Structure, Economy and Residence: A

STRUCTURE, ECONOMY AND RESIDENCE:

A RE-EXAMINATION OF NORTH AMERICAN

PATTERNS OF RESIDENCE

by

SAMUEL W. CORRIGAN B.A., The University of Manitoba, 1962

A THESIS SUBMITTED IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF ARTS

in the Department

of

Anthropology and Sociology

We accept this thesis as conforming to the required standard

THE UNIVERSITY OF BRITISH COLUMBIA August, 1964 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of

British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that per• mission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that, copying or publi• cation of this thesis for financial gain shall not be allowed without my written permissions-

Department of Anthropology and Sociology

The University of British Columbia, Vancouver 8, Canada

Date August 26, 1964. ABSTRACT

This paper is basically a re-examination of the existing ethnographic literature concerning Indian tribes in four subsistence areas of North America. The purpose is twofold: to outline the principles governing the size and composition of local groups, and to draw distinctions among rules and patterns of residence.

I suggest initially that patterns of residence are a key factor in the analysis of local group composition; that such patterns are at least partially rooted in ecological factors; that residence patterns will be similar in their effects on local group composition within given subsistence areas; and that major differences among residence patterns and the composition of local groups will be found only among residence patterns and local groups of different subsistence areas.

The tribes of the Northwest coast region were found to have corporate local groups and definite cultural preferences for permanent residence by adult males in those local groups in which they enjoyed the greatest advantage, in terms both of material wealth and prestige. This was termed a preferred rule of residence. Because of the preferred patterns of patrilateral and matrilateral cross cousin marriage, and the elaborate complex of status and rank, this was often the ii local group of ego's mother's brother, that is, avunculocal residence. More commonly, however, this would be the local group of ego's wife, that is, matrilocal residence.

The Sub-arctic region was divided into two areas. In the east, the local groups were corporate and residence was ideally patrilocal, i.e. with the local group of a man's father or brothers• This was based largely on status consi• derations and was termed a prescribed rule of residence. The ideal pattern was not always possible due to ecological factors however. In the west, local groups were not corporate. There was no ideal pattern of residence, although there was what was termed a statistical regularity of patrilocal residence.

Local groups were not corporate on the Plains.

Statistical regularities of both patrilocal and matrilocal residence were found, but these did not normally affect local group composition to any significant degree. The only ideal of residence was that of affiliation with a local group in which one had some kin.

In the Great Basin region, local groups were not corporate. Only statistical irregularities in residence pattern were found, due to ecological factors.

Several common elements were noted in the study.

Descent systems had little effect on local group composition, other than by establishing a dichotomy of kin and non-kin. iii

Local groups tended to be bilateral groups of kin, on the

Northwest coast due largely to sociological factors, and intthe Sub-arctic and Great Basin regions because of situ• ational factors. The local groups of the Plains region were clusters of bilateral groups of kin. Local groups were found to be limited in size, both maximum and minimum, by such factors as ecology. Within those limits precise patterns were based both on sociological factors and on such aspects of ecology as demography.

The final chapter of the thesis is a general discussion of the various factors affecting both local groups and residence patterns. iv

TABLE OF CONTENTS

CHAPTER PAGE

I. INTRODUCTION 1

II. NORTHWEST COAST...... 28

The Northern Northwest Coast 31

The Southern Northwest Coast • • . 45

III. THE SUB-ARCTIC 60

The Eastern Sub-arctic ...... 60

The Western Sub-arctic ...... 74

IV. THE PLAINS ...... 91

V. GREAT BASIN...... 112

VI. CONCLUSIONS . .125

The Local Group...... 125

Local Group Size 126

Local Group Composition. • •••••••• .129

Local Group Nature .133

Patterns of Residence • . 138

The General Hypotheses ••••• 141

The Specific Hypotheses . . . . .146

BIBLIOGRAPHY CODE .152

BIBLIOGRAPHY ...... 154 V

LIST OF FIGURES

FIGURE PAGE

I. Pa trilateral Cross Cousin Marriage 43

II. Ma trilateral Cross Cousin Marriage 43

III. Sister Exchange Marriage. . 44

IV. The Ancestral Family, the Kindred and the

Local Group 58 CHAPTER I

INTRODUCTION

This paper is basically a re-examination of residence patterns in North America. It was inspired by a dissatis• faction with current ideas on the subject and is an attempt at some new thinking on an old and stubborn problem. The results are purely tentative and indeed far less important than the methodology utilized in the study. It is to be hoped, however, that the very fact of new thinking, of a new type of analysis, will lead to more concrete results than previous forms have yielded.

I have formulated two general and four specific hypotheses for testing purposes. It must be bom in mind, however, that there could be a possible methodological error which could well serve as an invalidating factor against the idea of new thinking and new analysis, as stated above, unless the hypotheses are rigidly controlled. The use of hypotheses is a help in any study in the sense that they limit the field and must result in straight "yes" or "no" answers. Nonetheless, in a study based upon the re-examination of well known materials, hypotheses can prove to be a defeating element simply by virtue of their limiting the field 2. of enquiry; however, hypotheses were deemed advisable and are used here although they have been designed as broad outlines only and will be used simply to set the stage and to examine it again at the conclusion of the study. They are conceived of as an organizing tool, as it were, and will be referred to only briefly in the major part of the work.

Edmund Leach has recently suggested (1961b:3), with some justification, that there are very serious limitations to the Radcliffe-Brownian method of scientific analysis by mere comparison and the creation of never-ending typological series of structural forms. Some of his recent work involves another method instead, that of generalization, the examination of data concerning several aspects of social structure for a number of societies with the hope of identifying common patterns of structural relevance in human behaviour. It is this latter method which I propose to follow in this thesis.

The method represents a departure from the methods of anthro• pological analysis, particularly of social organization, normally used in North America. This specific project pro• vides an illustration of the difference between the two methods of scientific analysis.

Until a few years ago, two principal works formed the major body of data on residence theory and North American rules 3. and patterns of residence, George Peter Murdock's Social

Structure and Harold E. Driver and William C. Massey's

Comparative Studies of North American Indians respectively.

More recently, a journal, Ethnology, has begun to classify all world societies into a series of "known cultural types", using a series of typologies. The bulk of this work so far has been concerned with North American Indian societies. The first book, Social Structure.was based on an examination of some of the literature on some two hundred and fifty societies around the world, and resulted in the breakdown of the residence patterns of these societies into what are virtually three basic types and three sub-types. The former were termed patrilocal, matrilocal and bilocal, and the latter matri-patrilocal, avunculocal and neolocal respectively.

To these patterns was attached the term "rules of residence".

Although Murdock was careful to preface this phrase with

"culturally preferred" (1960: 16), he had in fact catalogued these forms in a very basic manner, placing statistical facts above behavioural patterns and failing almost totally to distinguish minor variations of the same basic social structure in relation to "general economic, social and cultural conditions" (Murdock, 1960: 17) in wide areas of the world.

In short, he compared social structures, including residence patterns, found some differences and attached labels. 4.

Sixteen of the societies under consideration in this paper have been so typed.

Driver and Massey followed this example by abstracting these six types and sub-types from the literature on North

American societies (1957: 400-401) and applying the terms in a blanket disposition of the Continent. The method in this case was exactly the same as that of the earlier work: the comparison was of social structure only.

I have one major criticism of these two works, then.

Although both studies took the initial step of examining the relevant literature with reasonable care, neither one looked for common elements of behaviour in certain spheres, pre• ferring instead to examine only structure and to catalogue by a single standard, that of locality. I suggest that these examinations might have produced quite different results than they actually did if they had been carried further to include studies of human behaviour and the reasons for that behaviour, rather than just the "algebraic and numerical indices" of social structure (Leach, 1964: 429).

There are two basic purposes to this study, both of which involve an examination of the ethnographic literature concerning four selected North American subsistence areas, and one of which will utilize specific hypotheses. The two 5. purposes are as follows:

1) I wish to demonstrate that local group

composition is based on a combination of

sociological and situational factors, and

that in basic design, local groups do not

vary greatly within given subsistence

areas, either in the situational factors

which limit them in certain respects, or

in the sociological factors which extend

them along certain lines. Further, I

hope to show that those major variations

which do occur among local groups are

based jointly on differences among sub•

sistence areas and on cultural factors

(which are themselves based partly on

aspects of subsistence).

2) I wish to outline a series of distinctions

among "rules" and "preferences" of specific

residence patterns. These distinctions will

be based partly on sociological and partly

on situational factors, and will be tested

by means of four hypotheses concerned with

both the theory and the practice of North

American patterns of residence. 6.

Before listing my hypotheses, it would be appropriate

to review some of the major theoretical considerations on

residence.

Residence is normally examined from the point of view of the male, and residence typologies are thus listings

of the place of residence of a male following a change in his residence or in his position in the social spectrum of

a society, in relation to his former place of residence.

There are two points here which are certainly as fundamental

as the actual place of residence: the relationships of each of the various social units of the society to the society as a whole, and the circumstances of the change of social position.

For the first, it need only be stated that there may be a vast difference between a change of residence from one family

dwelling to another in the same community, and a change from one community to another. In the latter case a change can usually be charted quite clearly. In the former, however,

there might not be any significant change in place of resi• dence, but only a change of status. For example, a man might marry a woman living in a house next door to his own and continue to reside in his own dwelling, bringing his wife

there. For both persons there is a major change of status, but for neither is there any significant change of residence. 7.

Nonetheless, the couple would normally be said to reside patrilocally. The same label, however, would be attached to the couple who resided in the man's pre-marital dwelling after marriage but who had formerly lived some hundreds of miles apart. This bring me then, to the next point,*: the meaning of residence and residence typologies.

Any analysis of residence is basically a consideration of the compositions of local groups of people, of geographi• cally isolated or socially distinct groupings. This being the case, we are not actually concerned so much with a possible change of residence by specific individuals, as with the composition of the local group through a given series of moves by certain individuals. For example, all the men of a given local group may be born and raised there, and spend the whole of their lives with that group. Or, the group may be composed of men under the age of, say, twenty, who were all born there and men over the age of twenty who moved there from other local groups. Similarly, the female population of a local group must be considered and charted as to possible moves. In other words, by examining the local group through time, as well as individuals through time, we may obtain a more complete picture of changes in the structure of a community. 8

Over the past fifteen years there has been a consider• able amount of interest in the problems of residence, and a number of exploratory articles. These have usually centered around two focal points, the place of residence and the meaning of the concept, and have tended to polarize into listings of types of residence patterns and general discussions of local group composition.

Murdock provided the major typology of residence now used in most cross-cultural and ethnographic studies. He noted that marriage initially results in the "dislocation of a residence" (1960: 16) and based his typology on the sex of the person moving and the place to which they moved at time of . marriage. Each of his six basic types was termed a

"culturally preferred rule of residence" (1960: 16). His three basic types were described as follows:

"If custom requires the groom to leave his parental home and live with his bride, either in the house of her parents or in a dwelling nearby, the rule of residence is called matrilocal. If, on the other hand, the bride regularly moves to or near the parental home of the groom, residence is said to be patrilocal. It should be emphasized that this rule implies, not merely that a wife goes to live with her husband, but that they establish a domicile in or near the home of her parents. "Some societies permit a married couple to live with or near the parents of either spouse, in which case such factors as the relative wealth or status of the two families or the personal preferences of the parties to the union are likely to determine whether 9.

they will choose to reside xnatrilocally or patrilocally. The rule of residence in such cases is termed bilocal." (1960: 16)

His three sub-types (which he did not describe as such) were termed avunculocal, for those societies "which prescribe that a married couple shall reside with or near a maternal uncle of the groom" (1960: 17), matri-patrilocal, described as "matrilocal residence for an initial period, usually for a year or until the birth of the first child, to be followed by permanent patrilocal residence" (1960: 17), and neolocal,

"when a newly wedded couple. • . establishes a domicile independent of the location of the parental home of either partner" (1960: 16). The two key points about this typology are that the types are considered rules. and that they are based on the change and location of residence following marriage.

In 1957, Driver and Massey noted that few, if any, of the Indian tribes of North America followed any specific pattern of residence unanimously. They suggested that for most tribes "a rule of residence was merely a preference or a tendency subject to change according to a variety of disruptive circumstances" (1957: 403). These writers retained Murdock's terminology and typing system, however. 10.

Bohannan argued against the concept of residence as

referring to marriage, insisting instead on its use in

connection with the household. The latter, he noted was "a

group of people who live together and form a functioning

domestic unit" (163: 86). It is the basis of the local

group which is of primary consideration in any study of

residence, whereas the family, the unit based on marriage, is

simply "a social group that may or may not be a local, spatial

group" (1963: 98).

In 1963 also, Dunning suggested that a unilocal pattern

of residence constitutes a rule and "is an imperative (whether

prescribed or strongly preferred) based on sociological con•

siderations" (1963ms: 13), but that any other pattern repre•

sents the absence of a formal rule. Beyond this, he implied

that in societies without a rule, situational factors such as

demography and ecology override any sociological consideration narrowing the range of possible choices of residence.

Several factors relevant to the discussion which

follows emerge from these four papers. Firstly, as noted earlier, we are concerned in any study of residence with the composition of the local group, from the basic unit, usually

the household, to the largest spatially distinct unit.

Secondly, although marriage is a primary factor in the 11. changing configurations of local groups, other social activities must be considered also. The most notable example, of course, is the practice common on the Northwest coast of a boy going to reside with his mother's brother at about the age of ten or so. Thirdly, it is incorrect to speak automatically

of patterns of residence as rules without qualifying the term

in some way; they may or may not be actual rules of behaviour

as the word is normally defined. And, fourthly there are a variety of factors, both sociological and situational, which

determine a precise pattern of residence.

The initial purpose of this research will be an

attempt to understand these first two points more clearly.

The second purpose will be a discussion of the third point.

The fourth point underlies the entire project.

With one exception, none of the works cited above has

touched on the actual crux of the problem, however, the

reason for a particular form or pattern of residence above

others. The exception was Dunning, who noted one major dis•

tinction in such patterns by pointing out that the ecology of

a group may prohibit certain regularities in residence

pattern. I will briefly outline some of the factors involved.

It is best at first to note certain patterns of residence which may actually occur. 12.

There are, for example, formal rules of residence which appear to require an individual male to reside with a specific local group. The most common focal point for this type of group is corporately-held land. Among the Yako, all land is held by groups of patrikin, and a man must normally reside with his patrikin if he is to obtain any land (Forde, 1950: 294).

There may be other reasons, however. The Lele are divided into local groups of matrikin holding rights over certain women.

To ensure both a wife and membership in a group holding wealth corporately (for these rights amount to a form of property), a man ijtust join the local group of his matrikinsmen

(Douglas, 1960: 251-252).

It may occur that a society has no form of corporately- held wealth, and yet the males tend to join together in certain local groups in fairly regular patterns. In this case, we can speak of a statistical regularity in the absence of any rule.

The Plateau Tonga, for example, tend to gather as groups of matrikinsmen, and yet changes of residence are frequent and the composition of local groups is highly fluid (Colson,

1962: 181-183). This is simply the operation of a behavioural norm with a quite valid sociological basis - one associates primarily with one's kin rather than with strangers - but with no specific focal point to permit the development of local corporate groups. 13.

Finally, there may be a specific pattern of preference in residential alignment, but a number of situational factors, such as demography and ecology, which tend to limit the actual operation of that preference to the point where the pattern is one of statistical irregularity. Among the Lozi it is consi• dered good to reside in the local group of one's father or father's kin. The size of Lozi villages is governed by fairly strict ecological limitations, however, and it may be necessary for a Lozi youth to settle in the local group which is best able to accomodate him (Gluckman, 1950: 172).

Dunning made a major distinction when he noted the difference between the rule of unilocal residence based on sociological factors and the absence of any rule of unilocal residence due to situational factors. While the dichotomy is valid, it does not go quite far enough. There are, in fact, several more distinctions which must be made.

On the basis of the foregoing material, then, I propose that there are four forms to be distinguished within the general concept of residence, and several possible patterns of residence, each one governed by slighflly different circum• stances. Two of these forms are termed rules, while two represent the absence of any specific pattern which I would ternua rule. The hypotheses for this paper are based on 14. these distinctions, and are as follows:

1) There are only two categories of actual rules

of residence. These are the prescribed rule,

which carries positive sanctions in the form

of property, titles or some other major asset

with a consequent influence over other individuals

if it is followed, and negative sanctions in the

form of the refusal of a local group to allocate

rights in these resources to the individual who

does not follow the rule, and the preferred rule,

the socially approved pattern of residence which

acts as an ethical force and enjoins compliance

by members of a society, but which does not carry

direct sanctions of any kind. These rules are

found only in those societies which have corporate

local groups, although not all societies with

corporate local groups will have rules as they are

defined here. The exceptions will be due to

ecological factors prohibiting any regularity of

pattern.*" The degree of rule - whether prescribed

or preferred - depends upon the precise ecological

conditions of the society. If there are situ•

ational factors of sufficient intensity to prevent 15. the utilization of the rule in a fair number of cases, so that some accomodation to the rule is necessary fairly frequently, then it will be preferred, and there will not be any direct sanctions involved if it is not followed, If, on the other hand, such situational factors are absent, then sanctions will operate in cases of violation of the norm and the rule will be prescribed.

Three forms of these rules can be noted, patrilocal, matrilocal and avunculocal. Patri• local implies a group of patrilaterally-related kinsmen as the core of the local group. This would mean the permanent settlement of a man in the local group of his father, the group in which he would normally have been born and raised. The basic kinship ties would be father--son and/or male siblingship. Matrilocal implies the opposite, a groupofmatrilaterally-related kinsmen as the core of the local group, and mother-daughter and/or female siblingship as the major elements of kinship among adults of the core group. This usually means the permanent settlement of a man in the local group 16.

of his wife following marriage. Avunculocal

implies a group of matrilaterally-related kinsmen

as the core of the local group, but with a mother's

brother-sister's son relationship, male/female

siblingship and/or male siblingship as the major

elements of kinship among the adults of the local

group.

2) Those patterns of residence which have hitherto

been termed "bilocal" and "neolocal rules of

residence" are not rules at all but simply patterns

of behaviour in the absence of any social rules.

They may be divided into two categories, the

statistical regularity, under which the society

shows a statistical balance in favour of one of

the unilocal patterns of residence, and the

statistical irregularity, the absence of any

significant statistical pattern of unilocal

residence. The former, the statistical regularity,

is found only in those societies which lack

corporate local groups. The choice of residence

is normally governed by what can be termed

"customary behaviour", but there are no sanctions 17.

operative in cases of failure to follow the

customary pattern. The latter, on the other

hand, is found in societies which may or may not

have corporate local groups. In this case, choice

of residence is based on purely situational

factors such as demography and ecology, although

there will normally be a customary preference.

If any specific terms must be used for these forms

of residence, they may be termed the (unilocal)

pattern of residence and the irregular pattern

of residence.

As noted earlier, the primary concern in any study of residence is the composition of the local group. When consi• dering rules and patterns of residence then, it is changes in local group structure through intermarriage or other moves of residence, that are important. Very broadly speaking, changes in residence within the local group do not concern us here.

Thus, in cases of marriage, or in other cases where a change of residence occurs within a single local group, it is sufficient to note any sociological preferences expressed for and of the pattern of residence. The patterns and rules considered in this paper, however, will be those concerned with changes in the composition of local groups. 18.

Notes and Queries on Anthropology defines the local group as "any aggregate of people, who inhabit a single clearly demarcated locality and who regard themselves and act as a unit in relation to other local groups" (Royal

Anthropological Institute, 1951: 64). Combined with this, we can consider the description of the local group by several

American anthropologists: "the maximal group of people who normally reside together in face-to-face association"

(Murdock, Ford et al, 1945: 29). This means that the local hunting unit, often smaller than the band, would be the local group among the Plains Indians, Sub-arctic Indians and Great

Basin Indians, while either the village or, in some cases to be defined later, a segment of the village, would be the local group among the Northwest coast Indians.

I have used the term corporate in this paper several times and it would be wise to define it before continuing.

Ottenberg and Ottenberg defined the corporate group as "a group of persons having an identity, rights and obligations distinct from those who make up the group" (1960a: 28). This was based on Radcliffe-Brown's definition, under which a group was corporate if it possessed "any one of a certain number of characters: if its members, or its adult male members, or a considerable proportion of them, come together occasionally to carry out some collective action - for example, 19. the performance of rites; if it has a chief or council who are regarded as acting as the representatives of the group as a whole; if it possesses or controls property which is collec• tive" (1950: 41). In addition, some writers have stressed continuity over time as a necessary criterion for such a group (Fortes, 1953: 27; Fried, 1957: 23). It is necessary, however, to distinguish between corporate activity by a defined corporate group and activity which is not corporate by members of that group. As an example, Leach noted that one of the major social tasks of a corporate descent group might be the negotiation of marriage for one of its members, but added that if a descent group ceased "to be a localized group, then, in general, it (jcease^) to be a corporate unit for the purposes of arranging a marriage " (1961c: 56).

Under Radeliffe-Brown's and the Ottenbergs' defini• tions, the local group, as the term is used in this paper, is always corporate. I have, however, selected a single criterion of corporateness over and above the Ottenbergs' definition, and will use the term corporate group only for those groups which own some property, including land, commonly ascribed to the group. In many cases this property may actually be owned by specific individuals; nonetheless, if the property is considered the province - whether by future inheritance rights or simply by association - of all members of the group, and 20. if all members are ascribed some status through that asso• ciation, then the group will be considered corporate. It will be noted that land, in the form of resource use rights and usually with some powers of control over individuals by

the holders of the land, is the most common locus of corpor•

ate behaviour over a period of time. In addition, wealth

tends most commonly to be associated with land, either

directly, as in the case of storable surpluses, or indirectly,

as in the case of titles associated with specific localities.

Land also implies a relatively sedentary population with

some form of permanent village and self-renewable exploited

resources (such as a fur bearing animal population or a fish migratory route), although these are not essential charac•

teristics of the corporate local group.

For this paper, what I have termed local groups among

the Northwest coast and Sub-arctic Indians are corporate groups; i.e. their economic behaviour is corporate and based on land rights in part, and much of the social and political behaviour of the groups is intimately connected with these rights. They are not corporate on the Plains or in the Great

Basin, however. In neither of the latter regions did a single

local group lay claim to a specific tract of land for economic exploitation under a concept of ownership. 21.

There remains one problem in this treatment of resi• dence rules and patterns, that of the endogamous local group. Murdock proposed the concept of the endogamous deme. which he described as "the endogamous local community which is not segmented by unilinear consanguineal groupings of kinsmen.

By virtue of the rule or strong preference for local endogamy,

the inhabitants are necessarily related to one another through intermarriage, although they cannot always trace their exact kinship connections. They are consequently bound to one another not only by common residence but also by consanguinity, as is, in fact, usually specifically re• cognized. Within such a group the only social structuring is commonly into families, which may be of either nucleur, polygamous, or extended type. Except for family ties, the strongest sense of identification is usually with the community as a whole, which is viewed as a consanguineal unit in relation to other communities" (I960: 62-63).

It would seem logical to suppose that such endogamous local communities would normally engage in some corporate activities. It is not by any means axiomatic, however, that they would be corporate as the term is defined here. A fundamental question is the relationship between the local group, as I use the term in this paper, and the endogamous 22. deme. Local groups, as they have been outlined for the

Northwest coast-'Sub-arctic and Great Basin regions, tend

to beiexogamous by kinship. It may well be then, that the

local groups described here are smaller than the endogamous deme and some attention will be paid to this in later chapters. The major problem concerns the utility of the idea of unilocal residence for a single endogamous community. It is conceivable that the concept is misleading, but again, this will be considered later.

Murdock describes three of the societies which are

included in the sample for this paper as having endogamous dernes, the Commanche, the Kiowa-Apache and the Shoshone

(1960: 63). Special attention will therefore be paid to these

tribes in the body of the paper.

For testing purposes I have formulated four hypotheses relating to specific economic areas in North America, pro• posing the form of residence which I expect will be found in each region. They are as follows:

1) Northwest Coast

This is the only one of the four areas in

North America in which ecological factors permitted the develop• ment of prescribed rules of residence in given social situations. I suggest that patrilocal or avunculocal patterns occurred in all societies of this area, and that in each case 23.

these were prescribed rules. Several factors combined here to make a prescribed rule the norm. Firstly, wealth accumulation

was possible on a major scale, and there were class differences

based in part upon this wealth. Secondly, the local group

was corporate and based on kinship, and a specific rule of

residence was a requisite for the maintenance of the corporate

group of kin. Thirdly, the development of a specific socio•

logical configuration in these societies, with a hierarchy

of prestige positions and a fairly rigid complex of behaviour

governing the attainment and validation of social positions,

contributed to the desirability of a specific unilocal pre•

ference in residence. Finally, the ecological factors of the

region, in combination with the nature of the local group and

the village, did not require any great flexibility in

residence pattern. Variations from the patterns were not

institutionalized norms based on ecological factors. I suggest

that the specific unilocal rule will follow the inheritance

and succession patterns of the local groups in each society,

either partilateral or matrilineal': through males, with

patrilocal and avunculocal patterns of residence.

2) Sub-arctic

In this area there was a preferred rule of

residence. The local groups were corporate and the unilocal 24. pattern of residence was patrilocal. There were two reasons for this patrilocality, the patrilateral inheritance of hunting territories within the local group and the ecological necessity of familiarity with the local group area by the hunter. It was not a prescribed rule. Wealth accumulation was not possible on any major scale within this region, and an individual had consequently to depend upon readily ob• tainable resources for subsistence. Storable surpluses were largely unknown. The ecological balance of such a region was precarious, however, and only small local groups could exist.

Flexibility in residence pattern was an essential element in the ecological adjustment of these tribes as a means of preventing demographic imbalance in the local groups. The combination of corporate land-holding groups, patrilateral inheritance of hunting territories and the ecological necessity of some flexibility in residence pattern combined to help produce a patrilocal preferred rule of residence.

3) Plains

This region, the tribes of which lacked corporate local groups, had only statistical regularities in residence patterns. There was a sociological preference for residence in a local group in which the individual had some kin, either affinal or consanguineal. The nomadic nature of the local groups, based on a migratory game subsistence 25. economy, worked against the development of any concept of land ownership, and hence against the development of cor• porate local groups. The absence of corporateness, in combination with bilateral descent patterns, contributed to enough freedom of choice to permit the growth of more than one unilocal pattern of residence in statistical regularity in each tribe. The lack of any necessity for rigid population control to limit the sizes of bands and local groups in the region permitted such regularities to occur.

4) Great Basin

Here the absence of corporate local groups meant the absence of any rules of residence. The tribes of the region were nomadic in an area of sparse food resources for local groups of this size. Subsistence was based largely on the collection of vegetable products in scattered areas at different times of the year. The limitations and vagaries of the food supply combined with the nomadic move• ments of the local groups to prevent the development of corporateness in local groups. In addition, the small size of the local groups and the tight limits placed on their size by ecological factors, together with an ecological necessity for a certain general demographic structure in each local group, meant a necessary flexibility in residence 26. pattern. The over-all result was a lack of any statistical regularity of any significance in residence patterns.

The method of this paper will be as follows. I will first of all outline the patterns of residence and certain facets of the economy and social structure of thirty tribes in four selected subsistence areas, describing the patterns to as great an extent as possible with direct quotations from the ethnographic literature. I will, for each area, provide a very brief analysis of the patterns outlined, noting the general pattern which obtains in the area and the basis of it. Each of the four area chapters will be largely descriptive with some analysis to provide a clear picture of the pattern of the area. I will then consider each area in a concluding chapter, hypothesizing the form of local group and the pattern of residence for the area. I will outline the general considerations with regard to these two factors, demonstra• ting their basis in economy and social structure. Finally,

I will conclude with suggestions for further areas of investigation and analysis. 27.

FOOTNOTES: CHAPTER I

One example is the Lozi, noted on page 14. The local group - the village - is corporate in the holding of land, but no specific rule of residence is possible due to the ecology of the society. Gluckman notes that "in practice all land is held by villages. . . If a man leaves a village he loses his rights in all land which he worked as a member of that village" (1951: 64). However, "Lozi villages can be inhabited only by small groupings of kindred. . . In a small family village. . . It is impossible for the headman to place all his own sons and brothers' sons on land. Therefore some of them seek new land from the king or beg land from other relatives" (1950: 172). CHAPTER II

NORTHWEST COAST

The seven tribes of the Northwest coast, ranging from southern Alaska down to Washington, developed a complex system of rank and status in an area of relatively isolated villages and small language groups. They were able to do so largely because the region was bountiful in natural re• sources. Their villages were built on the beaches of rivers and inlets, and along the coast in sheltered spots. Daryll

Forde noted that

within the range of any village group, there was a considerable number of districts and sites visited in turn for sealing, fishing, gathering and hunting at different seasons of the year. Places at which seals congregated along the coast in spring; portions of the rivers particularly suited to the erection of salmon weirs; tracts on which there were rich harvests of berries and roots and forest ranges frequented by deer, elk and mountain goat(1963: 92).

Sapir outlined two areas of the economy in some detail when he stated that

the streams teemed with various kinds of salmon throughout the year, and the sea offered a great variety of edible sea-mammals and invertebrates. It was thus possible for a rather large group of people to make a comfortable living in a quite restricted bit of coast territory. Access to the sea at a few points and the control of a few 29

streams at which the community could follow the salmon at their spawning periods were all that was needed to insure ample means of subsistence for all. Furthermore, the unusually great rainfall of the coast country made it necessary for the Indians to house themselves in substantial shelters, and at the same time gave them the ready means wherewith to fill this want. . . The inexhaustible supply of readily worked wood, particularly the red cedar, gave the Indians all that was necessary for the building of large houses. (1915: 357).

One of the results of these plentiful natural food resources was a high concentration of population, with an estimated total of some twenty-two thousand to twenty-seven thousand people in the three northern tribes, with a population density of .4 persons per square mile (Kroeber, 1963: 135; Jenness,

1960: 335; 339), and from twenty-eight thousand to thirty-six thousand in the four southern tribes, with a density of between .6 and .7 persons per square mile (Kroeber, 1963: 135;

Jenness, 1960: 339, 345, 347, 349).

Although these tribes all inhabited similar geographic areas, there were certain basic differences in their social organization. Accordingly, they will be considered in two separate groups, the Tlingit, Tsimshian and Haida of the north, and the Bella Coola, southern Kwakiutl, Nootka and

Coast Salish of the south. Because of the paucity of data available, the northern Kwakiutl will be omitted. 30.

The entire Northwest coast region was characterized by a single basic division of society, with all people either freemen or slaves. The latter were of relatively little importance socially, and will not be considered in any detail here. In all probability their numbers were rela• tively small, and they may best be treated as an economic asset rather than as a significant factor in the social structure of these tribes. Among the freemen, there was a grading of status, so that each individual occupied a separate rung on a hierarchical ladder of rank. As Drucker notes

the difference Cin status]) lay in extent of participation, not kind. One less high than the highest in rank, participated less fully in ostentatious activities. A person a grade above the lowest participated in these a bit more than the one on the bottom rung. And thus the manifestations of statuses of high and low degree shaded into each other. (1939: 56-57)

He added that "status in its minimum terms - membership in society - was derived from kinship and expressed in terms of wealth" (1939: 61).

Finally, one further trait was characteristic of all these tribes, the composition of the local group, which was described by Drucker as " a group of people sharing rights to the utilization of economically important places and occupying a common village" (1939: 58). It was almost always 31.

no more and no less than an extended family (slaves of course excluded) and was so con• sidered by its members. The individual of highest rank in the social unit was related to the lowliest (1939: 58).

The head of each local group, was the commonly-termed "chief", the individual of highest social, economic and political status in the unit. Each local group possessed rights of use over specific territories, although ownership was usually vested in a single individual. It is to be noted, however, that "individual ownership in these cases does not mean exclusive rights of use, but a sort of stewardship, and the right to direct the exploitation of the economic tract by the local group. The latter it was who held exclusive rights" (Drucker, 1939: 59).

The significance of these two factors, rank and the corporate nature of the local group, to the residence patterns of these tribes will become clear in the descriptive sections of this chapter.

The Northern Northwest Coast

The Tlingit Indians extended along the coast of Alaska and British Columbia, inhabiting both the mainland and many islands, for a distance of several hundred miles. Kroeber estimated their population at ten thousand at the time of 32. early European contact (1963: 135), although Jenness noted that they had decreased to less than six thousand by 1835

(1960: 331).

The Tsimshian Indians of the northern coastal mainland of British Columbia were divided into three distinct sub- tribes, each with a similar social organization, but with slight linguistic differences among them (Jenness, 1960: 337).

Kroeber estimated their population to have been about seven thousand all told at the time of early European contact

(1963: 135).

The Haida Indians inhabited the Queen Charlotte Islands and several of the southernmost Alaskan Islands, but maintained social contact with several tribes up and down the coastal mainland of British Columbia and Alaska. They are reported by

Kroeber to have numbered about nine thousand eight hundred in the early European period (1963: 135), but Jenness suggests that Newcombe's estimate of about eight thousand four hundred is more accurate (1960: 335).

These three northern coast tribes all had matriliheal descent systemsdirectly connected with a complex pattern of economic and political alliances. All were divided into a series of matrilateral clans grouped into exogamous phratries

(Jenness, 1960: 144). Generally speaking, the clans were not localized, but each village normally contained more than 33. one clan. Each localized segment of a clan possessed its own

territories and houses, and functioned as a corporate group.

It is this unit which I shall refer to as the local group.

As Murdock noted, speaking in effect of the localized clan segments in pre-European times,

land, in so far as it (wad) property, (pelonged) to the clan, which (Enjoyed) recognized rights to definite hunting grounds, salmon streams, village and camping sites, unusually abundant berry patches, rocky islands where aquatic birds GL&1

Although the clans - and hence the local groups - were

exogamous, few villages contained only clan segments of one

phratry, and were therefore not normally exogamous. On the other hand, they were rarely endogamous, so that a man would

be likely to have matrilineal:, kin in more than one village.

Inheritance and succession were both matrilineal,, so that a man could theoretically inherit resource use rights in other villages. In practice, however, an individual's rights to

resources rested in the local group of which he was a member.

A man often acquired rights of use over his wife's inherited resources as well.

The local group then, was an exogamous, corporate segment of a matrilineal elan, normally residing alongside 34. other local groups as a part of a village. The chief of the most influential local group was usually the generally

recognized leader of the village, but there were also other

local group chiefs in each village, graded in a hierarchy of status and power. Oberg pointed out that just as the

local groups were graded in terms of kin group status, so too

there was a grading within the group, so that "within the

clan (and local segment) every person (had) his or her rank which (was) definitely known" (1934: 146). The local group

then, was both a "sovereign group" (1934: 152), based on heritable wealth through kinship, and graded within so that it was "weakened by individual status" (1934: 146). In essence, it was united without and divided within.

The ethnographic data concerning Tlingit residence patterns is brief at best. Krause noted the first possible change of residence when he stated that

the estate of a deceased Tlingit goes to the son of his sister. . .The presumptive heir is required during his youth to serve his uncle without payment, but the future inheritance is expected to repay him"(1956: 161)

Since it would be unlikely that a married man would be a member of the same local group as his sister, it is probable that a youth moved to the local group of his mother's brother at a fairly early age, residing with him until his marriage. 35.

He also suggested a pattern of village exogamy (1956: 154) through tribal exogamy, but this seems unlikely in view of several other statements on residence. He reported that a man's mother's brother was likely to be in the same village, and went on to note that "a married man among the Tlingit always has the right to live with his father-in-law or move to his own home" (1956: 154), the latter presumably the local group of his father. Niblack was more definite, stating that a woman "goes to live with her father-in-law" following marriage while Davidson agreed (1928a: 20). Oberg found, however, that "in practice, it is a wife's clan that builds a man's house", indicating a matrilocal pattern (1934: 145).

The utilization of property was of considerable importance to the Tlingit and, as noted earlier, all exploitable property was held by local kin groups. "In the summer. . . they lead a half-nomadic life, for they scatter according to clan and family lines to their hunting and fishing territories" was how Krause expressed this Q.956: 85). Oberg and Swanton both confirmed this, the former noting that the "local division of the clan . . . possesses definite territories for hunting and fishing, (^nd) houses in the village" (1934: 145), while the latter added that the clan subdivisions - the local groups - "might occupy one or several houses" (1908: 398). 36.

The Tsimshian data offers more complete information

on the pattern of the avunculate so often noted for this

area. Garfield stated that if youths resided "with their

father they assisted him, though they were expected to aid

their uncles also. Even though boys did not live with their uncle they spent much time in his house, eating meals there,

playing with the children in the house and assisting in

fishing, wood gathering and other household activities"

(1939: 277). She also noted, however, that local group affiliation - i.e. affiliation to a residence unit - was

"inherited matrilineally as (was) kinship affiliation"

(1939: 175). There were also matrilineal patterns of inheri•

tance and succession (1939: 178). Garfield outlined these patterns in some detail:

The dwellings in which middle class Tsimshians lived were lineage property. Each was built through the cooperative efforts of the members of the lineage to which it belonged. Therefore, every person had a right to live in any dwelling that belonged to his group, since he had helped to build it. The man who held the name of highest rank in the lineage was the head. • .of the group. Headship of the dwellings and occupants, with the name, passed to a younger brother or nephew of the head, upon his death. The building, like other property, was transmitted within the lineage. If there were too many relatives to live in one dwelling, a second was built and headed by the man holding the name ranking next to that of the head of the lineage. . . A woman of the lineage, even though living with her husband in his home, contributed toward that of her lineage relatives (1939: 275-276). 37.

She also noted the composition of the group, stating that

in "the winter dwelling lived the head, his mother if she were a widow, his wife and small children, his widowed or

divorced sisters and their small children, also his younger

brothers and cousins (mother's sister's sons) and his grown

nephews with their wives and children. There would also be other relatives, according to circumstances, but these were

the usual ones represented" (1939: 277). She concluded by

stating that the head of a house - the term here refers to

the local group (1939: 174) - "had direct control over his unmarried nephews" (1939: 277), presumably all those residing in the local group plus the sons of his sisters resident in other local groups.

Boas suggested that if two people from different villages married they lived patrilocally, but that for intra- village marriage, matrilocal residence was possible

(1895: 575; 1916: 532).

The information concerning the Haida is quite extensive but much of the detailed data on residence is of fairly recent date and may be more representative of remembered ideals than of actual patterns. Swanton found that a "man's sisters' sons were his right-hand men. They, or at least one of them, came to live with him when quite young, were trained by him, and spoke or acted for him in all social matters. The one who it 38. was expected would succeed him was often his son-in-law as well" (1905: 70). He noted a pattern of bride service by a youth who "went to live with the girl's family, if not already there, and worked for them until his marriage" (1905: 50).

Following the actual wedding, "the youth's family took the bride home", indicating a patrilocal pattern (1905: 51).

Cross cousin marriages were apparently fairly common, both patrilateral and matrilateral types occuring (1905: 64, 68).

Finally, he stated that town chiefs - the acknowledged village leaders - "seem to have married in distant towns more often than the others, because they could not find anybody "great enough near by" (1905: 68).

Murdock noted approximately the same pattern:

A girl lives with her parents until her marriage and usually even thereafter. A boy, however, leaves his paternal home at about the age of ten and takes up his residence with his maternal uncle either permanently cr until he becomes independent (1936: 249-250).

There was also a period of pre-marital bride service lasting one or two years during which a man normally, but not always, resided with his future father-in-law (1934: 373; 1936: 250).

Following the wedding, "the young couple live for an entire season in rotation at the homes of the groom's sisters, where they are lavishly entertained. They then take up their residence with the bride's father, unless he is dead or the 39. groom is himself a house chief" (1936: 251). Earlier Murdock had noted that residence was "prevailingly matrilocal"

(1934: 374) and seems to confirm this with his outline of the

"typical household" which included "the owner or house chief, his wife or wives, his young sons and unmarried daughters, a daughter with her husband and children, a younger brother of

the chief with his wife and children, an unmarried nephew

(sister's son), a married nephew with his family, possibly some other poor relative, and a slave or two" (1936: 237).

I have noted the local group property patterns and the basic social organization of these tribes, and had earlier outlined the rank-status complex of the Northwest coast

Indians in general because I feel that the residence patterns can only be understood in the context of the particular motivations for specific residence patterns in this area. I suggest that there was a very definite ideal pattern, but

that there were many, almost institutionalized variations from it. I propose first to outline the ideal pattern.

Firstly, a man's wealth, and the wealth of his family and local group, was dependent in large part upon his access to economic resources. If his local group possessed rights of use over hunting territories and fishing sites, all well and good. Those groups which did not had to "resort to the 40. common grounds or wait until the owners of others were through with them" (Swanton, 1908: 425). Clearly, then, a man would settle if at all possible in a local group which possessed such rights, and this would always be a matrilin- eally determined local segment of a clan. There is clear evidence that a boy either moved to or became a participating member of the local group of his mother's brother at an early age. His father then, did not necessarily reside in the local group of his mother's matrilineal kinsmen. After marriage, a man joined the local group of his wife, but her brothers all moved to the local groups of their wivesJ This obviously does not fit the data presented above. I suggest then, a new approach, that rather than establish immediate patterns, we establish the reasons for any possible pattern.

With this in mind, it is clear that there is one dominant pattern in these three tribes: a man had certain obligations to his own kin which he fulfilled by service to matrikinsmen, normally his mother's eldest brother. They in turn had certain obligations to him. But a man also had obligations to himself and his children. He would, therefore, seek to attach himself to the local group through which he - and his children and his matrikinsmen of the immediate descending generation - would gain the most, through property, resource use rights, titles and the general power of manipulation of 41.

other individuals. Post-marital residence, then, involved

a variety of patterns.

If a man's father was a chief, his father's sister

would quite likely reside with them, and a preferred union

would be between he and his father's sister's daughter.

Residence would be both patrilocal and matrilocal, i.e. he

would continue to reside in the local group of both his father

and his wife, and his wife's brother would succeed his father.

Again, if a man's father was a chief, it is possible that his mother's brother would reside with them as well and avunculocal

residence would also obtain (see Figure I). If his maternal uncle was a chief, he would likely be born and brought up in his local group anyway. A preferred union would be between he and his mother's brother's daughter and he would continue

to reside in the same local group. In this case, patrilocal, matrilocal and avunculocal residence would be synonomous

(see Figure II). Since sister exchange marriage was also practiced - its desirability can easily be seen from Figure

III - it is likely that in many cases a man's mother's brother's daughter would also be his father's sister's daughter,

in which case the three forms of residence would again be combined. This latter procedure, of course, was the most practical from the standpoint of kin group wealth and position in that it tended to consolidate privilege in a small group 42.

rather than dessimate it through inheritance to previously

unrelated persons. Finally, if a man's brother became chief

- through any pattern of residence - it follows that he, in

seeking privilege as the second ranking male of his kin

group, would likely reside with him, either avunculocally or

patrilocally, but not matrilocally unless he had married his

brother's wife's sister or some other female of the local

group of his father or mother's brother.

In actual fact, however, as a general principle, men

tended to reside firstly in the local group which was most

advantageous for them, i.e. in the local group through which

they enjoyed special benefits by inheritance or position

through their own matrilineal kinsmen, and secondly, in the

local group in which they were able to assume a position

as an affine of a kin group, that is, in which they exercised

a husband's rights of use over a wife's inherited privileges.

In the first case, it is virtually impossible to chart a residence pattern other than by noting the emphasis on affiliation with the local group of the mother's brother. Even

though this might not involve any change of residence, it can be referred to as avunculocal in nature. In the second case,

the pattern of residence may be termed matrilocal, in that the man joined the kin-based local group of his wife. Although both of these patterns represented norms of behaviour, it could be misleading to refer to them as culturally preferred 4-3.

O • i — A EGO

/= lines of inheritance A= house chief

FIGURE I: A LOCAL GROUP - MARRIAGE TO FATHER'S SISTER'S DAUGHTER; PATRILOCAL, MATRILOCAL AND AVUNCULOCAL RESIDENCE.

FIGURE II: A LOCAL GROUP - MARRIAGE TO MOTHER'S BROTHER'S DAUGHTER; PATRILOCAL, MATRILOCAL AND AVUNCULOCAL RESIDENCE. 44.

A = O .A = O A

o O = A A

A = O 6 =>A A

A^= O 6 = A A

/ /= lines of inheritance A= house chiefs over time /

FIGURE III: SISTER EXCHANGE MARRIAGE; A FOUR GENERATION LOCAL GROUP. 45. rules per se; the only preference involved was that of residence with the local group in which one could achieve the greatest standing, socially, economically and politically.

This may or may not have involved a change of residence for either a man or his wife, depending upon their particular status and the power and position of their respective matri-

lineal kin groups.

It will be noted, incidentally, that cross cousin marriage was a preferred form, but that it actually occurred in only a minority of cases. For this minority avunculocal residence, as outlined above, was the standard procedure. For the majority, however, this was not practical following: marriage, and matrilocal residence was normal. In the case of marriage between slaves, residence would almost certainly be determined by the owner or owners of the slaves. For a union between a slave and a free person, the particular cir• cumstances would dictate the place of residence. It is to be noted, however, that inter-class marriages were frowned upon both for their social implications and for their politico- economic ramifications (Garfield, 1939: 232).

The Southern Northwest Coast

The Bella Coola Indians inhabited the area of two deep rivers, the Dean and the Bella Coola, and the inlets 46.

leading off them, occupying over forty villages along their

shores. They were almost surrounded by the Kwakiutl, who

occupied the territory both north and south of them along the

coast. The southern Kwakiutl, divided into several dialect

groups, ranged from the mainland Bella Coola area south to

Discovery Passage. They also inhabited the northeastern

section of Vancouver Island. The Nootka lived on the western

half of Vancouver Island and, like the Kwakiutl and the

Bella Coola, spoke a Wakashan language. The rest of Vancouver

Island and the Gulf Islands, together with the mainland down

through Puget Sound, was inhabited by Coast Salish groups,

in a number of separate villages with several different

dialects of the Salish linguistic family. Kroeber estimated

the population of the Bella Coola at about fourteen hundred

at the time of early European contact, and that of the Nootka

and the southern Kwakiutl at about six thousand and five

thousand nine hundred respectively (1963: 135). Mooney

placed the number of Coast Salish in Canada and the United

States at about fifteen thousand in 1778 (1928: 28). Driver and Massey listed all of these tribes as patrilocal in residence pattern (1957: 402), while Murdock termed the southern Kwakiutl patrilocal (1960: 230). 47.

Drucker described the social organization of this

sub-region as follows:

There were no moieties, clans or lineages. Descent was reckoned bilaterally, with only a slight preference for the male line. It is true that most Heiltsuk (Kwakiutl) groups . . . had divisions named after the Eagle, Raven, Blackfish and Wolf, which they themselves equated with the Tsimshian clans. However,they had neither a strict rule of descent determining affiliation in the groups, nor of exogamy, two concepts which are indispensable to true matrilinear organization. (1963: 120-121)

The local groups were basically bilateral extended

families, corporate in their holdings of economic resources,

(except in some Coast Salish groups) territories and family

privileges. There was, however, a recognition of rights

accruing unilaterally to individuals, so that a "man and his

wife might assign their first child to the father's so-called

"clan", the next to the mother's if she were of a different

."clan", depending on the names and rights they wanted each

child to share" (1963: 121). . As Sapir noted for the Nootka,

it is possible to inherit privileges in both the male and female lines, preference, where possible, being given to the former. This being the case, it is often hard to see exactly to which sept or clan a person properly belongs. . . In other words, a person steps into certain rights to which he has a claim by descent. . . (but) there are certain natural limitations to the inheritance of all privileges that one has a claim to (1951: 366-367). 48.

And finally, referring to the sub-region generally, he added that "owing to the structure of the village community, . . . there are few persons in a village that one is legally en• titled to marry; but it is important to note that the village community as such need not be exogamous" (1915: 367). The local group, it will be noted, is smaller than the village community, and has its own "chief" and privileges:

The chief of the extended family, like the chief of the northern lineage, was the custodian of all these rights. Membership in the extended family was inherited from one's parents. Although, when the parents stemmed from different local groups, a person was considered more closely allied to his father's side, he retained some rights in the maternal line (Drucker, 1963: 121).

In fact, the extended family, as the term is used above, is synonymous with the local group.

There are only two major ethnographic references for the Bella Coola, one of which is quite limited. Boas noted the bilateral character of the descent system while recog• nizing its unilateral possibilities (1892: 418). He went on to speak of residence, stating that after a wedding "the girl leaves her parents and goes to live with the bridegroom"

(1892: 418).

Mcllwraith described the social organization of the tribe in somewhat greater detail. The most important social unit was the ancestral family, a bilateral grouping defined 49.

as consisting of "those whose ancestral names, embodying definite prerogatives, are embodied in a single origin myth"

(1948: 127). This unit was held together largely by sentiment

based on consanguineal kinship (1948: 138), and a portion of

it was the basic land holding group, although a man was

theoretically "entitled to hunt on any hunting ground of any ancestral family of which he (was3 a member" (1948: 131,132).

These ancestral families were usually localized (1948: 375), and each village normally contained one or two such groupings:

two, or even more ancestral families, inhabiting one village tended to form a closely inter• related stock. In rare instances, two unrelated ancestral families may have become inhabitants of the same village, which would unquestionably lead to inter-marriage (1948: 140).

As for marriage, he noted that

there (was) a strong leaning to local endogamy, although this Cwas3 not compulsory. The restrictions on marriage Grere) of a bilateral nature, and the preferences (were) equally bilateral (1948: 375-376).

The major preferences, of course, were for marriage within the ancestral family, for reasons both of prestige and of land holding. However, since many people married outside their ancestral families, and since these were traced lineally from a pair of ancestors, all persons belonged to more than one ancestral family.

The result was an inevitable merger of several families at 50. each marriage. The economic result was that although "a man

(did) not own the land of his wife's ancestral family, he

(could) use it and his children(had) full rights to it.

Conversely, a wife Csould) gather berries on her husband's land, though it (was) not hers" (1948: 132). I will enlarge upon the ancestral family concept at the end of this chapter.

Finally, Mcllwraith noted that patrilocal residence was normal, but that under certain economic circumstances a man might move to the family home of his wife, the couple residing in the same house as her parents (1948: 399). In effect, this meant a change of local group, even if both families resided in the same village.

The southern Kwakiutl tribes differed from the northern groups in their lack of a unilateral descent pattern. The local group was a bilateral kindred, although there was an emphasis on patrilateral connections (Ford, 1941: 14-15).

Marriage within the immediate family was prohibited

(1941: 36), a factor which would normally make the local group exogamous by kinship. As Ford noted, there was a tendency to marry a person as close to oneself in rank as possible, so that privileges could be retained within the same general status position in the society, rather than diluted through union with those of inferior rank (1941: 36). This 51.

was confirmed by Boas, who noted a pattern of higher-rank

endogamy as a device to consolidate privilege and prestige

(1920: 117; 1921: 782), in combination with village exogamy

as a means of "obtaining new privileges" (1924: 329).

Each local group was normally a single household which

Ford< stated was composed of the leader and his wife and

children, together with

younger brothers of the head of the house and their families, or his sons if they were married. These were the people who had the most right to live in the household. Often, however, there would not be enough younger brothers and married sons to take up all the quarters available, in which case a sister of the household head or, perhaps his wife's sister might be invited to live there, especially if her husband belonged to the same village (1941: 12).

With specific regard to residence, Boas found a

standard patrilocal pattern following marriage (1890: 838;

1897: 364; 1921: 777; 1925: 277), which was later confirmed by

Ford, who stated that "men. . . generally brought their wives

to live in their own tribal district. A young woman went to

live in her husband's tribal district" (1941: 13). As will

become clear later, however, there was probably some variation to this pattern.

For the Nootka, Boas suggested that each tribe, apparently a separate village (1891: 583), held certain 52. territory "in common" (1891: 585). He noted a pattern of patrilocal residence when referring to chiefs, but did not enlarge upon the subject (1891: 595).

Drucker found a variable pattern of local group composition among the Nootka. Although the bilateral kindred normally formed the basis of the local group, there was a series of variations from this. "Residence was normally supposed to be patrilocal. That is, a man was considered to

•belong* to his father's house group, and to live with them.

Actually there was no fixed rule. Chiefs tended to stay most of the time with the group in which they owned property

• . . whether this came from the paternal or maternal line.

But even they moved about, and might spend a fishing season, a year, or even two years, with another group to whom they were related" (Drucker, 1951: 278). All homes and village sites were owned by the chiefs, as were all food resource territories (1951: 248), but these may be considered the vested interests of the corporate local group. Inheritance, as noted earlier, was bilateral, although there was a system of primogeniture for ranking purposes. Thus, the"eldest son of an eldest son was highest of all; his was the "eldest family". His brothers, while sharing in great part in his prestige, were slightly lower than he, his paternal cousins 53. still lower, and so on" (1951: 245). This system is excellently described by Drucker, who stated that

given privilege could be inherited by the eldest son, or shared by several children (all having the right to use it); it could be given to a daughter until her marriage and then bestowed on her brother; it could be given to her son-in-law who might, as the giver specified, have sole right to it or share it with his wife's brother. The nature of the privilege in question had some bearing; a name, song, or dance might feasibly be shared by a number of people, while a seat or a fish-trap was ordinarily held by but one person at a time. Cases are not lacking, however, in which several persons held rights to a fishing place. Ordinarily, a daughter would keep (or a sone-in-law be given), only such rights as were transportable (names, songs, dances, etc), and not such things as seats and fishing rights, unless her husband affiliated with her group. It sometimes happened, however, . . . that a woman might retain even such unportables. If a woman has no brothers her eldest child will inherit his father's rank and rights (if as high or higher than the mother's), and the next will be put in his mother's palace - taking her seat and all her rights. If a man had no children of his own, he might put a brother's or sister's child in his own place. If he had no choice, he would likely take one who otherwise would not have so many rights (1951: 267).

The ideal pattern, then, was patrilocal, although this was subject to certain variations. There was a core group of bilateral kin as the basis for each local group. Beyond this, persons of low status might move around visiting various of their kin for greater or lesser periods, associating themselves

j 54. entirely with the local group in which they happened to reside at the moment, in a form of temporary clientage to related chiefs. Drucker sums their position up by noting his impression of "a continual stream of people, mostly of low rank, pouring in and out of the houses" (1951: 279). Most married adults, however, would be associated with a definite local group, and

that was normally one in which the man had some kin, so that it is possible to speak in a general sense of a patrilocal pattern of residence.

There are a number of references for the Coast Salish

tribes, although there is not usually more than one for any single group. It is to be noted that while these groups were all somewhat similar socially, it has been suggested that they did not all have local group ownership of territories and rights to economic resources.

Boas noted what was apparently a patrilocal pattern for one Vancouver Island group, but stated that it applied parti• cularly to the sons of chiefs (1891: 575). He found the same pattern among the Coast Salish tribes of the lower Fraser

River area (1894: 457). Hill-Tout similarly noted patrilocal residence for the Fraser River tribes (1900a: 505; 1900b: 483;

1903; 408), but qualified this in one instance, stating that after a wedding, a wife "accompanies her husband to his own or 55. his father's house where they now live together. . . Should

the bridegroom have no parents or other near relations, it was not unusual for him in those circumstances to stay and

live with his wife's people" (1904: 319). He found a patri•

local pattern for the same Vancouver Island tribe outlined above by Boas (1907: 311), and went on to note that the "sons and daughters of chiefs intermarried only with those of their own caste" (1907: 308).

Barnett outlined several facets of the Coast Salish

local group, notably its role as the primary institution of social control as "the highest unit of common allegiance"

(1955: 241). They were composed of extended families, normally resident in a single dwelling, and based primarily on consanguineal kinship (1955: 241). Each one usually con• sisted of "a family head, his sons and their children, his unmarried daughters, his brothers and their children, and his unmarried sisters"(1955:242). It usually, but not always, had a patrilocal nucleus (1955: 242) and the leader was normally "in the direct line of inheritance from the tradi• tional founder of the family. . .(^nd) exercised control over the greatest number of its material possessions and its most fundamental rights" (1955: 244). Inheritance was bilateral, although males normally controlled, or were the 56. intermediaries for, heritable rights and possessions (1955:

189, 250-51), except of course, for names. Although inheri• tance normally followed same-sex lines, all but minor articles worn or used by women were held by men (1955: 250).

Marriage rules prohibited unions between first and second cousins bilaterally (1955: 184). Although the wealthier often married outside their villages in an attempt to create new "social and political connections" through family alliances (1955: 182), there was no rule of village exogamy

(1955: 184). The local group, however, was almost always exogamous by kinship.

With regard to residence, Barnett noted that "new family units generally took up residence with the husband's family", but that on occasion, "the new family unit moved into the home of the wife's father or brother" (1955: 242). He lists some of the reasons for matrilocal residence:

This might happen if the boy was wellborn but the youngest of several brothers, and if the bride's father was in a better position to aid the new family unit than the boy's. Or, perhaps more often, the youth was poor but skilled professionally and gave every assurance that he would be an asset to the family of an important man. Not being able to offer an acceptable sum for the aristocratic girl, he was taken into the household as a son-in-law on his professional merits. In effect, he worked most of the marriage gift out, though there was no contract or stipulation on that point. He remained in his father-in-law's household for life unless 57

he found it possible to set up a menage of his own. The residence of such men in the houses of their father-in-law was accepted as natural. For their children, it did not matter, because they were as much the heirs of their mother's father as of their paternal grandfather (1955: 193).

Generally speaking, then, residence was patrilocal, although variations occurred and it was possible for a married man "to shift residence from his father's people to his mother's" (1955: 193) or to that of his wife's parents and their local group associates.

In summary then, the tribes of the southern sub-region of the Northwest coast area appear to have had culturally preferred patterns of patrilocal residence, but with numerous variations depending upon the wealth and prestige of the respective parties to marriage. A common form of preferred marriage was that between two consanguineally related persons.

The patrilocal preference was almost undoubtedly influenced by the fact that property (in a broad sense, including titles and privileges) was very largely controlled by males, despite a general bilateral inheritance pattern. No sanctions against variations from patrilocal residence were evident under normal conditions of residence change. The local group was generally a corporate bilateral grouping of kin. (Figure IV illustrates the "ancestral family" of the Bella Goola as I understand the 58.

II III IV VI

/ --h

/ 4r--/ / / / I ! / / A = • A = • A = O A = O

--/ —/ A = O A = • A = • A = / l—L

/ ; A = O A = O A = O

A EGO

FIGURE IV: THE ANCESTRAL FAMILY, THE KINDRED AND THE LOCAL GROUP. TWO NON-FUNCTIONING KIN GROUPS ARE SHOWN, EGO'S KINDRED, INDICATED IN BLACK, AND THE ANCESTRAL FAMILY DERIVED FROM PAIR V-5, INDICATED BY SOLID LINES. WITH A PATRILATERAL EMPHASIS, THIS WOULD BE EGO'S "OWN" ANCESTRAL FAMILY. HOWEVER, SINCE EGO CAN TRACE DESCENT FROM ALL PAIRS IN GENERATION 5, HE IS A MEMBER OF THE ANCESTRAL FAMILIES OF ALL PERSONS IN GENERATION 2 AS WELL. LOCAL GROUPS ARE SEPARATED BY PARALLEL BARS. 59.

concept, a pattern which appears to have been more or less

standard for the region as a whole.) Twin forces tended

to encourage both village endogamy and exogamy on a class

basis, the former as a means of restricting the distribution

of wealth and the latter as a form of acquiring new wealth.

There is some evidence that the latter was particularly

prevalent among those of high rank, and it would seem likely

that the former was either common or preferred when possible

for those of low or medium rank. Persons of low status, if

the Nootka pattern was general for the region, tended to shift residence periodically, attaching themselves to individuals and local groups of high status in a form of clientage. CHAPTER III

THE SUB-ARCTIC

The huge Sub-arctic region of Canada embraces the entire northern continental forest area plus the Yukon

Territory and northern British Columbia. Although it was the home of a great many tribes of varying size and social configuration, only seven of these groups are discussed in this paper. Three of these are in what can be termed the

Western sub-region and the remaining four in the east. Each of these areas will be treated separately.

The Eastern Sub-arctic

I have selected four tribes of the Algonkian linguistic family for study, the Ojibwa, Cree, Algonkin and Montagnais-

Naskapi.*" These are not intended to be representative of all

Algonkians, or necessarily of the Eastern Sub-arctic sub• sistence area. Rather they were selected only for the large amount of ethnographic material readily available on their social organization.

The four tribes ranged from Saskatchewan in the west, across a substantial portion of Manitoba, throughout all but the southern sections of Ontario and Quebec, and into Labrador in the east. In addition, the Ojibwa occupied parts of the northern United States. Kroeber placed their total population 61. in Canada at the time of early European contact at 59,800

(including the Ottawa) in an area of 1,167,000 square miles

(1963: 140), with an average density of .051 persons per square mile.

All of these groups had certain basic social and cultural similarities, notably their division into small hunting bands with only amorphous political organizations.

As Henness noted when speaking of the Ojibwa, these bands, while "politically independent of one another (were) closely connected by intermarriage" (1960:277). This appears to be a case where one can speak of fields of marriage, as has been 2 noted in several recent studies, in which each band has ties of marriage with several neighbouring bands in a cluster, but no such ties with bands beyond a certain geographical distance.

In addition to the band divisions, each band was divided into hunting territories possessed by families. All four tribes had patrilateral descent systems, but formal descent groups were absent in the area.

The region in which these tribes lived abounded with game and fish, ranging from several species of moose and caribou through beaver and muskrat to rabbits and squirrels, and including both fresh and salt water fish and a variety of species of food birds. The principal foodstuffs were game, 62. both large and small, depending upon seasonal variations in the animal population, and fish, but birds and some collected vegetable products supplemented these. Nonetheless, Driver has pointed out that "while at its best the diet was adequate, famines were an annual threat to existence and a great deal of anxiety over food characterized this area" (1961: 26).

This region was early drawn into the European fur trade, and will be treated during a period of relative iso• lation combined with economic dependence upon trapping for trade. Although it is impossible to date this time specifi• cally, the description may be considered roughly valid for the period from about 1880 to 1920, except where otherwise noted for specific ethnographic statements.

The Ojibwa Indians ranged through Manitoba and Ontario in Canada and the northern part of the United States;their population in Canada is given by Kroeber as 30,000 (1963: 141) at the time of early European contact. Jenness (1960: 277) pointed out that "the majority of the bands were small, numbering probably not more than 300 or 400 individuals."

They were listed by Driver and Massey as patrilocal in residence pattern (1957: 402), and by Murdock (for the Emo

Ojibwa) as bilocal with a neolocal alternate pattern

(1960: 237). 63.

Two nineteenth century references, both by fur traders,

provide the first indication of a specific pattern of Ojibwa residence. Cameron, writing in 1804 of the Nipigon Ojibwa, noted that after marriage, "the husband must then go and live with his father-in-law for a year at least, and give him all he hunts during that time" (1960: 251). His colleague,

Peter Grant, writing at about the same time, confirmed this and added "the marriage is&>far consummated without further ceremony, but, to make it binding, it is necessary that (the groom) should live at least one winter with his father- in-law, during which the old man claims an undisputed right to all the produce of his hunt; but so soon as the young couple have a child of their own, they are released from any further dependence on the old people and are at liberty to go and live where they please" (1960: 320).

Skinner, speaking some years later of a more westerly band, however, does not mention any form of bride service, and states only that "the young woman was simply sent to live in her new husband's lodge at once" (1911: 151). In the same period Speck implied a purely patrilocal pattern for the

Timagami Ojibwa (1915: 23). It is conceivable that these two works, reports of contemporary practices mainly, had simply documented possible historical changes.

The 1930's saw two further references which did not agree with these. Jenness stated that among the Parry Island

Ojibwa of the late nineteenth century, "the accepted suitor 64. went to live with his bride's parents, for whom he hunted

and fished until the time when he could take his wife home

. . . The young couple often returned to the bridegroom's people after the birth of their first child, or at the end of about a year; but many men remained with their wives' bands permanently" (1935: 99). Landes, on the other hand, writing of the Emo Ojibwa of the early twentieth century, stated that "genealogies show no stress on matrilocal, patri•

local, or independent residence . . . Often there was an initial temporary matrilocal residence, followed by temporary patrilocal residence, followed by independent residence, possibly neighbouring either the paternal or maternal home"

(1937: 76). However, she also noted that "individual pre•

ferences or exigencies are the deciding factor" (1937: 76).

Rogers stated that "ideally, residence is virolocal"

Csio) for the contemporary Round Lake Ojibwa (1962:B55), but

that most men now reside "naolocally". This term is only

loosely defined, however, and it is difficult to discern the actual pattern. Dunning noted that "the institution of temporary uxorilocal residence for the first year or two is normal" for the Pekangekum Ojibwa (1959b:62), but that

"residence after marriage is a matter of choice" (1959b: 62).

Most men lived patrilocally, however. He estimated the number 65. of couples living matrilocally (permanently) in the 1910 to

1915 period to have been over twenty-five percent of the total (1959b: 63).

The woodland Cree Indians ranged from Saskatchewan to

Quebec, north of the Ojibwa but inhabiting the same type of country. They totalled some 17,000 at the time of early

European contact (Kroeber, 1963: 141), and were divided into bands and local hunting groups. Driver and Massey considered them patrilocal in residence pattern (1957: 402).

McDonnell, writing about 1797 of the Manitoba Cree, noted a pattern of bride service followed by independence, with a man "at liberty to choose a home for himself" following the birth of a child (1960: 278). This pattern was confirmed by Sir John Franklin two decades later, who described it thus ly:

When a hunter marries his first wife he usually takes up his abode in the tent of his father-in- law, and of course hunts for the family; but when he becomes a father, the families are at liberty to separate, or remain together, as their inclinations prompt them (1910: 66).

Skinner, in 1911, did not himself outline any specific pattern, but simply quoted Alexander Mackenzie as a reference for matri• local residence among the Ontario-Quebec Cree (1911: 57).

Nonetheless, he described the composition of the hunting group "Usually a man hunted with his sons and their families, 66.

but when these grew large, they separated" (1911: 57). This

would seem to preclude permanent matrilocal residence on any

major scale.

Speck found equal numbers of men residing with their

father-in-law's group and with their father's group following marriage among the Mistassini Cree of Quebec (1923: 462).

He pointed out, however, that "hunting companionship usually

includes fathers and sons, although the regular form of this

natural social grouping is generally broken up by local

social circumstances" (1923: 462). Davidson found the same

ideal pattern among the Tete de Boule Cree, but again there were exceptions to this, so that many men resided with the

families of their wives (1928b: 34). Honigmann suggested a

former pattern of "patrilocal residence, normally- (jEollowing) matrilocal bride service that lasted one or two years"

(1953: 811), but, speaking of the Attawapiskat Cree of the

late 1940's, he noted the absence of any definite pattern

(1953: 813-814). He later enlarged upon the earlier pattern

by stating that "a customary period of bride service followed

both a primary and a secondary union. It lasted one or two years, after which a young husband struck out for himself or

(ideally?) returned to his father's band. The next season, however, the couple might go to live with the wife's people 67 or even move away to live alone" (1956: 62). He went on to qualify this further by noting that "the pattern of patri• local residence previously asserted on the basis of limited testimony (Honigmann, 1953: 811) may not be true, Bilocal or neolocal residence appear to have been frequent following the initial period of bride service" (1956: 62, footnote).

The Algonkin Indians inhabited the woodlands of southern Ontario and O^iebec, adjoining the Ojibwa and the agricultural tribes of the lowland. They numbered about 7300

(with the Ottawa) in aboriginal times (Kroeber, 1963: 141).

Driver and Massey termed them patrilocal in residence pattern (1957: 402).

Speck found an ideal pattern of patrilocal residence among the Timiskaming Algonkin, but suggested that this was easily modified in times of economic stress, when it was not unusual for a man "to visit his wife's people and hunt, perhaps for several winters, on his father-in-law's grounds"

(1915: 6). He found the same pattern among the Dumoine River and Kipawa Algonkin as well(1915: 9). That is, the basic hunting group consisted of "individuals related by descent and blood together with other women married to the men of the family" (1915: 3).

The Montagnais-Naskapi are socially similar tribes and are here considered as a single unit. They ranged across 68.

Quebec into the Ungava district, and to the Labrador coast, numbering, according to Kroeber, about 5500 (with the Tete de

Boule Cree) at the time of early European contact (1963: 141).

Driver and Massey listed them as patrilocal (1957: 402), while

Murdock termed the Naskapi alone patrilocal with a bilocal alternate pattern (1960: 232).

Although Strong stated that " a constant shifting of families from group to group within the band, or from band to band, makes any generalization impossible" (1929: 286), he did note that Naskapi bands were "composed largely of kinsfolk most often a man's sons with their families" (1929: 286), thus indicating a predominantly patrilocal pattern.

Speck does not state any specific pattern for the

Lake St. John Montagnais, but he does note both matrilocal and patrilocal residence in the same tribe (1927: 392). Speck and Eisely also found both matrilocal and patrilocal residence among several bands of Montagnais-Naskapi in Labrador

(1942: 232).

Burgessei, found matrilocal residence as a form of bride service among the Lake St. John Montagnais (1944: 3), but added the words of an informant to the effect that "there is no fixed rule which demands that sons join the family of their father-in-law and that those who did so, acted from choice and because a young man had no hunting grounds of his 69.

own" (1944: 4). Later, however, he noted an apparent emphasis

on matrilocal residence (1945: 10).

Lips found a generally patrilocal pattern, at least

partly due to inheritance (1947: 421), while Garigue stated

that the traditional pattern was "patrilocal with certain matrilocal variations" (1957: 115).

Leacock noted an increase in "patrilocality, not only

as the ideal, but increasingly as the real pattern" (1955: 34)

representing a shift over a long term from "informal unstable units" to "formal stable" local groups (1955: 34).

Several peripheral points must now be noted. Firstly,

throughout this region there was a recognition of the indi• vidual family ownership of the hunting territory. This is

testified to, with certain reservations, by a number of writers, most of whom have pointed out that this ownership was conditional on the continued use of the land (Burgesse,

1945: 9; Davidson, 1928b: 32; Garigue, 1957: 122; Leacock,

1954: 6; Speck, 1915: 4; 1923: 459;). Secondly, it is quite clear, I think, that there was an ideal pattern of male inheritance of the hunting territory from the male owner and head of the family group, that this was basically patrilateral, but that on occasion a man might inherit the land of his wife's father (Burgesse, 1945: 10; Davidson, 1928b: 32;

Dunning, 1959b: 106; Honigmann, 1953: 812; Landes, 1937: 89; 70.

Speck, 1915: 5, 10, 12; 1917: 85; 1918: 146; 1923: 462;

1927: 389). Thirdly, the composition of the local group - the hunting group - was variable, depending largely upon the

resources available in a specific territory, and on the number

of sons and daughters in specific families (Davidson, 1928b:

35; Garigue, 1957: 117; Landes, 1937: 3; Leacock, 1954: 25;

Skinner, 1911: 57; Speck 1917: 88; 1923: 462; 1927: 392;

Speck and Eisely, 1939: 270). As Dunning has pointed out, however, it was basically a bilateral group (1959b: 58-61),

and seems generally to have been a patrilateral extended

family, or a male sibling unit, and their families, or a group of patrikin and their families, a fact which automatically suggests a preponderance of patrilocal residence (Davidson,

1928b: 34; Dunning, 1959b: 61; Burgesse, 1944: 4; Landes,

1937: 90; Speck, 1915: 3,4,6,9; 1923: 462). Fourthly, the size of the hunting territory itself varied with the abundance of game and the size of the local group which had to be supported (Hallowell, 1949: 43). And, finally, there is some evidence to suggest that in early times there was an over• abundance of land, and that new land could be utilized by expanding groups.^

Now, a few facts may be stated about marriage in general.

Firstly, the local group, while never exogamous as such, was often exogamous by kinship - that is, the relationships 71. within the group were such as to prohibit marriage on the basis of kinship alone. This was probably the case more often than not, since local groups often consisted mainly of male sibling groups and their families, as noted above. Where a group consisted of cross-sexed siblings and their families, however, intra-group marriage would have been possible under certain circumstances, notably in the case of cross-cousins

(Dunning, 1959b: 118; Garigue, 1957: 115; Jenness, 1935: 98;

Landes, 1937: 5; Speck, 1927: 398; Strong, 1929: 277).

Secondly, the bands (distinct from the local groups) were almost never endogamous, but neither were they often exogamous.

Rather, they fitted the pattern of what Lowie termed agamous- as much marriage within the band as without, in the absence of formal rules of band exogamy or endogamy (Lowie, 1948: 9;

Honigmann, 1953: 814; Speck, 1915: 8; 1927: 396; Strong,

1929: 278). Thirdly, in many groups there was an exogamous

totemic system, but each local group with a married couple contained representatives of at least two totems, by defini•

tion. Descent was normally patrilateral (Garigue, 1957: 115;

Jenness, 1935: 7; Skinner, 1911: 56; Speck, 1915: 7).

Now it is possible to discuss the data presented above concerning residence. Firstly, there very definitely appears

to have been a period of bride service following marriage, so that each man resided for a year or two, or until the birth 72.

of a child, in the local group of his father-in-law. This

was a universal pattern.

Secondly, the emphasis on patrilateral inheritance

of hunting territories indicates, I think, a preference for

residence by males with their f atheis and their local groups

where possible. It will be noted that this preference is

substantiated in a majority of cases by actual patrilocal

residence and by a preponderance of local groups which, while

basically bilateral in nature, have a definite patrilateral

emphasis. It is equally clear, however, that matrilocal

residence also occurred, and that the incidence of such a

pattern was substantial, particularly for the Montagnais-

Naskapi bands. Hallowell has suggested that the size of

Algonkian hunting territories is a function of ecological

adjustment (1949: 43). I suggest that the composition of the

local hunting group is also a function of ecological adjustment

to the extent that a given number of non-hunting people

required a given number of hunters of certain skill and age

etcetera, and vice-versa. Speck provides some of the back•

ground to this when he cites the reasons given by his

Mistassini Cree informants for residence by a male with his

father-in-law:

First, the necessity of rendering aid to him, should he have too few sons or none at all through these fatalities, second, that of 73.

avoiding overcrowding in a family of too many sons, should they be living and working the paternal holding" (1923: 464).

Davidson noted the same basic reasons for the Tete de Boule

Cree, namely, the prevention of either underpopulation or overpopulation, in terms both of numbers and of capabilities, of a specific local group (1928b: 35). The incidence of matrilocality varies considerably, however. Dunning suggests* over twenty-five percent for the Pekangekum Ojibwa between

1910 and 1915 (1959b: 63). Davidson noted an incidence of eleven percent for the Tete de Boule Cree (1928b: 36), and

Speck suggested from fifty to fifty-four percent matrilocal residence for the Mistassini Cree (1923: 462). The variation in rates of matrilocality is quite logical, however, given the generally small size of the local hunting bands, the variation in permissible size of the hunting bands, and the variation in game abundance and size of territories. It can reason• ably be assumed that the smaller the groups, the greater the variation which might result!.,, although this is not the only factor involved.

Finally, there is the question of status. Dunning noted that matrilocal residence imposes considerable social strain on a man (1959b: 133). He lives with his wife's parents "on sufferance and only by helping his father-in-law does he justify his position within the nucleur family of 74. affines" (1959b: 133). With the birth of a child, however, the couple have reached the point where they can successfully

(socially) leave the camp of the wife's kin for that of the man's, or else strike out on their own. As Dunning pointed put, "the only way for a man's status to change and advance is for him to develop a family of his own, and within the context of fraternal unity"? (1959b: 133). The period of matrilocal residence can also impose some economic strain, due to the lack of familiarity with the hunting/trapping terri• tory and the subservient economic position of the younger man in the camp of his wife's family. It can reasonably be assumed

I think, that this emphasis on status is not just a recent development, and was a factor of some importance prior to trapline registration.

I suggest, then, that there was a very definite cultural preference for patrilocal residence in the Eastern sub-arctic, but that such a pattern was not always possible due to the ecological limitations on local group size and composition in the region.

The Western Sub-arctic

I have selected three of the many tribes normally included in lists of the Western Sub-arctic sub-region, and 75. those entirely on the basis of the amount of information available on them. All are Athapaskan speaking, the Tahltan of the west, adjoining the Tlingit, the Kaska of the southern part of Yukon Territory, and the Slave of the Northwest

Territories. It is not easy to arrive at population figures or densities for these tribes due to the lack of agreement until very recent years on the exact divisions within and among the tribes of the region. Jenness suggests a maximum figure of about 4900 (1960: 376, 392, 399), based partly on

Mooney's figures. Kroeber suggested a total of 4250 for the three tribes (including the Taku-tine), in an area of some

135,000 square miles (1963: 141). This would give an average density of .03 persons per square mile.

The region is one of plateaus, intersected by mountain ranges and river valleys. It is rough country, largely primary growth vegetation even today, and abundant in wildlife.

Teit described the economy of the Tahltan in a passage which is probably accurate for the other two tribes as well. He stated that they were:

Pre-eminently hunters and trappers, and gain by far the greatest part of their livelihood by the snaring and trapping of fur bearing animals, and the hunting of large game, principally caribou and moose, which are plentiful in many parts of the plateaus and higher valleys. Bears of all kinds, big-horn sheep, mountain-goats, wolves, lynxes, foxes, wolverenes, martens, etc., are also abundant. Small game, chiefly Arctic hares and ptarmigans, are quite plentiful. Blue and ruffed 76.

grouse, ground-squirrels, and marmots are also plentiful in some localities (1906: 343).

In addition, fish were available to all three tribes, as were roots and berries, although neither of these types of pro• ducts formed the major part of the diet of any of the tribes.

Despite this apparent abundance of food supplies, economic stress and famine occurred periodically, and cannibalism, conceived of as a necessity for the survival of some, was not unknown.

The social organization of these tribes was variable for different band and local group sizes. Usually several families hunted together, but for many specific occasions, such as a fishing party, or times of economic stress, local hunting groups banded together. Each of these groups was variable in size and composition throughout the year, however,and there was a considerable amount of temporary intra-band movement in the region.

These tribes will be described for the period from about 1850 to 1920.

The Tahltan Indians ranged the northwestern corner of

British Columbia, adjoining and in contact with the Tlingit across the Coast mountains, from whom they are believed to have borrowed many facets of their social organization. They shared their eastern border with the Kaska. No reliable estimate of their population is available.' The Kaska 77. occupied the southeastern corner of Yukon Territory and the northeastern corner of British Columbia. They were divided into five sub-tribes, and numbered a maximum of 500 in the pre-European period, according to Honigmann (1954: 24).

Driver and Massey listed both tribes as matrilocal (1957: 402), and Murdock termed the Kaska matrilocal (1960: 247).

At the time of early European contact, the Tahltan were divided into six matrilateral clans grouped into two exogamous phratries. Each of the clans regarded itself as a distinct local, as well as social, group and "claimed the ownership of definite districts, (although) in practice they pooled their hunting territories" (Jenness, 1960: 373). The tribe as a whole maintained one permanent village near its principal fishing grounds, but resided there for only a part of the year. Most of the time they were scattered, presumably divided into six more or less localized bands.

I can find only three useful ethnographic references which mention any pattern of residence for the Tahltan.

Emmons (1911: 98) noted that "in the case of young couples, the husband more frequently comes to live with the wife's parents, thus adding materially to their support". He suggested that in pre-European times a newly-married couple might reside with the family of the groom for a short time and then join the bride's family (1911: 99). MacNeish mentioned a matrilocal pattern and noted a period of bride service, with a young man "helping his intended father-in-law 78. in hunting until the old people were satisfied with him as a son-in-law" (1957: 50).

Probably the most reliable reference on this group consists of the field notes of James Teit, referring to both the Kaska and the Tahltan Indians. He noted a period of bride service (1956: 143), and suggested that" it was the rule for a man to continue living among his people although he might have a separate house of his own. Thus most frequently wives lived among their husband's people. However visits more or less lengthy were occasionally made to the wife's people"

(1956: 145). He summed up the matter by implying that the specific place of residence of an individual was of relatively little importance in the society, but noted that "marriages between members of different clans and bands were encouraged because of the advantages of thus securing the right of hunting in different grounds" (1956: 146). The Kaska subr-tribes were divided into small hunting bands, termed the microcosmic band by Honigmann. Although these often consisted of a single matrilocal extended family, "sometimes a band formed around the nucleus of two brothers and their wives. . . Bands tended to be fluid among most Arctic-drainage Athapaskan groups and few pressures could be relied upon to maintain any particular type of organization" (1954: 75). It can reasonably be assumed, however, that these bands were normally exogamous by kinship. 79.

Honigmann suggested that in the pre-European period,

"matrilocal residence followed marriage" among the Kaska.

"After two or three children had been born a husband usually asked his father-in-law for permission to camp by himself"

(1954: 132). He noted that the pattern was highly flexible, though:

Daughters on occasion also left their parents' band after marriage to take up residence with their husbands' kin. . . A man lacking skill in hunting might join with his wife's family, expecting the latter's brother or father to furnish a dependable source of meat. Should his unproductive presence come to be resented, friction might send him to rejoin his own parental group. If his skill or 'luck' improved he would be able to fulfill the ideal expec• tations of a son-in-law or might decide to go off with his wife to shift for himself (1954: 75).

It is worth noting, however, that Honigmann also sug• gested patterns of patrilateral succession to hunting band leadership, under which a man was succeeded by "a son or brother whose duties included caring for the retired leader"

(1954: 87), and of patrilateral inheritance. In addition, he stated that "marriages did not secure new beaver grounds for the husband, who had the choice of returning to his parents' creek to hunt or of taking the animals on his father-in-law's territory" (1954: 88), although he discounted this "because it would result in bilocal residence" (1954: 88).

For the Kaska of the twentieth century Allard failed to note any specific pattern other than the residence of men 80. with their kin, either affinal or consanguineal (1929: 25).

Honigmann, writing in a period when trapline inheritance was apparently coming into vogue, found an ideal pattern of matrilocal residence with considerable variation based on such

factors as personal circumstances and the availability of

traplines (1949: 125, 193). The same author considered the

possible effects of matrilocal residence in a subsistence area of this type:

There is no evidence that matrilocal residence poses any difficult problems of social adjustment for the man. He assumes the authority in his own family, lives in his own dwelling and, except for offering meat and other assistance to his father-in-law, pursues his own economic interests (1949: 125).

The Slave Indians inhabited the region immediately

east of the Kaska, in the Northwest Territories, ranging as far

to the northeast as Great Slave Lake. Like the Tahltan and

the Kaska, they were divided into a number of "independent,

semi-leaderless bands named generally from the territories over which they hunted" (Jenness, 1960: 391), each of which was highly fluid in composition. Kroeber gave the Slave a population in excess of 2400 at the time of early European contact (1963: 141). Although this might seem rather high for

the area, it would give the Slave approximately the same population density as the Kaska, between 1 and 2 persons per

square mile. Driver and Massey termed the Slave matrilocal

in residence pattern (1957: 402). 81.

I can find only three major ethnographic references

to Slave residence and local groups, two of which have been contradicted by a more recent and thorough study. Mason, writing early in the twentieth century, suggested that among

the Indians of the Great Slave Lake area, a married man,

"generally Clived) with his wife's people unless his parents particularly (needed) his help" (1946: 32). He noted too

that the local group often consisted of several families, comprising a band, but that individual families often travelled and camped alone; (1946: 33). Each band was said to have had a specific territory for hunting (1946: 34). Honigmann stated that following a period of bride service and a marriage, a

"couple set up their own home as neighbours to the girl's family. . . Patrilocal residence appears to have been quite rare" (1946: 85). Helm contradicted both of these reports by noting that "the Slavey usage is for the new couple to reside with the bride's; parents"for a couple of years"or"until they have a baby". After that, the couple may remain near the bride's family, or strike out alone or, more commonly, return to live with the husband's family's band" (1961: 67). She also suggested that intra-band marriage occurred not infre• quently (1961: 67-68). In discussing the contradiction concerning residence patterns, Helm points out that Honigmann's 82. statement is based on material from informants concerning the kinship structure "as it was reported to have functioned" 4 in the past (Honigmann, 1946: 69). Finally, she goes on to note that "eventual permanent patrilocal residence among the

Slave is common but not inevitable" (1961: 67, footnote).

Several factors must now be noted for the tribes of this area. Firstly, there does not appear to have been any specific ownership of territory by individuals, although the variable-composition, bilateral hunting bands may have enjoyed certain territorial rights. The concept of corporate local groups does not apply to this area, however. There was no recognized system of hunting territory inheritance. There appears, however, to have been sufficient land for most cir• cumstances, and an individual was probably able to change his residence (i.e. move to another hunting group) - at least temporarily - without any actual ecological necessity for doing so. Now, on the basis of the foregoing data, with some consideration for the geographic and cultural region as a whole, I suggest that the residence pattern may be described as follows.

Firstly, as was the case among the Eastern Sub-arctic tribes, there was almost certainly a universal period of bride service in the local group of a man's wife's father. Following 83.

this, secondly, residence was normally with the local group

- the hunting band - of the male's father, although the Kaska may have been exceptional. The fluidity of the local groups, however, suggests immediately that ecological considerations

played an important part in deciding a place of residence.

Honigmann, for example, noted that the three hundred to five hundred Kaska were divided into five sub-tribes which were

further divided into local hunting groups. To suggest that

bands of from, say, fifteen to twenty-five people, would be able to maintain a specific pattern of residence with anything more than slight regularity would seem to ignore the ecological

conditions of the region. In addition, several authors have noted that residence often followed an alternating pattern and depended, at least in part, upon purely personal factors.

Finally, then, it can be stated that the Western

Sub-arctic patterns of residence and local group composition were similar to those of the Eastern Sub-arctic in their

flexibility based on ecological and personal factors. There was not, however, an expressed cultural preference for a single unilocal pattern of residence as there was among the Eastern

Sub-arctic,tribes. The only evidence which might indicate

such a preference - the patrilateral inheritance of property and the patrilateral succession pattern for local group

leadership noted for the Kaska - could well be exceptional, 84.

of course. Nonetheless, in the absence of any further

evidence, it would be best to speak of the residence pattern

of the Western Sub-arctic tribes as one of statistical regu•

larity in favour of patrilocal residence.

The positions of both the Eastern and Western Sub•

arctic sub-regions can now be placed in a more complete

perspective. I have suggested that there was a preferred

rule of patrilocal residence in all societies of the Eastern

sub-region considered in this chapter, and that there was

only a statistical regularity of patrilocal residence in the

Western sub-region. Some discussion of the differences

between these two areas is certainly in order. Firstly, the

distinction is not based on the differences between the

descent systems found in the two areas. The Tahltan and the

Kaska, at least in the western part of the Kaska region, were matrilateral in descent.^ The Slave were bilateral in descent

(Driver and Massey are the principal secondary source for all

three tribes; 1957: 419). The four tribes of the Eastern

sub-region are all patrilateral in descent. At first glance

it would seem inconceivable that a society with the local

group organization of the Tahltan and the Kaska could have a

pattern of patrilocal residence in conjunction with matrix-

lateral descent. However, if we examine the patrilateral

descent system of the Eastern Sub-arctic tribes, we see that 85.

it was of significance only for the regulation of marriage.

If operated, as it were, only to distinguish between kin and

non-kin, and had virtually no effect on the composition of

the local group other than through its effect on choice of marriage partners. The matrilateral and bilateral systems of

the Western Sub-arctic were of a similar nature: in this

ecological area they could not function in any other way.

In none of the tribes under study were there any descent

groups, for example. The ecology of the area simply did not

permit their development as functioning groups in the

societies.

Secondly, the presence of corporate local groups in

the Eastern sub-region indicates a cultural preference for a

specific pattern or complex of patterns of residence. Such

groups are lacking in the Western sub-region. This point

will be enlarged upon in the concluding chapter.

Thirdly, the distinction between the cultural pre•

ference of patrilocality in the east and the statistical

regularity of patrilocality in the west is based upon the

presence of patrilateral inheritance patterns in the Eastern

sub-region and their absence in the Western sub-region

(although the Kaska may be an exception), and on the presence

of status considerations in the east and the lack of such 86.

considerations in the west. The two factors may be considered manifestations of the existence of a cultural preference for

a single unilocal pattern of residence in the Eastern

Sub-arctic.

I do not mean to suggest any evolutionary sequence or

cause and effect relationship for the dual presence of the

patrilateral inheritance of hunting territories and the

cultural preference for patrilocality in the east.7 In fact,

the patrilateral inheritance of hunting territories was purely

an ideal, and actual inheritance was governed almost wholly

by economic and ecological circumstances. Basically, one

inherited rights of use over the territory exploited by the members of the local group of which one was a member. More

often that not, that was the local group of one's father or

brother. But it was not necessarily so, for it could well have been the local group of one's wife's father or brother, or,

in some cases, the local group of people to whom one was not

related. The point is that a preference for a specific

pattern of inheritance was economically unnecessary - one would inherit land use rights in a local group anyway. But

the existence of such a pattern functioned to maintain a

cultural preference for patrilocal residence, the ideal

pattern which was followed whenever possible. While it is 87.

doubtless true that the inheritance pattern influenced

choice where choice was possible, it was not crucial to an

individual's economic existence or well-being. An inheri•

tance pattern of some sort followed naturally from the

corporate local group principle; the specific pattern in

this case amounted only to an institutionalized expression of a cultural preference for a single unilocal pattern of residence.

Similarly, the presence of status considerations favouring patrilocal residence whenever possible, and negating any advantages in matrilocal residence when a choice was possible, served to emphasize a single unilocal pattern, patrilocality, as an ideal in the Eastern Sub-arctic.

The absence of these two factors in the Western Sub• arctic,, or of any other comparable emphases on a specific unilocal pattern, indicates the lack of any cultural preference

for a single unilocal pattern of residence.

In summary, I suggest that the composition of the

local group in each area of the Sub-arctic was bilateral, and

that this was based in part upon economic and ecological

factors. There was in both areas of the region a pattern of patrilocal residence, in the west as a statistical regularity probably based in part upon the position of the male as the hunter and the principal food supplier of the domestic group, 88. and in the east as a cultural preference based on corporate local groups and expressed in inheritance patterns and social status. In neither sub-region was the pattern abso• lutely possible for ecological reasons, and there was in each case considerable variation in pattern, notably through matrilocal residence. 89.

FOOTNOTES: CHAPTER III

Although actually two separate tribes, the cultural and social similarities of the Montagnais and Naskapi justify their consideration as one.

2 Dunning, 1959b: 167-188; Rogers, 1962: B51. This is a common pattern in several areas of the world, as noted by

Douglas for the Lele of Kasai (1960: 243-244) and Faronfor the

Mapuche of Chile (1961: 198).

3

This is indicated by a comparison of the population densities of specific hunting territories, which ranged, as an example, from 6.2 to 55.6 persons per square mile for the

Berens River and Grand Lake Victoria Ojibwa (Hallowell, 1949:

40), and the over-all tribal densities, which ranged from .01 to .24 persons per square mile (Kroeber, 1963: 141). 4

It is noteworthy that Honigmann's entire 169 page monograph was based on "about thirty-five or forty informant hours during the seven weeks in residence in Fort Nelson"

(1946: 4), a fact which indicates that considerable caution must be used when referring to his data.

It is fairly well established, incidentally, that this matrilateral pattern of descent diffused from the Northwest 90, coast where it was of considerably greater importance, as is noted by Jenness (1960: 373).

It is interesting to note that at least one tribe which adopted matrilateral descent from the Northwest coast

Indians was able to maintain a functioning phratric organi• zation only two months of the year (Jenness, 1937: 49).

^ Something along this order has been attempted, however. Leacock (1955) felt that the patrilateral inheritance

pattern had been one of the major factors in the growth of patrilocal residence since the inception of the fur trade.

We are here concerned only with the existence of the two

patterns, however. CHAPTER IV

THE PLAINS

There were some thirteen tribes in what Driver and

Massey have termed the buffalo subsistence area, corres•

ponding almost exactly with their "plains culture area".*"

The group included four linguistic families (Driver and Massey

1957: 170). The tribes, from north to south, were the Sarsi,

Blackfoot and Plains Cree of prairie Canada, the Gros Ventre

and Assiniboine of the border area, and the Crow, Teton

Dakota, Cheyenne, Arapahp, Jicarilla, Kiowa, Kiowa-Apache

and Comanche of the United States. Kroeber gave them a

combined population of 61,500 at the time of early European

contact, ranging an area of some 650,000 square miles,

excluding the Jicarilla (1963: 139). This is an average

population density of .1 persons per square mile. I have not

been able to obtain any reliable population estimate for the

Jicarilla.

The Plains tribes were all characterized by a division

into bands, each one a local unit, which came together in a major powwow on one or two occasions each year. Provinse

(1955: 341) noted that at these times "the centralized

authority of the tribes was exceedingly strong and well

defined". During the year, however, it was the bands and the various institutions within them, such as military and 92. dancing societies, and family groups, which were responsible for the conduct of their members. Further, it must be noted that the central authority of the tribe, vested in certain police officials, was valid only for tribal activities, and not for purely band, or even certain inter-band, matters.

Driver and Massey (1957: 178) have outlined the principal food supply of the area: "Buffalo were supplemented by elk, antelope, bear, and occasionally smaller game. The only plant food eaten in any quantity was berries, which was a standard ingredient in the pemmican made by all the tribes.

A few roots were obtained by women with the aid of a pointed digging stick. Tribes adjacent to the agricultural area to the east and to the southwest sometimes obtained a little corn in trade or in predatory raids on their more sedentary neighbours". Plains life, then, was nomadic, and the bands were largely dependent upon roving herds of large game for their subsistence. Nonetheless, the same authors were able to note that as long as buffalo were numerous, "food was seldom a serious problem" (1957: 179).

The three northern tribes, the Sarsi, Blackfoot and

Plains Cree, have all been termed patrilocal by Driver and

Massey (1957: 402). In addition, Murdock has typed the

Blackfoot as patrilocal (1960: 229). Although there were 93. 2 certain fundamental differences among the three tribes, they were socially quite similar, and are grouped together here simply for the sake of convenience.

The Sarsi, the northernmost of the plains tribes, are reported to have numbered about 700 people (Kroeber, 1963: 139), divided into five bands (Jenness, 1935: 23). There were no rules of exogamy or endogamy for the bands, and it can readily be appreciated that the average size of 140 people would pose immediate problems had there been such rules. I can find only three ethnographic notes on the Sarsi dealing with residence (Wilson, 1889; Jenness, 1938; and Honigmann,

1955), and they indicate the probable predominance of a patrilocal pattern. Jenness suggested the possibility of a requirement, in that the bride was expected to "erect her tent beside, her future father-in-law" after marriage (1938:

23). However, since the pattern of residence for conjugal unions within a band of but 140 persons would be relatively unimportant in determining group composition, and since bands of that size would not be likely to have fixed rules of inter- band marriage (i.e. as in a moiety system), the element of compulsion can be discounted.

The Blackfoot are said to have included some 15,000 people (Kroeber, 1963: 139), divided into three sub-tribes and at least thirty-six bands (Wissler, 1911: 21), each 94.

averaging over four hundred persons in size. Wissler,

writing of the nineteenth century pattern, noted that the

bands tended to be exogamous, despite the absence of any

formal rule of exogamy. It was apparently considered "not

good form, but not criminal" to marry within one's own band

(1911: 19). The residence pattern thus becomes far easier

to analyse from the literature.

Patrilocal residence was apparently the norm, in that

a woman, following marriage, went "to her husband and (was)

considered as belonging to his band" (1911: 18). Despite

this, however, "for a man to join the band of his wife at marriage (was) not unusual. The reasons for such changes

(were) usually selfish, in that greater material and social

advantages (were) offered, but we have no suggestion of such

transfers being made with the idea of recruiting a depleted

band" (1911: 19). A further indication of the frequency of matrilocal residence may be the statement of a belief "that

the father-in-law (pf a male) was for a time entitled to part

of the spoils of the chase and war" (1911: 10), a situation

which would not normally obtain with a straight patrilocal

residence pattern and virtual band exogamy, but which would be

unlikely to develop into an institutionalized factor without

a reasonable number of cases of matrilocal residence occurring. 95.

At any rate, it appears from Wissler's material to have been customary for men to remain in their own band after marriage, and for their wives to join them after payment of a bride price, but it was by no means obligatory and personal ad• vantage rather than any ecological necessity governed the choice. In cases of matrilocal residence, payment for the bride probably took the form of service to the father-in-law.

Two earlier writers, Wilson (1888: 192) and MacLean

(1895: 251), noted patrilocal residence, but did not suggest the existence of any hard and fast rule. Two later writers,

Hanks and Richardson (1945: 4), also noted a patrilocal pattern, but again did not state this as a rule.

The Plains Cree, who numbered about 3000 according to

Kroeber (1963: 169), were divided into twelve bands (Jenness,

1960: 317), averaging about two hundred and fifty people each.

There was no formal rule of band exogamy, but most men married girls of other bands (Skinner, 1914b: 517). The literature contains some contradictions which can, I believe, be understood if they are placed in the context of my hypothesis for this region. Skinner (1914a: 72) suggested a matrilocal pattern for the most easterly band of Plains Cree, and noted that two generations of married couples not infrequently lived in the same lodge. Mandelbaum, however, writing some years later of the nineteenth century pattern, 96.

stated that "the newly married couple usually lived near the

husband's parents although this was not an inflexible rule"

(1940: 245). He noted further, though, that matrilocal

residence occurred under certain circumstances, such as the death

of the bridegroom's parents, and quoted an informant as saying

that a young couple "went back and forth between his and her

parents" (1940: 246). The point I wish to make is that there

is absolutely no suggestion of any "rule" of residence, and,

although patrilocal residence was apparently customary,

matrilocal residence was not unusual and was not apparently

considered a second-class form.

The Assiniboine and Gros Ventre Indians both inhabited

the border regions of the Plains region, the Gros Ventre in

the west and the Assiniboine in the east. Again, these

tribes are considered jointly only for the sake of convenience.

Both are listed as patrilocal by Driver and Massey (1957: 402).

Kroeber gives the Assiniboine a population of 10,000 and the

Gros Ventre some 3000 at the time of early European contact

(1963: 139).

The Assiniboine are said by Lowie to have had seventeen

bands (1909: 33-34), but it is not known if they were either

exogamous or endogamous. The latter was certainly possible with an average band size of nearly six hundred, but it is unlikely in the context of Plains Indian culture which 97. generally lent support to inter-band marriage alliances.

They are stated to have had two patterns of residence, patrilocal following the payment of a bride price (Lowie,

1909: 41) and matrilocal which was associated with bride

service, under which a man joined "his wife's relatives"

(Lowie, 1909: 40). It is to be noted that this appears to have been the case among the Blackfoot as well, although

there does not seem to have been any definite rule in either society.

The number, and consequently size, of Gros Ventre bands is open to question. Kroeber stated that there were at

least eight (1908: 148), which would give an average size of

three hundred and seventy-five, while Flannery suggested that

there were twelve(1953: 30), in which case they would average about two hundred and fifty persons each. Both writers agreed that an individual's father's band was exogamous, and

there are indications that the band of ego's mother was

Included in the exogamous unit (Kroeber, 1908: 147: Flannery,

1953: 30). Kroeber indicated the variety of patterns that obtained in this tribe when he noted that "a young married man often lived with his father-in-law" (1908: 180) but that

"a woman, on her husband's death, went back to live with her parents" (1908: 181). Flannery (1953: 176) threw further 98. light on the matter: "There was no hard and fast rule in re• gard to residence. . .The tendency seems to have been towards patrilocal residence in general, although it was not unusual for a couple to move back and forth". Two further ethnographic statements provide an indication of the reasons for a specific choice. Flannery (1953: 176) noted that

"matrilocal residence in the same lodge was incompatible with the observance of the mother-in-law taboo and it was said that if a man wished his son-in-law to remain with him, he provided the separate lodge and all the equipment".

Kroeber stated (1908: 180) that at marriage a man "received a tent with all the proper furniture, usually from his prospective parents-in-law, but sometimes from his own parents if they were the wealthier". Both of these statements seem 3 to support Wissler's contention that the "greater material and social advantages" (1911} were a major factor in choosing a place of residence.

The Teton Dakota, the western division of the Dakota tribe, numbered about 10,000 at the time of early European contact, while the Crow Indians to the west totalled about

4000 (Kroeber, 1963: 139). Both tribes are listed as patrilocal by Driver and Massey (1957: 410). Murdock termed the Teton patrilocal with a bilocal alternate pattern

(1960: 232) and the Crow patrilocal (1960: 247). 99.

The Teton were divided into seven sub-tribes, each of which was further sub-divided into hunting bands. Dorsey

listed over seventy such bands in all (1897: 218-221), an

average size of slightly over one hundred persons, a figure which seems rather low for the region.

Hassrick noted both patrilocal and matrilocal residence

for the Teton Dakota (1944: 339), and suggested a balance in

favour of patrilocal residence in the absence of any formal

rule. Dorsey, writing in the nineteenth century, confirmed

this: "Sometimes the young man goes to live with his-wife's kindred, but in such matters there is no fixed rule" (1897:

222). Hassrick added, however, that "matrilocal residence was a difficult adjustment because of the Giother-in-la^)

avoidance taboo, and consequently it was always of a temporary character" (1944: 339).

It is not possible to consider the Crow Indians in any detail here, because of the paucity of material available, particularly concerning the local group. The only data I can

find which is relevant is Lowie's statement of initial patri•

local residence (1912: 223). The specific context, and the pattern following this, if any, is unknown, but I think it would be reasonably safe to assume in the context of this statement that there was no formal rule of residence. 100.

The Cheyenne and the Arapaho are very often considered jointly in the literature, and will be here also. The two tribes adjoined each other on the middle plains, the Arapaho south of the Cheyenne. Their relationship has been summed up by Eggan, who noted that "the two tribes lived in close alliance although each tribe maintained its individuality"

(1955a: 36). Driver and Massey listed both tribes as matri• local (1957:, 402), as did Murdock (1960: 229). Kroeber gave them,a combined population of 6500 at the time of early

European contact (1963: 139).

Mooney stated that there were ten bands of Cheyenne "

(1907: 403), while Kroeber listed four bands of Arapaho

(1902: 7), giving each of the fourteen groups slightly more than 450 persons. Although Kroeber stated that the Arapaho bands were not exogamous (1902: 8), he did not discuss the basis for this. Eggan, speaking of both tribes, noted that the only exogamy rule was a kinship one, but that it might have covered the band "more or less completely" (1955a: 59).

Although Mooney suggested that Cheyenne bands tended towards endogamy for political reasons (1907: 410), Eggan pointed out that the kinship exogamy rule was probably far more important

(1955a: 84). 101.

Both tribes had a matrilocal extended family organi• zation, with each family grouping operating as an economic unit (Eggan, 1955a; 37). All of the references consulted are quite clear on the pattern of residence: Grinnell

(1962: 144-145) noted that matrilocal residence appeared to be normal, while Eggan (1955a: 61) spoke of it as a "general custom". Two statements suggest that this was not by any means a formal rule, however. Mooney spoke of "the husband usually (but not always) going to live with his wife's people" (1907: 411), while Eggan stated that "a change of band affiliation was possible, and the family might return to the husband's band or join another" (1955a: 37). Two reasons for patrilocal residence may have been a difference in wealth of the respective parents-in-law, since this was a factor to be considered in marriage in a tribe with both bride price and bride service (Eggan, 1955a: 59-60), and, possibly, a desire to maintain a certain number of hunters in a particular band (Eggan, 1955a: 83).

The Arapaho were considered by Eggan to be almost identical to the Cheyenne in residence pattern, with an emphasis on matrilocal residence in the absence of any rule either requiring matrilocal or forbidding patrilocal residence

(1955a: 58-62). Kroeber (1902: 12) noted a general matri• local pattern, and did not mention any rule. 102.

Although there appears to be an emphasis on one specific pattern for these two tribes, there does not appear to have been any system of sanctions against variant forms at all, nor does there appear to have been any reason other than customary behaviour for the one pattern.

The Kiowa and the Kiowa-Apache, although of different language families, were closely associated, the latter being geographically encircled by the Kiowa. Kroeber listed their combined population as 2300 at the time of early European contact (Kroeber, 1963: 139), while Mooney gave the maximum population of the Kiowa-Apache alone as 350 (1898: 353).

Driver and Massey termed the Kiowa bilocal, and the Kiowa-

Apache matrilocal, in residence pattern (1957: 402), while

Murdock listed the Kiowa-Apache as matrilocal with a bilocal alternate pattern (1960: 229).

The Kiowa were divided into six bands (Mooney, 1898:

227), each averaging slightly over three hundred people.

They were not exogamous (Mooney, 1898: 227), Although Mooney stated that "in general the husband goes to live among his wife's people instead of taking her to his own camp"

(1898: 233), he did not suggest that this was expressed in any formal rule. Further data are lacking for the Kiowa, however. 103.

The Kiowa-Apache did not have a band organization exactly similar to that of the other tribes of the Plains region. They were endogamous in practice, despite the occasional case of marriage with the Kiowa or other nearby

tribes, and the network of kin relationships through which all members of the tribe were related to all other members

(McAllister, 1955: 99).

Murdock termed the Kiowa-Apache an endogamous deme society, as noted earlier. If by this he meant that they constituted a single local group, however, he is in error.

McAllister noted for this tribe, for example, that marriage

"may not have necessitated a change from one community to another for either spouse" (1955: 147), and that two spouses

"may not even have known each other previously" (1955: 145), an unlikely, if not impossible occurrence in a single local group with a maximum population of 350 and the social con•

figuration of the Kiowa-Apache. Indeed, McAllister provided concrete evidence of local groups much smaller than the tribe as a whole in his description of the various groupings of the Kiowa-Apache. After one annual tribal event (held with the Kiowa), they "split into various groups which might align themselves differently each year" (1955: 165). I think it is clear that while the tribe did not have the band organization 104.

of other Plains groups, it was divided into a series of

local groups.

The extended domestic family, consisting of "a group

of relatives who lived in separate tipis which were located

close together, but who did not as a rule eat together"

(McAllister, 1955: 165) was the basic social and economic unit of the tribe. Very often, however, several of these

families would group themselves into larger clusters, which

would vary considerably in size on a seasonal basis, and

which might vary in basic composition on an annual basis ,.

(McAllister, 1955: 166). It is the choice of residence be•

tween two or more of these larger clusters that is of interest here. McAllister has noted that "there was no fixed rule

regarding residence" but that "the couple lived with,.; or had

their tipi near, one family and then the other" (1955: 147).

Although matrilocal residence was probably the commonest

pattern, he suggested, I think, correctly, that for a group of

this size organized in this manner "place of residence and change of residence may not be very important" (1955: 165).

For the Kiowa-Apache specifically, there was not always a change of community at marriage, but often only a change of dwelling, a factor which automatically decreases the con• ceptual importance of any one pattern. In addition to this, 105.

Bellah pointed out that "a considerable number of elementary families" were patrilocal, despite the general matrilocal extended family organization (1952: 43).

The Jicarilla Apache of the southwestern plains were really only marginally in the buffalo subsistence area, since

they practiced some agriculture and recognized private ownership of farm land^ (Opler, 1936: 206). Nonetheless,

they were still largely dependent upon game for their food supply, and were still largely nomadic. As noted earlier,

I have been unable to obtain any reliable estimate of their early population. They were divided into two bands, neither of which was exogamous in rule or in practice, but each of which was divided into six local groups. Each was described by Opler as "a cluster of families associated through blood

relationships, marriage, common interests, or strong

friendship, living in the same district or around a specific

landmark under the control of a recognized chief or leader"

(1936: 203). These local groups were probably exogamous.

Driver and Massey classified the tribe as matrilocal in

residence pattern (1957: 402).

Opler noted that the matrilocal extended family was

the primary social and economic unit of the Jicarilla,

"comprising a married couple, their unmarried children, their married daughters and sons-in-law, and the children of 106.

these" (1936: 203). Bellah added that "ordinary hunting and gathering activities need involve no personnel outside

this group" (1952: 15). Although both writers noted the normal pattern of residence as matrilocal, it was Opler who expanded on this, qualifying the residence pattern with

the statement that "it (was) accepted that residence (should) be matrilocal whenever this (yas) possible" (1947: 453).

He also noted the solidly established norm that "the married man GhoulS) provide for the household of his parents-in-law as well as for his own larder" (1947: 453), adding that all cooking for the extended family was done in the lodge of the senior woman of the family.

The i.Comanche were the southernmost of the Plains

Indian tribes of North America, ranging almost to the Gulf of Mexico. They numbered about 7000 at the time of early

European contact (Kroeber, 1963: 139). Driver and Massey

termed them patrilocal (1957: 402), while Murdock considered

them to be neolocal, although with an alternate pattern of patrilocal residence (1960: 229).

The -.Comanche were actually "a congerie of bands held together as a peace group by the bonds of a common

tongue and culture" (Hoebel, 1940: 11). They varied from the 107.

"single family camping alone, through the small camp of related individuals who formed a composite extended family, up to the large band of several hundred persons" (Hoebel,

1940: 11). Thus, if by endogamous deme Murdock means a single local group, then the CComanche did not have endogamous demes.

Both of the major writers on Comanche residence have noted a considerable variety in patterns. Hoebel found that

"JComanche residence tended to be more patrilocal, but equally subject to easy change" (1939: 446). Later, however, he found that "marriage was commonly intra-band when a person was a member of a large group, but there was no fixed rule of residence. Since marriage seems to have been to a great extent within the band, a couple lived customarily in the same group as both their parents" (1940: 12). Gladwin spoke of a patrilocal tendency (1948: 77), but noted that matri• local residence occurred not infrequently. He suggested that it began as bride service, with the husband later persuaded to remain with his wife's parents indefinitely

(1948: 84). The two families often became an economic unit

(1948: 89). This occurred most frequently with marriages arranged by the girl's family, after which "the son-in-law

(moved) in to live with or near them" (1948: 89). This was not always a permanent arrangement, however, since, "after 108. two or three years, the husband often had accumulated enough property, either as gifts from his in-laws or through his own efforts at raiding and such, to move away and set up for himself" (1948: 89-90). Even this might not be the end of the changes, however: he also reported that in this event the young couple might be joined later by the girl's parents

(1948: 90).' Finally, Hoebel summed up the flexibility of the pattern by noting that "there was nothing to prevent change of residence to another band at any time, mere whim being sufficient cause. . . Qtn^D temporary changes were frequent"

(1940: 12).

It is now possible to summarize the data presented for the tribes of the plains region. Of the thirteen tribes one, the Crow, may be excluded from discussion because of the paucity of material available. Of the remaining twelve, there is a reported emphasis on patrilocal residence among four tribes, and on matrilocal residence for a further four. The remaining four tribes had both forms reported as common. The possibility of a pattern alternate to the major one noted is mentioned for four of the eight tribes with only a single general pattern. Of the twenty-nine references utilized, not 109. one suggests the presence of any particular "rule" of residence. I think these statistics demonstrate my point quite clearly: although there were indeed some preferences for certain patterns of residence among the tribes of this region, there were no rules, and there was not apparently any discrimination against those who chose an alternate pattern. Although extended families with matrilocal resi• dence in the younger or middle generation were common among some of the tribes, often as an economic unit, these were apparently established more on the basis of personal choice and compatibility than on any recognized rule of behaviour.

With few exceptions, the general pattern of residence does not appear to have affected significantly the basic composition of the bilateral hunting band. In other words, the behaviour patterns of matrilocal and patrilocal residence were generally of equal import on local group composition, and if this latter is thus of key interest, it is somewhat pedantic to consider the two patterns radically different. The typing of these tribes as either matrilocal or patrilocal in residence pattern is thus of only minor value, if not actually an obscuring fact, to any understanding of the composition and significance of the local group. 110.

It can readily be appreciated that in the absence of unilineal descent there were no descent groups among these tribes. Further, the local group often reached a size of several hundred people, a factor which was easily possible and, for some, highly functional in the context of Plains

Indian subsistence culture. There were no reasons whatever for the development of corporate local groups here; indeed such a development was undesirable in the context of Plains

Indian culture. Ill

FOOTNOTES: CHAPTER IV

The one major exception is the Lipan, who are included in the Plains culture area, but not the buffalo subsistence area (Driver and Massey, 1957; 170, 177).

2

The Plains Cree had moved out from the woods and were in contact with and influenced by woodland Ojibwa

(Jenness, 1960: 317) while the Sarsi were strongly influenced by the Blackfoot (Jenness, 1960: 325). The Plains Cree and

Blackfoot were both Algonkian ; speaking tribes, while the

Sarsi spoke an Athapaskan language (Driver and Massey,

1957: 170). 3

Made for the Blackfoot, but applicable, I suggest for this tribe as well. 4

No specific rule of unilineal inheritance of this land developed, however (Opler, 1936: 206). CHAPTER V

GREAT BASIN

The tribes of the Great Basin area of the western

United States inhabited a semi-arid region of valleys and plateaus. They were primarily collectors and gatherers of seeds, nuts and roots, and, during certain periods of the year, deer and antelope hunters. The tribes of this region were all roughly similar in both society and culture.

Rather than define the boundaries of the area, however, I think it is best simply to speak of the general territory of the Southern Paiute, Northern Paiute, Shoshoni and Ute tribes occupying what Driver and Massey termed the Desert culture area (1957: 170). Driver summed up the economy of the region as follows:

In the Great Basin there was a large area where wild plant foods predominated over animal foods. The pinon, from which pinenuts were obtained, was the most common single species used for food, but seeds of many; other plants were also eaten, and a few roots were obtained with the aid of the digging stick in the northern part of this area. There was no single food which could be regarded as a staple because even the pine nut crop failed frequently. Deer inhabited the mountains of the Great Basin, where they were probably the most common large animal, and mountain sheep were to be found on the higher crags and summits over most of the area. Antelope became more frequent 113.

toward the north and, because they lived in open country in large herds, were usually hunted communally. These hunts were led by medicine men who were supposed to be able to attract antelope by their magical powers. Compared with other areas, however, large game was scarce and probably furnished a smaller part of the year round diet than did rodents, reptiles, and insects. Rabbits in particular were hunted in communal drives similar, except for the absence of the medicine man, to those for antelope. Famines were common in the Great Basin and people were hungry much of the time, according to old Indian informants (1961: 29-30).

The tribes of the Great Basin shared similar character• istics in their social organization. They were bilateral in descent pattern, and divided into family bands and band groups. The former was the local group much of the year.

Steward described it as the "nucleur, biological or bilateral family, consisting of mother, father, and children. Unlike many primitive peoples,the Shoshoneans were not organized in extended family or lineage groups and, although, as we shall see subsequently, the immediate family was frequently enlarged through plural spouses and different families were closely allied by marriage, the functioning unit was the nucleur family, augmented only by a grandparent, aunt, or uncle who otherwise would be homeless" (1955: 102-103). On occasion, two, or possibly more related families might travel together as a unit, but this was not always economically possible 114.

(Steward, 1955: 105). For some events, such as periodic rabbit drives and antelope hunts, a number of family local groups would come together for communal economic activity.

The same was true for occasional fishing parties. Nonetheless the normal pattern was a "fragmentation of Shoshonean society into nucleur family units, which moved about the country seasonally and annually in an unpredictable itin• erary" (Steward, 1955: 107).

These tribes have been quite extensively treated in the literature, generally as bands, properly the term for the largest political grouping normally found in this region, although some bands were noted for the unity of the group during a short period of each year when it would be assembled at one "headquarters" point with permanent houses. These groups are termed villages rather than bands. I propose, however, only to examine several sources on these groups at random, rather than just select a few bands for intensive study. I anticipate similar general patterns of behaviour for the entire region, and I think this form of random sampling should be sufficient to enable general conclusions to be drawn.

Kroeber listed the population of all these tribes at about 15,000 at the time of early European contact (1963: 137).

They were all of the same linguistic family, Uto-Aztecan

(Driver and Massey, 1957: 170). Driver and Massey termed them 115. all bilocal in residence pattern, with four matrilocal exceptions (1957: 402). Murdock termed the Hekandika Sho• shone bilocal and the Surprise Valley Paiute neolocal

(1960: 229).

One writer has already discussed the Great Basin area and provided an answer for the variety of seemingly contra• dictory data concerning residence and marriage patterns. I think it will be of some value to consider several tribal groupings in order, however, demonstrating the type of material available for these groups.

Kelly, speaking of the Surprise Valley Northern

Paiute, pointed out that while the only formal bar to marriage was a blood relationship, there "was no feeling for exogamy" and a man normally "married within his own band or a related

Paviotso band" (1932: 165). She reported "a definite feeling for band membership" (1932: 182), with each band governed by a single leader, a "man of influence" (1932: 182). She went on to note that "the chiefly office was loosely inherited, but if a person were thought not capable, someone from another family was chosen by general consent" (1932: 183). With regard to residence, Kelly found that "initial matrilocal residence was the rule. . . Independent residence seems to have been established after the birth of one or two children. Patrilocal residence occasionally obtained if a youth were an only child"

(1932: 183). 116.

For the Owens Valley Northern Paiute, Stewart found the same pattern of patrilateral inheritance of leadership for each "district1' of people (1933: 304), with each family group maintaining a permanent but not always occupied village

(1933: 238). Since each village "comprised enlarged families plus a few others, all regarded as relatives", and since blood relationship prohibited marriage, each village may be assumed to have been virtually exogamous (1933: 294). He described the residence pattern as follows:

Residence was matrilocal for about one year, ... the husband providing his family-in-law food, sometimes aided by his brother-in-law ... A year of patrilocal, then independent residence followed, the place chosen by the girl, the house built by the boy's family. They visited their families often and per• manent residence tended to be matrilocal. Their parents, especially the wife's, lived with them, housekeeping and raising the grandchildren. (1933: 295)

In a survey of three Northern Paiute bands, Steward found one with a variable pattern of residence, one with both a variable pattern and a known, but not normal, pattern of matrilocal residence, and one with both variable and matri• local residence and a known, but not normal, pattern of patri• local residence (1941: 311).

Steward surveyed the twenty-one Northern Paiute bands and concluded that "leadership, . . . was a;,reward of person• ality and experience, not inheritance" (1939: 130). He found 117. an absence of anything other than a very vague concept of individual resource-use rights, "inherited matrilineally and patrilineally", and an equally indefinite idea of territorial ownership (1939: 130). He also noted that "no definite rule for residence or exogamy existed" in any of the twenty-one bands (1939: 130).

Later, the same author conducted a more intensive survey in which he clarified the residence patterns of twelve bands as patrilocal, matrilocal, variable and/or matrilocal until the birth of a first child. Of the twelve bands, five had all of the first three patterns, five had only matrilocal residence, one had all four patterns and the twelfth had all but patrilocal residence (1941: 405).

Lowie found a general pattern of initial matrilocal residence among the Northern Paiute, but did not state if it was normally a life-long pattern (1924: 276). Park found matrilocal residence among the Northern Paiute (1937: 367).

In a survey of four southern Paiute bands, Stewart noted matrilocal residence for all four and an additional pattern of variable residence, known but not a definite normal pattern, in one band (1942: 296). In addition, his informants specifically denied any knowledge of either patrilocal resi• dence or any preferential form of marriage (1942: 331). Lowie had earlier found a pattern of matrilocal bride service for 118. the Shivwits Southern Paiute, but did not elaborate on the subject of residence generally (1924: 275).

For the Ute, Stewart found matrilocal residence for all eight bands covered by a survey, with a variable pattern of residence known, but not established as normal practice, in one band (1942: 296). Again, informants specifically denied the possibility of patrilocal residence (1942: 331).

Lowie noted two cases of patrilocal residence, however, one among the Uintah Ute where a couple would first live matri-

locally and then "with the husband's family" (1924: 275), and among the southern Ute generally, where no initial period of matrilocality was noted (1924: 275).

There are a large number of references for the Shoshoni bands by three writers over a period of almost twenty years.

Lowie found a pattern of matrilocal residence among the Wind

River Shoshoni, but noted that it "was not universal"

(1924: 278). Steward stated that residence among the Lone

Pine Shoshoni "was matrilocal until childbirth, then independent" (1933: 296). For the Nevada Shoshoni, he found

that "where true bands occurred, they were composite, that is, consisted of unrelated families who intermarry within the band. But most localities lacked even bands and had only scattered, independent villages which, by reason of lack of consistent post-marital residence rules, comprised unrelated 119. and therefore dntermarriagable families" (1941: 252). He i went on to state that "residence following a marriage consummated through present exchange (the normal,pattern) tended to be matrilocal at first, in some cases even being preceded by a kind of bride service of some months. . . but was subsequently variable. Permanent residence depended upon food supply, location of other relatives, and various matters of individual preferences" (1941: 252). In a survey of fifteen Shoshoni groups he noted a variable pattern of residence for all fifteen, a combination of patrilocal and either definite or previously known, but not normal, matrilocal residence in addition in eight, and previously known, but not normal, matrilocal residence in addition in one (1941: 311).

He also noted independent residence for some groups (1941:348).

For the Hekandika Shoshone, Hoebel found that

"residence was bilocal in extended family households or groupings within the band. • . Since marriage was largely within the band, it was easy to shift residence to either parents and back" (1939: 446). The band as a whole may well conform to the endogamous deme of Murdock (I960: 62-63), but if he means just a single local group, the local group, if it was smaller than the band, was not endogamous.

In discussing the Northern and Gosiute Shoshoni,

Steward noted that "residence was usually matrilocal for a 120.

year or so, which was more or less a period of bride service.

After that, it was independent, the location being determined

by circumstances of subsistence and relationship" (1943: 279).

For the Lemki and Idaho bands of Northern Shoshoni,

Steward found that while the local group was not exogamous by

rule, local exogamy was probably a norm in practice (1938:

194). He went oh to state that "residence was matrilocal

for a time, during which the couple often visited the husband's family. Eventually they moved to permanent patri•

local residence. If the husband's father were dead, however,

they might remain in matrilocal residence" (1938: 195). In

a survey of four Northern Shoshoni bands he found variable

patterns of residence in three, and one band with temporary matrilocal residence followed by patrilocal residence

(1943: 337).

For the Western Shoshoni, the principal economic

activities were conducted by "independent households under

the leadership of the household head" (Steward, 1938: 239).

These consisted of nucleur families of parents and children,

together with any other relatives who had no other home

(1938: 239-240).

In a survey of three Western Shoshoni bands, Steward

found variable residence for two, and an initial period of matrilocal residence followed by patrilocal residence for one 121.

(1943: 337). He discussed the Grouse Creek band of Western

Shoshoni in some detail. One informant claimed that

"marriage should be matrilocal for about a year, when the couple visited the husband's family. They then returned to the wife's family, where they lived until two or three children were born. Then they made a separate home near the wife's family" (1938: 176). Steward's data did not indicate any period of initial temporary residence, however

(1938: 176). Instead, he found that out of fourteen cases, six were patrilocal, three following marriage to a member of a distant group, and four were matrilocal, two following marriage to a member of a distant group, while in three cases both spouses moved to a new locality and in one case, both remained in the same place (1938: 176-177). It was his opinion that "the bulk of residence would be neither patri•

local nor matrilocal with respect to locality" (1938: 177).

He noted that marriages were considered "alliances which bind families" (1938: 177), and concluded by rejecting the idea of

"rigid rules of exogamy or post-marital residence" (1938: 177).

Steward then summed up the Western Shoshoni pattern of residence as follows:

Practices respecting post-marital residence, though slightly weighting woman's status, failed materially to alter family balance among Western Shoshoni. Tempprarily, residence was almost everywhere matrilocal, the husband often spending a year or more in a kind of 122

bride service during which his general acceptability and hunting prowess were ascertained. Sometimes residence was matri• local until several children had been bom, because the aid of the maternal grandmother was sought in childbirth and infant care. Such temporary residence seems to enhance the status of women. . . But permanent residence among the Western Shoshoni was determined solely by certain practical and personal circumstances: 1, abundance of local foods; 2, the whereabouts of relatives; thus, when a brother and sister married a sister and brother, the two families prepared to live near each other; also, proximity to other favourite relatives might be sought; thus, if the parents of only one spouse were living, the couple might settle near them; 3, whether a headman of some kind was in the family; relatives often preferred to live near such men; 4, whether marriage were with a cross-cousin, pseudo cross-cousin, or unrelated person from the same village, none of which required change of residence; 5, whether marriage were by capture, the woman always being brought to the man's village (1938: 243).

The data for the entire region show- some contradic• tions and conflicts. Nonetheless, Steward has summed up what

I consider the most likely pattern of residence for the area.

It will be remembered that the basic subsistence unit of this region consisted of from one to three families; the territory simply would not support any greater number of people in a permanent local group. A non-corporate cluster, this unit wandered the collecting and hunting grounds most of the year, 123.

settling in one spot for a few months during the winter,

often with other families, but only rarely interacting ec•

onomically with other local groups. As Steward implied in

his statement concerning marriage alliances among various

bands, marriage was highly functional in the context of Great

Basin human ecology; should one local group be depleted in

population, the remaining members had kin to whom they could

turn; similarly, the risks ofoverpopulation was reduced by

the fact that each local group had additional units to which

one or two members would likely be able to attach themselves

in time of need; He noted later that "the preferred arrange• ment was several marriages between the children of two families

(1955: 118), a factor which "meant that a family in a given

locality could probably find consanguineal or marital kin

of one kind or another among a large proportion of the families

which ranged its own territory and among many families father

afield" (1955: 118-119).

With specific regard to the pattern of residence of

the Great Basin Shoshoneans, I can only state that, as in

all societies, the specific residential affiliation of an

individual male was based both on sociological and on what

Dunning termed situational factors (1963: 2), but that in

this case no specific unilocal sociological preference was 124. workable to anything more than a very slight extent. That is,

residential alignment was based to such a large extent on

the specific situations of individual local groups and persons

that it is not possible to speak of a rule, preference or

even statistical regularity of unilocal residence, and that

it would be better to refer to the area as one characterized

largely by statistical irregularity in residence pattern. CHAPTER VI

CONCLUSIONS

I stated at the beginning of this paper that I had two main purposes ih mind, the consideration of factors governing local group composition and an examination of various patterns of residence in an attempt to draw valid distinctions among rules and preferences of residence. So far, I have outlined the basic patterns of economy, social structure and residence that appear to be relevant to these two purposes for a total of thirty North American Indian tribes in four selected subsistence areas of the Continent.

Now it is possible to sum up the theoretical points which can be drawn from the descriptive material on these societies.

The Local Group

Man as a social animal almost always exists within a localized unit of society founded on the basis of kin ties, with the other members of which he interacts socially, economically and politically, and which is socially and spatially distinct from other units of the population. It is this association which I have termed the local group. This is a minimal definition only, and there are undoubtedly groups which fit this pattern but which might, under certain circum• stances, be considered either more than one local group or but 126. a segment of a single local group. Nonetheless, precise definition of such a unit is impractical, since it would probably exclude many cases of relevance and include many of little significance. As will become clear later in this discussion, a more exact definition of the local group must rest on specific cases. For the time being it can be accepted that the local groups with which we are here primarily concerned - as examples for a series of theoretical premises and propositions - are the clan or bilateral kindred village segment of the Northwest coast, the "family" hunting band of the Sub-arctic, the large hunting band of the Plains and the small "family" collecting band or village of the Great Basin region. All of these have been defined in greater detail in the descriptive chapters of the thesis. There are certain general factors which govern the size, composition and nature of the local group which are now evident. These will be considered in order.

Local Group Size

Ecology has been defined as the adjustment or balance of all living things to a given non-organic region. In anthropology, it is principally the study of the "balance of man and nature" and includes such variables as human 127. technology. It is, of course, a key point in any considera• tion of local groups. The size of any local unit may be said to have two limits, a minimum below which the group may not fall if it is to continue to exist, and a maximum above which it may not grow for the same reason. Although neither of these is definable in absolute terms, they are both keyed to the ecology of a given area. We have seen, for example, that the hunting band of the Eastern Sub-arctic tribes cannot be allowed to grow too large; if it reaches a point in size where the economic return of the hunting area is insufficient to enable the group to maintain itself, or even to maintain a reasonable standard of living as that is con• ceptualized by the members of the local group, then some individual or individuals must leave the unit. Similarly, if it grows too small to permit the effective utilization of the resources available and runs the same risks as the overpopulated group, new members must be attracted from outside the unit. The size of a specific area exploited by a local group varies with the resources, both human and animal, available. This can be seen in the Ojibwa bands, among which there is considerable variation in population density, based on human (number and skill of hunters and their support organization of women and children etc.) and natural

(conditions etc., number and type of animals and the cyclical environmental conditions) resources available. There are, then, 128. certain definite but not precisely definable (in absolute

terms) limits to the size of local groups, based on purely ecological circumstances.

There is, however, a wide diversity of range possible between the upper and lower limits of local group size. We have seen, for example, that on the Plains, bands ranged in size from slightly over one hundred people to several hundred people. On the Northwest coast the range was, if anything, even greater, while in the Sub-arctic and Great Basin regions it was comparatively narrow. This diversity of range is based on ecological and economic factors such as the amount and period of year of food supply availability, the type of food supply and the technology in use by the people, whether surpluses could be obtained, and if so, if they were storable,

transportable etc. Among the Great Basin tribes, for example, some food supplies could be stored, but only some, and it was not worthwhile to store more than a certain amount at any one

time. While the placing of a limitation on the amount stored could cause some hunger in the local group, the alternatives of missing, say, an antelope hunt to continue laying in a store, or of slowing group movement by having to transport

large quantities of goods, could cause equally great problems.

On the Northwest coast, however, most food resources were relatively nearby and bountiful most of the time. Both surpluses and storage were more easily possible. 129.

It must be understood that the precise size of a

local group was determined by a variety of factors. Ecology

alone did not force groups into such rigid limits that they

had to maintain a certain specific size. Rather, the

ecology of a society simply set outer limits on the size of

local groups, above which they could not rise and below which

they could not fall. Within those limits there was room for

considerable variety of size, however; thus the range of sizes

possible for local groups, but not the precise size of any

specific local group, was based primarily on ecology and

economy. These may be termed situational factors, after

Dunning (1963ms: 2). He described them in part as the "life

cycle, technological skill, (an

individual persons" and demography. In short, the limits on

size of any local group were governed largely by a series of

factors over which the members of the local group did not

normally have direct control, the variables of a specific

ecological situation.

Local Group Composition

Given the situational factors of a local group and the

outer limits on the size of the group, its composition will vary according to what may here be termed the sociological

factors, the individual and group predelictions for attachment 130. and/or affiliation with a particular individual or set of individuals. Included in this classification are factors such as descent system and kinship terminology, marriage and mating patterns and regulations, status and political power, religious beliefs and practices and, finally, the individual idiosyncrasies which drive some humans into patterns of behaviour contrary to or just different from those of their neighbours. Despite the wide range of patterns which prevails in North America, it is possible to make some general obser• vations about the sociology of local groups.

Firstly, despite the presence of unilateral descent systems in many tribes, nowhere did we note a local group which constituted purely and simply a unilateral descent group.

Aside from the addition of affinal kin in the form of spouses, there were always, in virtually every tribe, some additional persons in the group. On the northern Northwest coast, j individuals attached themselves to those of their kin from whom they might gain some advantage. In the south the descent system was bilateral. In the Sub-arctic and in the Great

Basin, such purely unilateral arrangements were not possible because of the situational factors involved, and on the Plains the descent system was bilateral. The basic effect of the descent systems in every case was simply the creation of the dichotomy of kin and non-kin. Everywhere, however, individuals tended to associate with their kin wherever this was possible 131. and desirable, so that in virtually every case the members of each local group were related in one way or another to some, if not all, of the other members of the local group.

Secondly, there was everywhere an emphasis on the freedom of the individual wherever this was possible in light of the situational factors involved. Indeed, although some societies maintained certain institutionalized forms of inducement for the affiliation of certain individuals to specific local groups, this may be considered in one sense as a method of overcoming the individual tendencies of associational choice recognized in some societies, and as a method of encouraging specific general types of local group composition.

Thirdly, a variety of other sociological factors influenced the precise composition of local groups. Inheri• tance patterns and marriage preferences tended to lead towards certain sociological groupings of people, for example, while individual factors such as religious position, economic skill or personal characteristics often made the affiliation of an individual with a particular local group either desirable or undesirable, both from his point of view and from theirs.

There is still the question of size within the limits set by situational factors, however. It should be pointed out that sociological factors can limit size as well, 132.

although only, within the outer limits imposed by the

situational factors. For example, the Northwest coast size

limits are quite far apart. Nonetheless, most local groups are well within the possible limits. In some cases, marriage within the local group is a preferred form, one

factor which helps to maintain a smaller size than need be.

Again, while there are advantages in terms of prestige and the ability to manipulate other persons in being the leader of a large local group, there is a point at which the "law" of diminishing returns begins to operate, and it may well be

that a man is actually in a better social position as leader of the largest of several local groups in a village than as

leader of the same village if it constituted but a single local group. All of these sociological factors tend to operate within the framework of the situational factors, however, and may be slightly altered by them to meet specific requirements. For example, there is the" so-called "fictive" marriage of the Northwest coast, under which, if situational factors fell below the ideal, a man might make a sociological arrangement to compensate for the failure and; thus in effect maintain the sociological norm by an usual and abnormal method. 133.

Local Group Nature

The precise nature of a specific local group, its

size and composition and the various ties which serv:e,: to

keep it distinct from other local groups, will depend upon a

balance of situational factors and sociological factors. The

former will provide certain limits beyond which the group

cannot exist, and the latter will provide the specific detail,

the sociological configuration of the group, within those

limits. Generally speaking, it would seem safe to suggest

that the greater the range between the two limits on size,

the more economically secure the local group and the greater

the possibility of sociological variation and complexity.

Size alone is relatively unimportant; this can be seen by a

comparison of the small possible size of a Northwest coast

local group, with its complex ranking system, and the size of

the Sub-arctic local group which may be only slightly smaller

than the Northwest coast group, and of the large size of the

Plains band in relation to the size of the largest conceivable

local group on the Northwest coast, which could not exceed more than several dozen people.

One further area of interaction between sociological

and situational factors may now be noted. This is the 134. corporate local group. I have already defined the corporate group (in Chapter I) and it is now possible to discuss some of the factors governing their presence or absence in a specific area.

Firstly, the situational factors of a given area may not permit the development of such groups. One of the re• quisites of corporateness is a specific territory used regularly by a certain group and only by that group. Another is the ecological possibility of a fairly consistent compo• sition of the local group. Neither of these is strictly possible in the Great Basin region, for example, and such groups are accordingly lacking. In the Sub-arctic, this form of corporateness was possible in the east, but only barely, and we can note enough variation from the ideal of local group composition to suggest that, while the groups were cor• porate, they were very often that by just a slim margin and no more. Such groups were also permitted by situational factors in the Northwest coast region, but here the ecology of the society did not place such rigourous limits on group size and composition, and corporateness was workable to a much greater degree, to the extent that variation from the ideal was not nearly as frequently required.

This brings up a second point, the influence of sociological factors on group corporateness. On the Northwest 135. coast, the situational factors created the possibility, but it was the sociological factors which determined that the local group would actually be corporate. On the Plains, however, the specific cultural milieu of war and peace, opposition and alliance, a sociological factor, combined with the migratory nature of the buffalo herds and the demography of the region, both situational factors, to make corporate local groups impractical in terms of Plains culture. Although strictly speaking they were probably feasible in terms of situational factors alone, the sociological factors of the area rendered them undesirable.

Finally, then, there is the possibility, outlined by

Murdock, of the endogamous deme. In two of the three tribes of this sample listed by Murdock as having endogamous demes, the Kiowa-Apache and the Shoshone, the endogamous unit was larger than the local group. In the third case, the Commanche, the larger local groups were apparently endogamous, and changes of residence within these generally had no effect on the composition of the groups.

A comparison of the nature of local groups in the four selected areas of this paper runs as follows:

1) Northwest coast:

Considerable variation in size was permitted by

situational factors, from a single nucleur

family to several dozen families. The largest 136.

local group was several times larger than

the smallest. Sociological factors governing

local group composition as ideals and norms of behaviour were virtually unhindered by situ• ational factors. Complex patterns of sociological factors determined precise local group composition. Local groups were corporate, and the corporations were based almost entirely on consanguineal relationships.

Plains:

There was a high ratio of the size of the largest possible local group to that of the smallest, although the ratio was probably not as great as on the Northwest coast. The sociological factors governing local group composition were, as on the Northwest coast, almost unhinctered by situational factors.

Although the latter normally operated only to determine the limits of local group size, they occasionally played some part when a decision of local group affiliation had to be made.

The local groups were not corporate. 137.

Sub-arctic:

Here the ratio of largest possible group size to smallest possible local group size (and here it is important to note that this refers to possible local group size in a single given area) was quite small. Accordingly, the sociological preferences which tend to set ideal standards for local group composition were regularly subordinated to situational factors. In the eastern section there were corporate local groups, but they were based as much upon affinal relationships as on consanguineal relationships. Local groups were not corporate in the Western sub-region.

Great Basin:

In this region the ratio of the size of the largest possible local group to that of the smallest was very low, so much so that while certain basic ideals of local group composition did exist, the sociological factors which ever actually came into play to determine that composition were minimal in scope and even then not always possible. There were no corporate local groups. 138.

There is, then, a continuum of sociological and situ• ational factors governing the nature of the local group. The two exist in balance, but in some societies the sociological factors may outweigh the situational factors in determining local group nature, while in others the reverse may be true.

With these basic aspects of the local group now established, it is possible to move on to a general discussion of patterns of residence.

Patterns of Residence

I have noted a wide variety of patterns of residence in the thirty tribes outlined in the descriptive chapters of this thesis. Very broadly speaking, these can be classified into three general changes of local group affiliation by a male: changes prior to marriage, changes at the time of marriage, and changes after marriage, but not directly con• nected with the union. In addition, several common elements can be noted. Firstly, when individuals changed residence it was almost invariably a move to another local group in which they had kin. Secondly, these changes of residence usually, but not always, meant a change in social status for the person moving. Thirdly, the reasons for change were only rarely directly related to situational factors, and then normally only in specific subsistence areas, a point which 139. will be considered in detail later. Finally, each of these

residence changes must be considered apart from the insti•

tution of marriage, even though the classification above is

keyed to the time of marriage. Each of the three types of

residence change can now be discussed in detail.

Changes prior to marriage. Two basic forms of

behaviour can be noted here. Firstly, there was the insti•

tutionalized pattern of movement from ego's own local group

to that of the person from whom he would inherit, his maternal uncle, on the Northwest coast, This move usually

came at.about the age of ten, and may or may not have been

permanent, depending upon the precise social position of the

two individuals in question. This move was only slightly,

if at all, related to situational factors.

Secondly, there was the move by a male to the local

group of his future wife's father or brother. This is

normally termed a period of bride service, and may or may not have been followed by further moves after marriage. The

question of just when the couple were married - of whether

the bride service was pre- or post-marital - is purely academic

and of little significance here since it can fairly be assumed

that actual marriage occurred some time following the move,

recognition of the formal nature of the union coming perhaps 140. over a period time. This type of move was found in all areas surveyed, in greater or lesser degree, although in some specific cases it was either not practical as an institu• tionalized and regularly followed form, or was undesirable due to a combination of situational and sociological factors.

Changes at marriage. These occurred in virtually all tribes surveyed, but, from the point of view of a male, were a standard form in only some of the tribes. In not all cases, however, did a change necessarily occur at the time of marriage, although there were no local groups found which were either endogamous or exogamous as such. (Some, of course were exogamous by kinship.) Generally speaking, changes at this time were almost always to the local group of a man's wife. Only rarely, and then usually for situational factors, did a man associate with a local group other than that of his own family of orientation, or that of his wife, on a per• manent basis after marriage. The two terms matrilocal, referring to residence with the local group of a man's wife, and patrilocal, referring to residence with a man's own local group (and, in many cases, referring to a lack of local group change), are standard for the two basic patterns of residence found at the time of marriage. 141.

Changes after marriage. In many cases, a man moved to another local group at some time after marriage, usually, if on a permanent basis, for personal reasons, the socio• logical factors. Often, however, a man might alternate his residence between two local groups, or wander among several, affiliating with first one and then another, either for personal reasons or because of situational factors. Even so, the general preference was always for residence with a local group in which one had some kin.

The three standard possible patterns here were the matrilocal and patrilocal patterns noted above, and the avunculocal pattern of residence with the local group of the maternal uncle, noted earlier. In each case, these patterns led to certain standard groupings of kin in local groups, a point which will be outlined in greater detail later.

The General Hypotheses

It is now possible to consider some of the reasons for specific types of residence change and local group affiliation. This is best done, however, in the context of the two general hypotheses governing rules and preferences of patterns of residence. The two hypotheses must be consi• dered two parts of a single unit, with the basic distinction that of the presence or absence of corporate local groups. 142.

1) I proposed initially that in all societies

with corporate local groups there would be

either prescribed unilocal rules of residence,

defined as cultural preferences for a unilocal

pattern of residence, with both positive and

negative sanctions involved for success and

failure in following the pattern, or there

would be preferred unilocal rules of residence,

defined as cultural preferences for a unilocal

pattern of residence, but with the absence of

any sanctions for success or failure in following

the pattern.

In actual fact, in none of the societies examined were

there any sanctions of the type noted above directly connected

with residence. Where and if ever such sanctions came into

play, they were either omitted from mention in the literature

or were so rare and abnormal as to be insignificant for this

discussion. Obviously, then, if the word rule is to be

retained, some other criteria must be noted for its definition.

Two basic forms of behaviour were noted here. There were those societies which had a cultural preference for a

specific unilocal pattern of residence, and those with a very definite unilocal pattern of residence but a cultural ideal 143. only for affiliation with the local group in which an individual had kin and would find the greatest possible advantage. The

former case may be termed the prescribed rule of residence.

It was not always possible, however, for individuals to

follow this, due to situational factors. Nonetheless, the

cultural preference for a single unilocal pattern was present and was expressed in such factors as social status, ideal

land rights inheritance, and political leadership and power.

In the other case, the unilocal pattern occurred for reasons other than a specific unilocal cultural preference. This was almost always the case for reasons of material gain

following a pattern of inheritance. Interestingly enough, variations from the pattern in these societies usually occurred for reasons of sociological rather than situational

factors, although the latter was possible. Naturally, the

cultural preferences of those societies with prescribed rules

sharply decreased the possibility of variation due to

sociological factors.

It should be noted here also that the effect of cor•

porate local groups was to maintain a given cultural

preference. The very fact of corporateness acted as a stimulus

to certain patterns of residence, either the single unilocal 144. pattern of the prescribed rule, or the most advantageous pattern, not definitely expressed as a single unilocal pre• ference, of the preferred rule. The effect of corporateness was to deny prestige and high status to the individual who failed to follow the preference. It was, in fact, a failure to recognize other possible patterns of residence as socially acceptable, unless situational factors decreed an alternative as necessary. In the one region where this latter was rarely a factor, a failure to follow the cultural preference may have meant an inability to affiliate with any local group of more than very low status, but this did not apparently constitute an institutionalized sanction, although the literature is not clear on this point.

There are, then, two forms of rules, the prescribed rule with a single unilocal pattern of residence as a cultural preference expressed in certain institutional mechanisms such as ideal inheritance and status-movement patterns, and the preferred rule with a cultural preference not expressed as a single unilocal pattern per se. but resulting in such a pattern nonetheless.

2) I proposed that all those societies which did

not have corporate local groups would be totally

lacking in cultural preferences for single

unilocal patterns of residence, but that they

would be divided into two types, those with a 145.

single unilocal pattern of residence in

statistical regularity, and those with

only statistical irregularity in residence

pattern.

The hypothesis continues to stand. No cultural pre• ferences involving high status for one pattern and low status for another were found. Some societies had more than one unilocal pattern of residence. In some there were definite statistical regularities, permitted by the situational factors of the local groups, while in others the situational factors involved prohibited any statistical regularities of significance.

In the case of both hypotheses there were three basic patterns of unilocal residence, the patrilocal, under which males affiliated with the local groups of their fathers, in which case the local group tended to be based on father/son and male sibling consanguineal relationships and female affinal relationships, the matrilocal, under which males tended to affiliate with the local groups of their wives, under which the local groups tended to be based on mother/ daughter and female sibling consanguineal relationships and male affinal relationships, and the avunculocal, under which 146. males affiliated with the local groups of their mothers' bothers, in which case the local groups tended to be based on mother's brother/sister's son and male sibling consangu•

ineal relationships and female affinal relationships. These were ideal patterns only, however, and were rarely absolute

types in terms of actual local group composition.

The Specific Hypotheses

I had originally proposed the precise pattern of

residence which would be found in each subsistence area, and

the reasons for those patterns. These hypotheses can now be considered.

1) For the Northwest coast, I proposed a

prescribed rule of residence either

patrilocal or avunculocal, depending upon

the sub-region of the area. This was based

on four premises; the possibility of large

scale wealth accumulation and class

differences keyed in part to wealth, the

corporate and kin-based nature of the local

group, the sociological configurations of

the societies, and the situational factors

which did not require any great flexibility 147.,

in residence pattern. I suggested that

the unilocal pattern of residence would

follow the inheritance and succession

patterns of the local groups.

In fact, the local groups were corporate and kin-based, wealth accumulation was possible and there was little danger of overpopulation. Class differences were those of freemen and slaves only, but a ranking system existed, based partly on wealth. The early definition of the prescribed rule was found to be invalid, and, with.one exception, there do not appear to have been any single unilocal cultural preferences.

In the northern sub-region there was an initial avunculocal prescribed rule of residence prior to marriage, followed by a matrilocal preferred rule of residence after marriage.

Variations for sociological reasons were common. In the southern sub-region there was a patrilocal preferred rule of residence following marriage, but variations occurred, again due largely to sociological factors. In all cases, the unilocal patterns tended to follow inheritance and succession patterns, although this was not always true in individual cases. 148.

2) For the sub-arctic region, I proposed a

preferred rule of patrilocal residence and

corporate local groups. I stated that the

pattern would be based on a norm of patri•

lateral inheritance and the ecological

necessity of familiarity with the local

group area by the hunter. I suggested that

wealth accumulation would not be possible

on any major scale, and that flexibility

in residence pattern would be necessary due

to situational factors.

In fact, differences were found between the two sub- regions of the area. In the eastern section there were corporate local groups, but wealth accumulation was not possible on any great scale. There was a cultural preference for patrilocal residence, however, and despite a flexibility imposed by the situational factors involved, this may be termed a prescribed rule, as defined earlier in this chapter. The ecological necessity of familiarity with the local group area by the hunter and the norm of patrilateral inheritance, while definitely wielding some influence over the unilocal pattern, were not vital factors. The latter was simply an expression of the cultural preference. In the western section there 149. were no corporate local groups. In two tribes there was a statistical regularity of patrilocal residence. In the third, the pattern was uncertain, due to conflicting state• ments in the literature. In both sub-regions there was a period of pre-marital matrilocal residence, in the east as a prescribed rule, in the west as a statistical regularity.

3) I proposed that the tribes of the Plains

region would lack corporate local groups,

and that the situational factors of the

area would permit the development of large

bands of people. Further I suggested that

while unilocal patterns might occur in

statistical regularity, there would be no

single unilocal pattern of residence as a

cultural preference, but rather only a

sociological preference for residence in a

local group in which the individual had

some kin. I suggested that this would be •

based in part upon the lack of corporate

local groups and on the bilateral descent

system.

In fact, this was the case found after an examination 150.

of the ethnographic literature. There were both patrilocal

and matrilocal statistical regularities in the. region,

usually only one in each tribe, but there were no expressed

cultural preferences as( defined above, In addition, some

societies showed a pattern of matrilocal residence prior to marriage, as an institutionalized form.

4) For the Great Basin region, I proposed the

absence of any corporate local groups,

cultural preferences for specific patterns

of unilocal residence, or statistical

regularities of unilocal residence of any

significance, all due to situational factors.

In fact, this was correct and the hypothesis stands.

There were no corporate local groups, expressed cultural

preferences as defined above, or significant statistical re•

gularities. There was a general sociological norm of pre• marital matrilocal residence among some tribes, but this

was not always possible due to situational factors.

The latter two regions, the Plains and the Great Basin,

together with the Western sub-region of the Sub-arctic, none

of which areas had corporate local groups, appear to have had

only sociological preferences for residence with kin, a type 151. of preference found in all societies surveyed in this paper.

No specific single unilocal preferences were found in these groups.

As I had originally postulated, then, a close examin• ation of the ethnographic literature demonstrates a wide difference between the accepted brief "rules of residence" and the actual patterns of behaviour governing residence and the local group in North America. 152.

BIBLIOGRAPHY CODE

AA American Anthropologist, New Series, Menasha, Wisconsin. AAA-M Memoirs of the American Anthropological Associa• tion, Menasha, Wisconsin. AMNH-B Bulletins of the American Museum of Natural History, New York. AMNH-M Memoirs of the American Museum of Natural History, New York. AMNH-AP Anthropological Papers of the American Museum of Natural History, New York AN Anthropologica, First Series, Ottawa. ANTH Anthropos. APS-P Proceedings of the American Philosophical Society, Philadelphia. APS-T Transactions of the American Philosophical Society, Philadelphia. AR University of California Anthropological Records, Berkeley. BAAS Annual Reports of the British Association for the Advancement of Science, London. BAE-B Bulletins of the Bureau of American Ethnology, Washington, D.C. BAE-R Annual Reports of the Bureau of American Ethnology, Washington, D.C. CNM-B Bulletins of the National Museum of Canada, Ottawa. CU Columbia University Contributions to Anthropology, New York JRAI Journal of the Royal Anthropological Institute, London. PM Primitive Man, Washington, D.C. UCPAAE University of California Publications in American Archaeology and Ethnology, Berkeley and Los Angeles. 153.

USNM Annual Reports of the United States National Museum, Washington, D.C. Y Yale University Publications in Anthropology, New Haven.

The first Arabic number following a listing indicates a primary reference, i.e. volume or series number; the second Arabic number indicates a secondary reference, i.e. issue number or part number. Arabic numbers separated by a hyphen indicate a page reference; a second set of hy• phenated Arabic numbers indicates a year, as does a set of four numerals following a reference. 154.

BIBLIOGRAPHY

Allard, E. (1929): Notes on the Kaska and Upper Liard

Indians, PM, 2, 1 and 2, 24-26.

Barnett, Homer G. (1955): The Coast Salish of British Columbia.

Eugene, Oregon.

Bellah, Robert H. (1952): Apache Kinship Systems. Cambridge,

Mass.

Boas, Franz (1890): First General Report on the Indians of

British Columbia. BAAS, 1889.

(1891): Second General Report on the Indians of

British Columbia. BAAS, 1890.

(1892): Third Report on the Indians of British

Columbia. BAAS, 1891.

(1894): The Indian Tribes of the Lower Fraser

River. BAAS, 1894.

(1895): Fifth Report on the Indians of British

Columbia. BAAS, 1895.

(1897): The Social Organization and the Secret

Societies of the Kwakiutl Indians. USNM, 1895.

(1916): Tsimshian Mythology. BAE-R, 31,311-738,

1909-10.

(1920): The Social Organization of the Kwakiutl.

AA, 22, 11-126. 155.

Boas, Franz (1921): Ethnology of the Kwakiutl. BAE-R, 35,1,

43-794, 1913-14.

(1924): The Social Organization of the Tribes of

the North Pacific Coast. AA, 26, 323-332.

(1925): Contributions to the Ethnology of the

Kwakiutl. CU, 3.

Bohannan, Paul (1963): Social Anthropology. New York.

Burgesse,, J. Allan (1944): The Woman and the Child Among the

Lac-St-Jean Montagnais. PM, 17, 1 and 2, 1-18.

(1945): Property Concepts of the Lac-St-Jean

Montagnais. PM, 18, 1 and 2, 1-25.

Cameron, Duncan (1960): The Nipigon Country, in Masson,

Volume II (q.v.).

Colson, Elizabeth (1962): The Plateau Tonga of Northern

Rhodesia. Social and-Religious Studies,

Manchester.

Colson, Elizabeth and Gluckman, Max (Editors) (1951): Seven

Tribes of British Central Africa. Manchester.

Davidson, D.S. (1928a): Family Hunting Territories in

Northwestern North America. Indian Notes and

Monographs, 46, New York.

(1928b): Notes on Tete de Boule Ethnology. AA,

30, 18-45. 156.

Dorsey, J. 0. (1897): Siouan Sociology. BAE-R. 15, 213-44,

1893-94.

Douglas, Mary (I960): The Pattern of Residence Among the Lele.

in Ottenberg and Ottenberg, 1960b (q.v.).

Driver, Harold E.(1961){Indians of North America. Chicago,

London.

Driver, Harold E. (1961): Indians of North America. Chicago,

London.

Driver, Harold E. and Massey, William C. (1957): Comparative

Studies of North American Indians.APS-T. New

Series, 47, 2.

Drucker, Philip (1939): Rank. Wealth and Kinship in Northwest

Coast Society. AA, 41, 55-65.

(1951): The Northern and Central Nootkan Tribes.

BAE-B, 144.

(1963): Indians of the Northwest Coast. Garden

City.

Dunning, R. W. (1959a): Rules of Residence and Ecology Among

the Northern Oiibwa. AA.,Q- .806 -Q16.

______(1959b) : Social and Economic Change Among the

Northern Oiibwa. Toronto. 157.

Dunning, R. W. (1963ms): An Aspect of Domestic Group Structure.

Among Eastern Canadian Eskimo.

Eggan, Fred (1955a): The Cheyenne and Arapaho Kinship System

in Eggan, 1955b (q.v.).

Eggan, Fred (Editor) (1955b): Social Anthropology of North

American Tribes. Enlarged Edition, Chicago and

London.

Emmons, G.T. (1911): The Tahltan Indians. University of

Pennsylvania, The Museum, Anthropological

Publications, 4, 1, Philadelphia.

Faron, L. C. (1961): Mapuche Social Structure. Institutional

Reintegration in a Patrilineal Society of Central

Chile, Illinois Studies in Anthropology, 1, Urbana.

Flannery, Regina (1953): The Gros Ventres of Montana: Part I.

Social Life. The Catholic University of America,

Anthropological Series, 15, Washington, D.C.

Ford, Clellan S. (1941): Smoke From Their Fires. New Haven.

Forde, Daryll (1950): Double Descent among the Yakb. in

Radcliffe-Brown and Forde (q.v.).

______(1963): Habitat. Economy and Society. London.

Fortes, Meyer (1953): The Structure of Unilineal Descent

Groups. AA, 55, 17-41 158.

Franklin, John (1910): Narrative of a Journey to the Shores

of the Polar Sea in the Years 1819-20-21-22.

London.

Fried, Morton H. (1957): The Classification of Corporate

Unilineal Descent Groups. JRAI, 87, 1, 1-29.

Garfield, Viola E. (1939): Tsimshian Clan and Society.

University of Washington Publications in

Anthropology, 7, 3, 167-340, Seattle.

Garigue, Philip (1957): The Social Organization of the

Montagnais-Naskapi. AN, 4, 107-135.

Gladwin, Thomas (1948): Comanche Kin Behaviour. AA, 50, 73-94.

Gluckman, Max (1950): Kinship and Marriage Among the Lozi

of Northern Rhodesia and the Zulu of Natal, in

Radcliffe-Brown and Forde (q.v.).

(1951): The Lozi of Barotseland in North-Western

Rhodesia, in Colson and Gluckman, (q.v.).

Grant, Peter (1960)5 The Salteaux Indians, in Masson, Volume

II (q.v.).

Grinnell, George Bird (1962): The Cheyenne Indians: Their

History and Ways of Life. 2 Volumes, New York.

Hallowell, A. Irving (1949): The Size of Algonkian Hunting

Territories: A Function of Ecological Adjustment.

AA, 51, 35-45. 159.

Hanks, L.M. and Jane Richardson (1945): Observations on

Northern Blackfoot Kinship. Monographs, the

American Ethnological Society, 9, New York.

Hassrick, Royal B. (1944): Teton Dakota Kinship System. AA,

46, 338-347.

Harrison, Charles (1892): Family Life of the Haidas (Queen

Charlotte Islands), BAAS, 1891.

Helm, June (1961): The Lynx Point People: The Dynamics of a

Northern Athapaskan Band. CNM-B, 176.

Hill-Tout, Charles (1900a): Notes on the N'tlaka'pamuq of

British Columbia, a Branch of the great Salish

Stock. BAAS, 1899.

(1900b): Notes on the Sk.qo'mic of British Columbia.

a Branch of the great Salish Stock of North

America. BAAS, 1900.

(1903): Ethnological Studies of the Mainland

Halkome'lem. a division of the Salish of British

Columbia. BAAS, 1902. i (1904): Ethnological Report on the Stseelis and i

Sk»aulits Tribes of the HalkSmeTem Division of

. the Salish of British Columbia. JRAI, 34, 311-376.

(1907): Report on the Ethnology of the South-Eastern

Tribes of Vancouver Island. British Columbia.

JRAI, 37, 306-374. 160.

Hoebel, E. Adamson (1939): Comanche and Hekandika Shoshone

Relationship Systems. AA, 41, 440-457.

______(1940): The Political Organization and Law-ways of

the Comanche Indians. AAA-M, 54.

Honigmann, John J. (1946): Ethnography and Acculturation of

the Fort Nelson Slave. Y, 33.

. (1949): Culture and Ethos of Kaska Society. Y, 40.

______(1953): Social Organization of the Attawapiskat

Cree. ANTH, 48, 809-816, Fribourg, Switzerland.

(1954): The Kaska Indians: An Ethnographic

Reconstruction. Y, 51

(1956): The Attawapiskat Swampy Cree: An

Ethnographic Reconstruction. Anthropological

Papers of the University of Alaska, 5,1, 23-82,

College, Alaska.

Jenness, Diamond (1935): The Ojibwa Indians of Parry Island:

Their Social and Religious Life. CNM-B, 78.

__ (1937): The Sekani Indians of British Columbia.

CNM-B, 84.

(1938): The Sarcee Indians of Alberta. CNM-B, 90.

(1960): The Indians of Canada. Fifth Edition,

CNM-B, 65. 161.

Kelly, Isabel T. (1932): Ethnography of the Surprise Valley

Paiute. UCPAAE, 31, 3, 67-210.

Krause, Aurel (1956): The Tlingit Indians. Translated by Erna

Gunther, American Ethnological Society, Seattle.

Kroeber, Alfred L. (1902): The Arapaho. AMNH-B, 18, 1.

_ (1908): Ethnology of the Gros Ventre. AMNH-AP,

1, 4.

(1963): Cultural and Natural Areas of Native

North America. Berkeley and Los Angeles.

Landes, Ruth (1937): Oiibwa Sociology. CU, 29.

Leacock, Eleanor (1954): The Montagnais Hunting Territory and

the Fur Trade. AAA-M, 78.

(1955): Matrilocality in a Simple HuntingEconomv.

Southwestern Journal of Anthropology,. 11, 31-47.

Leach, Edmund (1961a): Rethinking Anthropology. London.

(1961b): Rethinking Anthropology, in Leach, 1961a.

(1961c): The Structural Implications of Matri-

lateral Cross-Cousin Marriage, in Leach, 1961a.

(1964): Letter to the Editor. Comment on

Scheffler's Note on the Mangaian Kopu, AA, 66,

427-429.

Lips, Julius E. (1947): Naskapi Law. APS-T, 37, 4, 379-492. 162.

Lowie, Robert H. (1909): The Assiniboine. AMNH-AP, 4,1.

(1912): Social Life of the Crow Indians.

AMNH-AP, 9, 2.

(1924): Notes on Shoshonean Ethnography.

AMNH-AP, 20,3.

(1935): The Crow Indians. New York.

• (1948): Social Organization. New York.

MacNeish, J.H. (Editor)(1957): The Poole Field Letters.

AN, 4, 47-60.

Mandelbaum, David (1940): The Plains Cree. AMNH-AP, 37,2.

Mason, J. Alden (1946): Notes on the Indians of the Great

Slave Lake Area. Y, 34.

Masson, L. R. (Editor) (I960): Les bourgeois de la compagnie

du nord-ouest. 2 Volumes, New York.

Maclean, John (1895): Social Organization of the Blackfoot

Indians. Transactions, the Canadian Institute, 4,

249-260, 1892-1893, Toronto.

McAllister, J. Gilbert (1955): Kiowa-Apache Social Organization,

in Eggan, 1955b (q.v.).

McDonnell, John (1960): Some Account of the Red River (About

1797). in Masson, Volume I (q.v.).

Mcllwraith, T.F. (1948): The Bella Coola Indians. 2 Volumes,

Toronto. 163.

Mooney, James (1898): Calendar History of the Kiowa Indians.

BAE-R, 17, 1, 141-444, 1895-96.

(1907): The Cheyenne Indians. AAA-M, 1, 6,

(Lancaster, Pa.).

(1928): The Aboriginal Population of America North

Mexico. Smithsonian Miscellaneous Collections,

80, 7, Washington, D.C.

Murdock, George Peter (1934): Kinship and Social Behaviour

among the Haida. AA, 36, 355-385.

(1936): Our Primitive Contemporaries. New York.

(1960): Social Structure. New York.

Murdock, George Peter, Ford, C.S., Hudson, A.E., Kennedy, R.,

Simmons, L.W. and Whiting, J.W.M. (1945):

Outline of Cultural Material, quoted in Murdock,

1960 (q.v.).

Niblack, Albert P. (1890): The Coast Indians of Southern

Alaska and Northern British Columbia. USNM,

1887-88.

Oberg, Kalervo (1934): Crime and Punishment in Tlingit Society.

AA, 36, 145-156.

Opler, M.E. (1936): A Summary of Jicarilla Apache Culture.

AA, 38, 202-223.

. (1947): Rule and Practice in the Behaviour Betigeen Jicarilla Apache Affinal Relatives. AA,49,453-462. 164.

Ottenberg, Simon and Ottenberg, Phoebe (1960a): Africa and

Its Peoples, in Ottenberg and Ottenberg,

1960b (q.v.).

(Editors)(1960b): Cultures and Societies c£ Africa.

New York.

Park, Willard Z. (1937): Paviotso Polyandry. AA, 39, 366-368.

Provinse, John H. (1955): The Underlying Sanctions of Plains

Indian Culture, in Eggan, 1955b (q.v.).

Rogers, Edward S. (1962): The Round Lake Oiibwa. Occasional

Paper 5, Art and Archaeology Division, Royal

Ontario Museum, University of Toronto, Toronto.

Radcliffe-Brown, A.R. (1950): Introduction, in Radcliffe-Brown

and Forde (q.v.).

Radcliffe-Brown, A.R. and Forde, Daryll(Editors) (1950):

African Systems of Kinship and Marriage. London,

New York, Toronto.

Royal Anthropological Institute (1951): Notes and Queries on

Anthropology. Sixth Edition, London.

Sapir, E. (1915): The Social Organization of the West Coast

Tribes. Proceedings and Transactions, Royal

Society of Canada, Third Series, 9, 355-374, Ottawa. 165.

Skinner, Alanson (1914a): Notes on the Plains Cree. AA, 16,

68-87.

(1914b): Political Organization. Cults and

Ceremonies of the Plains Cree. AMNH-AP, 11, 6.

- (1911): Notes on the Eastern Cree and Northern

Saulteaux. AMNH-AP, 9, 1.

Speck, F.G. (1915): Family Hunting Territories and Social

Life of Various Algonkian Bands of the Ottawa

Valley. Memoir, Geological Survey of Canada, 70,

Ottawa.

(1917): The Social Structure of the Northern

Algonkian. Papers and Proceedings, American

Sociological Society, N.S. 12, Philadelphia.

______(1918): Kinship Terms and the Family Band among the

Northeastern Algonkian. AA, 20, 143-161.

______(1923): Mistassini Hunting Territories in the

Labrador Peninsula. AA, 25, 452-471.

(1927): Family Hunting Territories of the Lake St.

John Montagnais and Neighbouring Bands. ANTH, 22,

387-403, Vienna, Austria.

Speck, F. G. and Eiseley, Loren C. (1939): Significance of

Hunting TerritorySysterns of the Algonkian in

Social Theory. AA, 41, 269-280. 166.

Speck, F.G. and Eiseley, Loren C. (1942): Montagnais-Naskapj

Bands and Family Hunting Districts of the Central

and Southeastern Labrador Peninsula. APS-P, 85,

215-242.

Steward, Julian H.(1933): Ethnography of the Owens Valley

Paiute. UCPAAE, 33, 3, 233-350.

(1938): Basin-Plateau Aboriginal Sociopolotical

Groups. BAE-B, 120.

(1939): Some Observations on Shoshonean Distribu•

tions . AA, 41, 261-265.

(1941): Culture Element Distributions: XIII.

Nevada Shoshone. AR, 4, 2, 209-359.

(1943): Culture Element Distributions: XXIII.

Northern and Gosiute Shoshoni. AR, 8, 3, 263-392.

______(1955): Theory of Culture Change. Urbana, Illinois.

Stewart, Omer C. (1939): The Northern Paiute Bands. AR, 2, 3,

127 - 149. (1941): Culture Element Distributions: XIV.

Northern Paiute. AR, 4, 3, 361-445.

______(1942): Culture Element Distributions; XVIII.

Ute - Southern Paiute. AR, 6, 4, 231-356. 167.

Strong, William Duncan (1929): Cross Cousin Marriage and the

Culture of the Northeastern Algonkian. AA, 31,

277-288.

Swanton, John R. (1905): Contributions to the Ethnology of the

Haida. AMNH-M, 5, 1, Jesup North Pacific

Expedition.

(1908): Social Condition. Beliefs, and Linguistic

Relationships of theTlingit Indians. BAE-R, 26,

391-512, 1904-05.

Teit, James Alexander (1906): Notes on the Tahltan Indians of

British Columbia, in Boas Anniversary Volume,

pp. 337-349, New York.

(1956): Field Notes on the Tahltan and Kaska

Indians: 1912-1915. AN, 3, 39-171.

Wedel, Waldo R. (1953): Some Aspects of Human Ecology in the

Central Plains. AA, 55, 499-514.

White, C. Langdon, and Foscue, Edwin J. (1954): Regional

Geography of Anglo-America. Second Edition,

Englewood Cliffs, New Jersey.

Wilson, Edward F. (1888): Report on the Blackfoot Tribes.

BAAS, 1887. 168.

Wilson, Edward F. (1889): The Sarcee Indians. BAAS, 1888.

Wissler, Clark (1911): Social Life of the Blackfoot Indians.

AMNH-AP, 7, 1.