwas at firstsupposedto be a necessaryconattreacts nectionis reducedto a mereconjunctionof elements. Thought of the mysticunderlying is no longerarrestedby a contemplation units and theirrelationswith the observedelements;to deterrnine of theseelementsthemselvesbecomes the natureand interrelations the highest,nayonlypossible,goalof investigation. In the domain of physicalscienceaa criticalreformationof this type has been, within recent decades, effectedby ProfessorErnst Mach. In aim, ethnology,the school which has set itself a corresponding ethmystic in belief traditional whichendeavorsto supplantthe only still be, nologicalcomplexeswith a deeper,though,it may proximateanalysisinto provisionalelements,is the schoolheaded by ProfessorFranz Boas. Under-his influenceKroeber and Wisslerhave shownthat the same patternis subjectto varying even within the same tribe: designand interpreinterpretations to distinctpsychologicalprocesses. tationare foundto correspond to the conjunctionof story regard with An analogousconclusion by the presentwriter. drawn has been plot and cosmicphenomena Independentlyof Boas, but in thoroughharmonywith his point of view, Seler, in criticizingPreuss,and Haddon,in criticizing (Sulin,have pointedout en passant that the associationof ritual with formsof diversionis a secondarydevelopment. What all LOWIE] A NEW COSYCEPTION OF TOTEMISM 203 these writershave attemptedin the study of theirown problerns, has done for the far more complicatedsubject Dr Goldenweiser of totemism. He has shownthe futilityof attemptingto connect any definiteconceptionof concreteethnologicalfactswith the term "totemism." He has shownthat thereis no justificationfor assuminga commonsubstratumunderlvingall the "totemic"complexes:a completestatementof all the social, religious,esthetic, as foundin a given and other correlateswith their interrelations, area, exhausts the possibilitiesof descriptionand explanation. However,as alreadyshown in the resumeof the sectionon doesnot abandon "The Complexin the Making,"Dr Goldenweiser by a reretention its justify the term "totemism,"but seeks to definitionof the word from a dynamicstandpoint. It is here that he passesbeyond the limits reachedby his fellow-students of secondaryassociations. For, while the latter are generally contentto indicatethe factthat a secondaryassociationof elements boldly undertakesto define,vvith has occurred,[)r Goldenweiser some precision,the processitself of the association. That is to say, he doesnot merelyholdthat totemismis the resultof a secondary associationof social units with variousfactors. He holds,in addition,that the associationresultedfromthe fact that objects and symbolswhichwereoriginallyof emotionalvalue only to individualsbecame,throughdescent,valuesfordefinitesocialgroups (P-97) Beforeenteringinto a critiqueof this conception,it is worth noting that many formsof associationnot;ordinarilyconsidered totemicwouldbe classedas such accordingto the new definition.A phratryand a local groupmightillustratethe dynamicprocess in questionas well as any "totem kin" of other writers. In particular,the fact that the name occupiesno favoredposition, but appearsas but one factorof many that may be associated, seemsto render"totemism" almostall-inclusirJe.Thisis especially the casewhenwe considerthat, on the author'stheory,it is not at all necessarythat the namesbe derivedfrom animalsor plants. DiscussingIroquoistotemism(p. 96, footnote),Dr Goldenweiser arguesthat even here, wherethe totem is merelya name,it, at an emotionalvalue,inasmuchas otherleast formerly,represented 204 AMERICAN ANTHROPOLOGIST I9I I [N- S., I3, wisethe namewould-nothavebecomefirmlyfixedin socialgroups. Obsriously,the same reasoning which the reviewercannot considerconclusive-wouldapplyto localunitswithnon-animal names. It is not clearwhether,or where,the authorwoulddrawthe line here;indeed,the data bearingon namesof totemgroupsrequire moreextensivetreatmentthanthat givenin thepresentpaperbefore it will be possibleto forma clearview of Dr Coldenweiser's conceptionof this specialpoint. Dr Goldenweiser's definitionof totemismmay be considered fromtwo pointsof view. In how far does it accuratelyrepresent the phenomenacommonlydesignatedas totemic? And, to what extent does it representthe totality of phenomenawhich seem psychologicallyand sociologicallyrelated with these totemic phenomena? In replyto the firstquery,it mustbe admittedthat the author's definitionoutlinesa plausiblecourseof development. Nevertheless, it is possibleto conceivethat conditionsotherthanthosedefinedby Dr Goldenweisermay lead to typical totemism. Assume two locallydistinctgroups,eachwith its own taboosagainstthe eating of a certainanimal. Then the unionof these two groupswould leadto a typicaltotemicsociety,in-the ordinarysenseof the term, if we add the featureof exogamy. Sucha hypotheticaldevelopmentin no way militatesagainstthe author'sgeneralpointof view. Nevertheless,it is perfectlyeasy to understandthe process,from what we knowof the developmentof taboos,withoutrecourseto the theorythat the taboowas originallyof only individualsignificanceand afterwardsbecamesocializedthroughdescent. Or,to take a casewhichis not hypothetical. Whatevidenceis thereto show that amongthe Iroquoisthe clan name was originallyan individualpossessionwhich, throughdescent,becamesocialized? To excludeinstancesof thistypefromthe list of totemicphenomena by a rigorousapplicationof the definitionwouldreducethe whole discusFionto a logomachy,which would be entirely beside the author'spurpose. Forwhathe attemptsto do is preciselyto define the essentialfeaturesof the processresultingin whatare ordinarily calledtotemicphenomena.The fundamentalobjectionto such a definitionas Dr-Goldenweiser hasattemptedis, that it is frequently LOWIEl A NEW CONCEPTION OF TOTEMISM 205 impossibleto determinewhetherit correctlyrepresentsthe historical process of association. If we assume the association of name and social group as the starting-point of totemism and, as the author himselfhas shown, this combinationsometimesexhausts the content of totemism it is, in our ignorance of the actual history of the development,impossibleeither to prove or to refute the theory that the group names, not only in the Iroquois, but in the Australian cases as well, ever served to designate individuals. The inherent probabilitv of such a condition does not seem very great. If the associationof tabooand socialgroupis taken as the startingpoint, the a priori probabilityof a socializingprocesswill presumablyappear considerablygreaterto the majorityof ethnologists.. Nevertheless, the hypothetical instance given above seems to indicate that socialization is not a Denknotwendigheit for the comprehensionof the established associatiotl. The critic is thereforeof opinion that a non-committal attitude on the process of association (so far as it ellldes observation) is highlv advisable. Totemism would then be defined, not as a socializationof various elements of (at least potentiallv) emotional value, but merely as the associationof such elements with social groups. The second question is, does Dr Goldenweiser's conception embrace all the phenomenaessentially related to those of totemic phenomena generally recognized as such? The writer feels that, inclusive as is Dr Goldenweiser'sdefinition, it limits the field of totemism too narrowlyby an exaggeratedemphasisof the element of descent. By a "complete social unit" Dr Goldenweiserunderstands one group of at least two within the tribe, each including both men and women, and perpetuatedby descent (pp. 93, 94, 97, 98). Accordingly,in dealingwith the resemblancebetween totemic institutions and religious societies whose membersshare the same gt1ardianspirit, he does not discover a genutnehomology. "While a certain psychological ainity between the two institutions is not improbable, their genetic relationship, claimed by some, calls for demonstration" (p. 94). The matter of genetic relationshipmay be dismissed at once as irrelevant, for as Dr Goldenweiser,on the very next page, states his belief in the convergent evolution of totemic phenomena,absenceof genetic connectionsurouldnot, from AM. ANTH. N. S., I3-I4 206 AMERICAN ANTHROPOLOGIST [N. S., I3, I9I I hispointof view,barreligiousorganizations fromthe foldof totemic institutions. Their exclusion,then, rests essentiallyon the definitionof a socialunit. Now, the definitiongivenby Dr Goldenweiserseems to the writer quite arbitrary. If the peculiarity of totemicphenomena liesonlyin the relationobtainingbetweenthe elements(p. 92), the psychological resemblance of this relationship wouldseem to be the predominant issue,whilethe precisenature of the socialgroupbecomesnegligible. Amongthe GrosVentre (Atsina),whereevery man passessuccessivelythrougha seriesof age-societiestthese gradesare well-definedsocial units. The associationwith eachof themof a certainanimalfor whichseveral of the societiesarenamed,doesnot seemto differin principlefrom theassociation of a clanwithits crestoreponymousanimalancestor. It ma not be out of placehereto referto the fact that Schurtz has alreadydarklyhintedat a connectionbetweentotemismand the age-gradesof the northernPlains Indians(Altersklassenund Mciznnerbunde, p. I54). The argumentjust advancedin behalf of age-societiesis obviouslyapplicableto the type of religious societiesspecificallymentionedby Dr Goldenweiser,as well as to still other formsof social units. Is their exclusionjustifiable froma pointof viewthatemphasizes merelsr the relationof elements enteringinto a "totemic"complex? In advancingthese comments,the writeris fullv awareof the fact that he maynot havefullygraspedDrGoldenweiser's meaning. The subjectof totemismis not yet quite in the positionof those metaphysical problemsof whichCliffordhassaidthat, in discussing them,peoplefindit peculiarlydifficultnot only to makeout what anotherman meansjbut even what they mean themselves. But that it is peculiarlydifficultto discoveranotherman'sconception of totemism,is amplyattestedby the recenthistoryof ethnologv. However-thismay be, Dr Goldenweiser himselfknowsquite well that his analyticalstudy is not definitive,but programmatic; that the next step mustbe a moreextensiveethnographic investigation of the field. What he has alreadygiven is a statementof first principles. Whateverdeficienciesmay be foundin his definition, he has beenthe firstto showat length,andwith irrefragable logic, that totemismcan not be treatedas an integraldatum, the first, LOWIE] OF TOTEMISM A NEW CONCEPTION 207 as alreadystated, to apply the doctrineof secondaryassociation to the subjectof his inquiry. Fromthis point of view, his paper constitutes a landmarkin the history of totemic study, the prolegomenato all positiveattemptsat a sane interpretationof "totemic" institutions. MUSEUMOF NATURALHISTORY, AMERICAN NE\07YORKCITY.