OIIBWA EMOTIONAL RESTRAINT MID THE
SOCIALIZATIGN PROCESS

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
THOMAS H. HAY
1968

 

IIIIIIIIIIIIIIIIIIIII

This is to certify that the

thesis entitled

OJIBWA EMOTIONAL RESTRAINT AND

THE SOCIALIZATION PROCESS? -., : '

presented by _ .
Thomas H. Hay

has been accepted towards fulfillment
of the requirements for

‘ M degree in Mr I ‘13 9

. _ . 7 a
Major professor: s

I

Date It: May 1‘16?

0-169

tree-J‘s. aux-H“ - . _;4 f

, . r 2L!
L [B 13 :1 It 1

. . I

lVlICIllgf r1 Smte I
Univcrsity

Tm “nu—r

 

l
I PLACE IN RETURN BOXto remove this checkout from your record.
TO AVOID FINE return on or before date due.

MAY BE RECALLED with earlier due date if requested.

 

 

 

 

 

 

 

DATE DUE DATE DUE DATE DUE
O l “ IL
.mfin 9 2831 ”SW l 260%
APR 2 3 2:13
64 2 5 r C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m www.mu

 

 

 

 

 

 

 

 

 

ABSTRACT

OJIBWA EMOTIONAL RLSTRAINT
AND THE SOCIALIZATION PROCESS

by Thomas H. Hey

This study focuses on the motivation of the
extreme restraint characteristic of the typical
Ojibwa Indian and on the learning process through
which motivation for this restraint is developed.
Hypotheses derived from two different theories of the
motivation of the restraint and from two different
theories of the development of extreme restraint are
tested in this study.

One of the theories of the motivation of
restraint traces this restraint to an exaggerated fear
of retaliation for openly aggressive behavior. The
other theory traces restraint to an exaggerated
fear of doing serious injury to others by openly
aggressive behavior. The first of the theories of
the development of restraint, derived from stimulus-
reSponse learning theory, attributes fear of retal-
iation for aggressive behavior to frequent punishment
in childhood for behavior offensive to adults. The

other theory, derived from cognitive learning theory,

attributes exaggerated fear of doing serious injury
to others to adult acquiescence in the offensive
behavior of children.

Since the Ojibwa consider children incapable
of effective retaliation against adults, but do not
consider children tobe immune from injury by adults,
it was reasoned that these theories of the motivation
of restraint could be tested through observation of
adult reactions to the offensive behavior of children.
The observations could also be used to test the
learning theories when analyzed from the point of view
of the experience of the child.

The two motivational theories were partially
formalized in the expected value framework employed
by some contemporary decision theorists. This formal-
ization permitted deduction, fromezach theory, of
quantitative estimates of the probable choices by
adult Ojibwa of four different courses of action in
a situation where the adult had been offended by a
child. These formalizations also provided estimates
of the frequencies of punishment or of acquiescence
required by the learning theories.

The data to test the hypotheses derived from
these theories was gathered through observations of
the interaction of adults and children in nine Ojibwa

households at Lac du Flambeau, Wisconsin and four

Ojibwa households at Berens River, Manitoba. These
households can not be taken to be representative of
file populations at either place. The sample is
definitely biased toward the more culturally conser-
vative families. The number of observations recorded
was sufficient to permit tests (with the Kolmogorov-
Smirnov test of goodness of fit) for twenty-nine
adults of the hypothesis derived from the motivational
theories. The binomial test of goodness of fit was
used to test the learning hypotheses for thirty—two
children.

The theory that Ojibwa restraint is motivated
by fear of retaliation is not supported by the data.
It implies that Ojibwa adults should punish offending
children with high relative frequency. This hypothesis
could be rejected at the .01 level for twenty-seven
of twenty-nine adults.

The theory that Ojibwa restraint is motivated
by fear of doing serious injury to others is neither
clearly supported nor contradicted by the data. There
is, however, reason to believe that minor modifications
of this theory would lead to deduction consistent with
the data.

The theory that fear of retaliation is learned
through punishment in childhood is unsupported by the

data -- as might be eXpected from the results for the

theory that restraint is motivated by fear of retal—
iation.

Finally, the theory that exaggerated fear of
injuring others is due to parental acquiescence in
the offensive behavior of children can be rejected for

only two of thirty-two children.

OJIBWA EMOTIONAL RESTRAINT

AND THE SOCIALIZATION PROCESS

by

Thomas H. Hay

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
' for the degree of

DOCTOR OF PHILOSOPHY

Department of Anthropology

1968

0/!
4/

r 2
I
I'/

ACKNOW LEDG EtENT S

Like most dissertations, this one owes much
to the help and encouragement of many others. I
particularly wish to express my appreciation to the
following people.

To my committee, Professors Charles Hughes,
Seymour Parker, Lucy R. Ferguson, and Moreau Maxwell,
whose encouragement and helpful suggestions have been
indiSpensable.

To Professor Harry Grater, now of the Univer—
sity of Florida, without whose help I would not have
been able to write a dissertation at all.

To Professor Stephen Boggs of the University
of Hawaii for many helpful discussions in the form-
ative days of the project and much encouragement since
then.

To Professor Melford E. Spiro of the Univer—
sity of Chicago and to Professor John Messenger of
Inidana University for their long-continued encourage-
ment and support.

And, finally, to my wife, Nancy Harden Hay

who gave more than anyone.

TABLE OF CONTENT

Chapter
I. INIROD JCTION.......................

11. 11-42:: Two comm 711117 1: s.
111. 17:11: FARIILIES 3111:3121)

TV. THEORIES OP TIE T"O'I‘IVIII'I‘IOI'I AND

DSVELOPHENT OF OJIBKA RESTRAI"T. .

V. RESULTSOO0.000000000000000.000.0000

v:. 17:11", 71: me o 171:: 1.1.7317 AND T1113 "PEA? or

7": II T rr‘rv‘nr

.‘ ‘r'r
C77.
Q14)“; Luql'tlooooooooooo00.000000.

VII. CONCLUSIONS AND I'IPLICATIOIIS.......

ppBIIDI-J\oooooooooooooo00000000000000.0000.

hhh~§rfilrc S

Lexi—4. CITEDoooo00000000000000.0000...

Ho
Ho
Ho

12
37

61!

i‘

110

161

p...
O‘

H
\ i)
[\J

P\)
I!)
\l

U1
0
IO

U1

U1
CD

\51
\3

LIST OF TABLES

Sample Application of K-S Test to Ten
Incidents of adult-Child Interaction
Under a Hyoothesis From the “Fear of

C’thCI‘S" Theoryooooooo00.00000000000000000000

Distribution of Observed Responses of
Twenty—nine Ojibwa Aaults Compared
with the Theoretical distribution from

the "Fear of Others” Theory.................

Observed RCSponses of Adults to Offending
Children and Test of Hypothesis from

"Fear of thers” Theory for each of
Twenty-nine AdUltSoooooooooooooooooooooooooo

Distribution of Observed Responses of
Twenty-seven Ojibwa Adults Compared with
Theoretical Distribution from ”Fear of

C: :‘I “ ..
uCli IRCOL‘a...co.00.000000000000000coco...

Observed Responses of Adults to Offending
Children and Test of Hypothesis from

"Fear of Self” Theory for each of Twenty-
SCVCH AdUltSoooooooooooo00000000000000.00.00

Comparison of Observed Distribution of
Choices of Berens River Agults with Mod-
ified Theoretical Distribution..............

Comparison of Observed Distribution of
Choices of Lac du Flambeau Adults with
Modified Theoretical Distribution...........

Observed ReSponses of Adults of Offending
Children and Test of Hypothesis from
”1"

Berens River modification of rear of
Self" Theory for Each of Seven.Adults.......

Observed RCSponses of Adults to Offending
Children and Test of Hypothesis from Lac

Q

du Plambcau Kodification of ”Fear of

83 f” Theory....o.....oo..oo................
Outcomes of 767 Situations in which Children
Behaved in a Manner Offensive to Adults.....

1V

Page

103

Ill

:13

IBM

J

135

H
K»
C)

l 1:3

Table

U1
f—J
(j)

.16

VT

List of Tables -- cont nued

 

Outcomes of Situations in which a Child
Behaves in a Manner Offensive to Adults.
Children Under Two Years of Age Omit-

tGdOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO...

Outcomes of Situations in which a

Child Behaves in a Iann r Offensive to
Adults. Comparison of all Observations

at Berens River with all Observations

at Lac du Flambeau.........................

G1oup I Learning Theory. Number of Of—
1ensive Acts, Number of Punitive ReSpon-
ses Received, Relative Frequency of
Punishment, and Binomial Probability

of Observed Number of Punitive Responses
under the Null Hypothesis, for each of

Thirty-LIVO Ct]. ildr‘enooo00.000000000000000...

Outcomes a: sit ations in which a Child
Behaves in a _r Offe ensive to Adults.
All Observations at Lac du Plambeau and

Bercns Riverooooooo0000.00.00.00...0000-000

Eic
o
3

Outcomes of

in which a Child
Behaves in a Manner Offen
s

ive to Adults.

1‘

Outcomes 01 oi ations in which a Child
Behaves in a anner ffensive to Adults.
Comparison of all Observations at Berens
River with all Observations at Lac du

FlambcauOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Outcomes of Situations in which a Child
Behaves in a 1.1anner Offensive to Adults.
Comparison of Children Two Yea ars Old

and Old oer at Lac du Flambeau with

Similar Children at Berens River...........

Outcom s of Situations in which a Child
Behaves in a Lionner Offensive to Adults.
Children less Than Two Years Old:

Berens River and Lac du Plambeau com-

parcd..............................o..ooooo

W

s
of Age Cmmitted...

ILL

‘ J
ILLb
1&6

5—;
V1
H

H
U1
OJ

1....
U1
UL

a
l“)
U.
H
C)

8.3

C7)
0

Ho

U)
(-1'

 

Group II Learning Theory.
Offensive Acts, Number of Acquiescent

Resoonses Received, Relative
of Acquiescent and Binomial Probability

l-lumb e r o f

T“

:requency

of Observed Number of Acquiescent

Resoonses Under the Null Hypothesis, for
:aCh Of 32 ChilerHOOOOOOOOOO00......O...

Assumed Subjective Probability of
Occurrence of

each of
Outcomes Given Choice of

Of ACtiOHOOOO0.0...OOOOOOOOOOOOOOOOOOOOOO

the P

unitive
each Course

Subjective Probabilities of Primi-
tive Outcomes Given Choice of C1, C2,

C3, or Ch.

Probabilities of o3......................

”Fear of Others” theory,
Near Perfect Maximization, Minimum

Subjective Probabilities of Each of
the Outcomes, Given choice of C2 as
Originally Estimated and as Modified
for each Locality. “"Fear of Self"
Theory; Adults in Interaction with

ChilerHOOOOOOOOOOOOOOOOOQOOOOOOCOOOOOOOO

Probabilities of Choice of Each of

the

Courses of Action, as Originally Cal-

culated
cality.
in

and

OS

lbd

. H.

12].

ed

101‘

each lo-

"Fear of Self" Theory, Adults

Interaction with Children.............

Pace
————M—

157

206

CO

21

LIST OP ILLUSTRATIONS

5.1 Distribution of Observed Resoonses
of Adults to Offending Children
Compared with the Theoretical Dis-
tributions from the "Fear of Others"

Theory and the "Fear of Self" Theory....... 1

Pa)
\p

Apnendix Page

I Pormalization of the Two Theories
Of ”Otivationooooooooooooo0.000000000000000 192

CHAPTER I

INTRODUCTION

Pur,ose

The purpose of this project is to test some
contrasting hypotheses about the functioning and develop-
ment of the typical personality of the Ojibwa Indians --
a typical personality which is characterized by hos-

1

Ho
Ho
ri-

t y and emotional restraint. This type of personality

H 0
U)

of interest to both psychiatry and anthropology,
since there is evidence that it is a factor in the
development of psychosis and of some psychosomatic
disorders and because there are theoretical reasons
for believing that it may tend to perpetuate the
"atomistic" social organization of the Ojibwa.
iypotheses derived from four theories will be
tested. Two of these theories offer eXplanations of
emotional restraint in terms of underlying values and
beliefs. The other two offer explanations of the way
in which this emotional restraint is learned.
Hallowell has proposed a theory which explains
Ojibwa restraint as a consequence of their fear of
retaliation through sorcery. An alternative theory which
is implicit in the writings of some psychoanalysts ex-

plains this restraint as a consequence of the Ojibwa's

l

2
fear of what they might do to other people if they lost
control of themselves. From each of these theories a
hypothesis has been deduced concerning the relative
frequencies with which Ojibwa adults should choose
punitive behavior toward children who offend them.
These hypotheses will be tested with the same data used
to test the developmental hypotheses described below.
Overview

rrom two groups of theories of learning of
personality characteristics which differ in the em-
phasis placed on the child's drawing of inferences about
his social environment, are derived two kinds of hypo-
these about the child rearing antecedents of emotional
restraint and hostility. The hypothesis from one group
of theories implies that the antecedent of emotional
restraint is frequent and severe punishment. The hypo-
thesis from the other group of theories traces emotional
restraint to the absence of adult control.

From these hypotheses are derived measures of
parental punitiveness and acquiescence in situations
which are comparable in different households. Theo-
retical expected frequencies of parental choice of
punitive and acquiescent behavior are derived to permit
statistical tests of the hypotheses from which these
expected frequencies were derived. The data to test

these hypotheses were obtained through participant

3

observation of interaction in several households in

each of two differently acculturated groups of Ojibwa.
Interaction of adults with children was observed in a
total of thirteen households. Nine of these households
were located at Lac du Plambeau, Wisconsin and four at
Berens River, Manitoba. Six hundred seventy-one obser-
vations of adult reactions to offending children were
recorded, permitting statistical tests of hypotheses
about the behavior of twenty-nine adults. Twenty-two

of these adults were from Lac du Flambeau and seven

from Berens River. Seven hundred sixty—seven observations
of the adult reactions eXperienced by offending children
were recorded. These were sufficient to permit statis-
tical tests of the hypotheses about these experiences
for thirtyhtwo children. Twenty-four of these children
were observed at Lac du Flambeau and eight at Berens

River.

Ojibwa typical pers nality

Ojibwa personality has been studied by numerous
anthropologists, but the work of A.I. Hallowell has been
particularly extensive and thorough and will serve as
the principal background for this study. He has studied
three groups of Ojibwa, each at a different ”stage of

acculturation." Two of these groups are located in

Manitoba and one in Wisconsin. Hallowell calls the

group at the earliest stage of acculturation the

” and the group at an intermediate stage

”Inland Group,
of acculturation the "Lakeside Group." The third,
and most acculturated, group is located at Lac du
Flambeau, Wisconsin and gets its designation from that
location. On the basis of a sample of Rorschach tests

from each group (uh, 58, and 115 subjects, reSpectively),

he concludes:

The most striking fact is the persistence of the
same basic psychological pattern through t-ese
stages of acculturation. There is a persistent
core of generic traits which can be identi“ird
as Ojibwa (1955:351)

Hallowell does not indicate in this article which traits
constitute this "core," but his other works indicate
that he considers hostility and emotional restraint to
be the central characteristics of the Ojibwa typical

personality. In Chapter Six of Culture and bxperi ence

 

(1955) in which he attempts to reconstruct the person-
ality typical of the early northeastern Indians from the
accounts of missionaries, eXplorers, ani traders, the
major characteristics discussed are emotional restraintand
hostility. In the same chapter, referring to the

Rorschachs of the two Canadian groups, he says:

q
/
The most prominent feature in the great majority

of the records is the emphasis on strong restraint
and control (1955:1h9).

Other students of Ojibwa personality concur in
the centrality of emotional restrain and hostility as
traits of the Ojibwa typical personality (Landes in a
personal communication to Victor Barnouw, quoted in
Bernouw 1963:1h7; Barnouw_1950:22, 60—61; ames 1961:738,
7h0; Wallace 1952:10h—IOS; Lowenfels, quoted in Barnouw
1963:282). Lowenfels, in her analysis of Ojibwa child-
ren's drawings, finds 63% "constricted" and "over-
controlled" and the remaining 37% also "manifesting

Y

constriction and over-control' but less "inflexible"

(Barnouw 1963:282). These reports are taken to justify
the assumption that a large majority of the Ojibwa
disolay hostility ani emotional restraint which

Hallowell appears to regard as the "core" of the typical

personality.

Psycholocical and secial significance of this personality
We

(—1—

 

Personalities marked by this combination of
hostility and restraint are of considerable interest
to psychiatry since there is evidence linking this type
of personality to the development of some psychoses

and to the development of essential hypertension and

6

neurodermatitis. (See Buss 1961, chapters 11 and 12
for a summary of the evidence.)

Since personality plays an important part in
some theories of cultural integration (Kardiner 1939:
20-21; whiting and Child 1953:310) and in some theories
of cultural change (Linton 1939zx; Wallace 1951:61;
NcClelland l961:ch2; Hagen l962:ch 8), the process
of personality development is also of concern
theories. These theoriesof cultural change amp
that the typical personality of the members 0. a society
tends to limit the society's acceptance of cultural
changes to those which are congenial to the typical
personality. The "social atomism" (Barnouw 1950:15—16)
or "particularism" (Hickerson 1962: ch 1) of Ojibwa
society can be explained on this basis through a hypo-
thesis proposed by Berkowitz, building on earlier work
by Newcomb (19h ) and that by Thibault and Coules (1952).
Berhowitz hypothesizes that people who are hostile and

also inhibited about eXpressing aggression

...sought to cut off communications with a person
who, they believed, might heighten their aggressive
inclinations and therefore, would also increase
their anxiety about this hostility (1962:99).

If considerable communication is necessary for the

7

effective organization of groups larger than the
extended family household, then the Ojibwa reluctance
to form larger social groupings would be expected from
this hypothesis.

It seems, therefore, that information about
potentially significant factors in the functioning and
development of this type of personality is relevant
to the concerns of psychiatry, psychology, and anthro-

pology.

Hypotheses to be tested

 

The fundamental assumption of this study is
that a large majority of contemporary Ojibwa children
are in the process of developing the hostility and
emotional restraint which is regarded as the “core"
of the typical personality.

The hypotheses to be tested are derived from
two groups of theories which embody different concep-
tions of the learning process. The conception of the
learning process in the first of these groups of theories
includes only a minimum use of reasoning on the part of
the child who is learning. This kind of theory is often
labeled "stimulus-reSponse" or "8-?" theory. Hereafter,
these will be called "Group I Theories. " The conception
of the learning process embodied in the second group of

theories includes much more use of reasoning by the child

8

who is learning. These theories are often called

"cognitive theories." These will be called "Group II

.

theories." As used here, the child is said to be

reasoning if he draws inferences about relationships

 

among events, real and hypothetical. Reasoning is, of
course, inferred from more readily observable behavior.
This difference between the two groups of theories is
important since, for high degrees of hostility and
enotional restraint, the antecedents predicted by the
Group I theories are opposite to those predicted by
the Group II theories.

All of the hypotheses entertained in the course
of the original field work were drawn from Group I
theories. bince no alternative to this group of theories
had been considered, it was not until after the original
work had been completed that the possible inconsistency
of theciata with the implications of these theories
was fully appreciated and a search was begun for
theoretical formulations in which personality character-
istics were conceived to be differently related to the
actual events experienced by the child.

After such an alternative was found, the original
study was recast as an effort to test hypotheses derived
from the two different groups of theories. The theories

being tested are thus primarily explanatory of what has

 

9

been observed rather than predictive, although their

 

specific forms and implications were developed before
the date were analyzed formally. It should be noted
here that the formal, logical structure of explanation
and prediction are identical. That is, a formally

adequate, deductive explanation is sufficient for

 

prediction (Hempel 19h2). Partial formalizations of
the two theories of Ojibwa restraint are to be found
in the ap endix.

Hypotheses derived from the two theories
purporting to explain Ojibwa restraint concern the
relative frequencies with which individual adults
should be observed to choose Specified courses of action
when they have been offended by children. hither
theory (or both theories) could be rejected if the
deduced frequencies of choice were not observed for a
sufficient proportion of the adults in each of the two
groups studied. There are thus two statistical prob-
lems. First, the determination for each adult of
whether or not the observed relative frequencies of
choice of the Specified courses of action could be
eXpected if the hypothesis derived from a particular

theory were true and, second, whether or not a suf-

ficient prppprtion of the observed adults of both

 

groups diSplay the frequencies of choice implied by

10
the particular theory.

The learning theories are concerned with the
relative frequencies with which each child received
certain kinds of treatment from adults. The statis-
tical problems are essentially similar to those in—

volved in testing the theories of restraint.

The sample
AS is frequently the case in child-rearing

"

studieS, selection at a probab

I—‘o

lity sam le was impos-

'd

Sible. Therefore, any conclusions can be generalized
only with caution. In testing hypotheses about the
typical personality of the Ojibwa and its transmission,
the important issue is not the extent to which the
sample represents the population of Lac du Flambeau or
Berens River, but the extent to which the people studied
can be taken to be typical Ojibwa. An effort has
therefore been made to estimate the extent to which the
families which were studied can be taken to be typical.
(See Chapter Three.)

The families which were studied were from the
“Lakeside Group” at Berens River, Manitoba and from the
Lac du Flambeau Reservation in Wisconsin. The differences
between these two groups provide some control over
the possible significance of Specific child-care

techniques -- which is not entirely a dead issue

ll
deSpite Orlansky's conclusions.
Orlansky concluded, "...we are led to reject the
thesis that Specific nursing disciplines have a Specific,
invariant psychological impact upon the child" (19h9:39).

"specific nurs1no dis-

The differences between these
ciplines” at the two locations are not clear-cut,

but there are differences in the relative frequencies of
different "disciplines" between Lac du Flambeau and
Berens River.

A few families at Berens River use the cradle-
board for infants and some use the cradle-board wrap-
pings without the board itself. At Flambeau neither
cradle-board nor wrappings were observed, except on
diSplay in business establishments. At Berens the
hammock or swing (WEwEbizon) in which infants sleep is
nearly universal; at Flambeau, it is used by a very
few families. At Berens River about two out of three
infants are breast-fed (based on a limited sample),
while, at Lac du Flambeau, all infants known to me
were bottle-fed. Boggs (1958:50—53) has discussed
these and some other differences between his two samples
from the same Indian groups.

The data to test the hypotheses consist of nar-
rative accounts of family interaction gathered by me and

my wife during participation in everyday activities with

the sample families.

CHAPTER II

THE TWO COMMUNITIES

Introduction

In each of the two communities in which field
work was done there is a considerable range of variation
in the degree to which individual and family behavior
deviates from the reported traditional norms. In
most reSpects, the deviations from traditional behavior
are smaller at Berens River than at Lac du Flambeau.
For this reason, I shall consider the community of
Berens River and the behavior of its people in com-
parison to reported traditional behavior first anl then
the community of Lac du Flambeau and the behavior of

its people.

Berens River, Manitoba
fihysical Setting

The community of Berens River is located at

 

the mouth of the Berens River on the east side of Lake

Winnipeg about half-way between the northern and southern

end of the lake. The land is low and swampy with fre-

quent outcroppings of granite bed-rock. The swamps

I

are muSkig, that is, they are thickly grown with trees,

12

13

rather than open. The characteristic forest cover is
Spruce and birch with some balsam. The reserve extends
inland on both Sides of the river and north from the
river along the rocky shore of the lake. There are

no roads or railroads nearer than 100 miles and access

to the "outside" is primarily through the weekly steam-
ship when the lake is free of ice and by plane or

tractor train. During the Fall freeze and Spring "break-
up", there is only radio contact with "the outside".

The log houses of the Indians are 100 to 300
yards apart along river and lake shore in the reserve.
(In Canada, a "reserve" is primarily an area for housing
and the Indians have the right to hunt and trap over
the area outside the reserve.) In the summer commercial
fishing season, some of the families from up-river

live in tents along the lake shore near their relatives.

Pqpulation

 

The 19h8 annuity list, supplied by the Hudson's
Bay Company factor, with some correction for births
and deaths by him, and some by myself, lists 391
enrolled members of the band. In addition to the
Indians, there are about 35 whites, mostly the families
of government officials, traders, and missionaries,

and something less than 100 "half-breeds." These last

1h
are Indian in identification and style of life.
They differ from technical "Indians" in not being

eligible for annuity payments, and in being eligible

to vote in Canadian elections and to buy beer.

Subsistence

 

Aboriginally, the Ojibwa were hunters and
gatherers. With the arrival of Europeans and the fur
trade, this basic occupation was not disrupted but
intensified by the new emphasis on fur-bearers (Kroeber
l9h7z96; Hallowell 1955:119).

In 1952 hunting and trapping had declined in
importance, but were still Significant in the subsis-
tence of many families. The decline in importance
is indicated by the increased importance of four other
sources of income.

Chief among these is commercial fishing. Most
of the families at Berens River were engaged in this
pursuit in the summer of 1952. As in hunting and in
trapping for the fur trade, a man is still his own boss
in commercial fishing. A man gets his nets, the rental
of a boat and motor, and the purchase of some goods at
the fish company store on credit which is paid off by
his catch. There are three seasons when commercial

fishing is open in different areas of Lake Winnipeg.

15
There is a Spring and early summer season at Berens River;
an early Fall season in areas toxvhich many families move
for the season; and a winter season, in which few
people at Berens River are involved.

Cutting fire wood for sale to the steamship
company is a winter occupation which engaged many men
until the winter of 1952, after which the ship was con—
verted to another fuel.

rs
..
-—.

mace work for th wish companies, the schools

('3

and the United Church, and the forestry ‘epartment em-
ployed a few men seasonally, while the Haunted Police
and the Hudson's Bay Company employed one each the year
around.

The fourth source of subsistence is government
payments, usually in the form of rations 01

to those who are in need.

A few families gather wild rice for the com-

I

Ho
('9‘

mercial market, but is not abundant locally.
The decline in the importance of the fur trade

decline in fur prices

0
"1
O.)
"J
H.
D.

has been accelerated by tn
in the early 19508. Few amilies in Berens River move

to their winter trapping grounds. As a rule, those men

still engaged in trapping go out to their trapping
grounds periodically, leaving their families in the

settlement. Hallowell (1955:119) reports this pattern in

the 19308 when morenen trapped. Information supplied

16

by Dr. Nufeld of the United Church mission indicates
that about half of the men, assisted by their older
sons, were out trapping in the winter of 1951-52.

The continuing importance of hunting and trapping
plus fishing is indicated by the unavailability of
animal foods at the stores. In addition, the standard
footwear of the Indians is moosehide moccasins over
which are worn rubbers purchased at the store -- a
great boon in wet country. A surplus of moccasins is
produced and sold to the stores for tourists. DeSpite

the summer emphasis on fishing, several moose were shot

and the meat smoked and hides tanned while we were there.

Social Organization

 

Marriage

Landes' description of marriage at Manitou
in Ontario (1937:103—105) largely applies to Berens
River. It seems unlikely, however, that the disappear-
ance of cross-cousin marriage can be attributed to
contact with the Dakota, but must be attributed to other
causes which certainly include Christianization and
probably the development of relatively large and per-
manent settlements. Boggs (lQSh: ChZB) notes in a few
families a tendency for second and third cousins to

marry, which could be a persistence of the old pattern

17
modified to suit the Christian tabu on first-cousin

marriage. Landes' generalization that "

...most marriages
are fairly short and very stormy," (1937:10h) applies
to Berens River as well as to Hanitou.

Adultery and illegitimacy are fairly frequent,
but only sometimes lead to separation of a couple.
The illegitimate child of a married woman is often
knowingly raised by her husband. In the case of an
unmarried girl, the illegitimate child is often brought
up in the household of her parents until she marries
when it will go with its mother to her husband's house-
hold. (This information is derived from a household
survey made by Dr. Nufeld of the United Church mission
dated January 1952 and from field notes.) There seems

to be little stigma attached to either the mother of

llegitimate child or to the child.

5—»

an

The Household

According to Dr. Nufeld's survey, forty-one
of the seventy-one households of enrolled Indians were
nuclear family households, eighteen included three
generations, three households were composed of one or
both grandparents anl grandchildren, three households
included some relative other th>n nuclear family and
grandparents, and five men and on“ wo.er lived alone.

”923 nouseholds are not stable in membership as people

Q
L)

rd

shift their living arrangements readily as is indicated

T“.

by the shifts of Spouses mentioned before. bones

\JV

(lQSh: Ch 2) concludes that family and household sta-

bil ty cannot be shown to have decreased as a result of

No

contact.

The Extended Family
The extended famil

unit at Berens Ri er. As indicated above, many house-

holds includethree generations. hdditionally, it is a
common practice for men to build their houses near those
of their fathers or fathers-in-law. Tishing, trapping,

1‘
i

and cutting wood or the steamer are carried on by groups
of men related in either of these ways and sometinng

by brothers after the death of their father.

Community Organization

As of 1952, government was nominally vested
in a hereditary chief and a group of councellors, but
they lacked effective powers to initiate action or

maintain social control. Social control, as in ab—

3

nal times, is still primarily a matter of in

O
’1
Ho
(13
[-40

d
ual, internalized control. Hallowell writes of the

Berens River people:

19

ince there are no organized penal sanctions and,
n fact, no adult has any authority to impose any
sort of penalty upon another person, and since
openly expressed moral disapproval is avoided
because it may be taken as a sign of hostility,
fear of disease is the major social sanction

(1955:29k).

S
O
1

These internal controls are quite effective; crimes are
very rare. External controlis maintained by a single
officer of the Royal Canadian Nounted Police who is
resident in the community, but whose territory must
cover more than 5,000 square miles. The justice of
the peace for the area lives fifty miles down the lake --
a six-hour round trip in the police boat.

Trapping grounds are registered with, and super-
vised by, the provincial Game Guardian.

No organized community activity occurs. Coop?
erative group activity on the scale of the extended
family does occur, but no cooperative activity on a

larger scale.

Religious Behavior

Virtually all of the Indians in the community
are Christians and most attend services with considerable
regularity. They are about evenly divided between
the United Church and the ioman Catholic Church.

Observances of Ojibwa religious ceremonies

20
have long been forbidden in the community and apparently
do not take place -- certainly none were observed.
Hallowell (1936:h9) concludes for this community,
"The last lidewiwin must have taken place over lifty
years ago." Aboriginal herbal remedies are regularly

used, and Boggs reports the use of "love potions,"

L-

as well (1 951

:35).

De5pite the departure from traditional Ojibwa
religious observances, there is a strong religious
orientation at Berens River. Although many people
express disapproval of the traditional religion, they
also evince considerable interest in it (Boggs lQSh:

Ch 2). Boggs discusses religious behavior at some length
and concludes that, although Christi anized, the

Berens River Ojibwa are closer to the aborigi1nal

pattern than the people of Lac du Flambeau who are

mostly secular in orientation, although a few people

at Lac du Flambeau still observe traditional rites.

Lac du Flambeau, Wisconsin

Flambeau in 1052

 

The following account applies to Lac du
Flambeau as it was seen in 1952. A few important changes

in subsistence, political organization, and religious

 

21
behavior were noted in 1966. These will be reported in

a concluding section.

Physical Setting

The village of Lac du Flambeau Sprawls among
tall white pines on the high ground which separates
the clear, cool lakes which form the Flambeau chain.
A dozen stores and the post—office are concentrated on
two blocks of a single street. The houses of the
Tndians Spread out from this center among the pines and
along the shores of the lakes.

Two mi-es to the west of the village proper, the
"Old Village" stretches along the north shore of
Flambeau Lake. It is here in the Old Village that a
few people still build wigwams for summer use and the

Ojibwa religious ceremonies are generally held, al-

Ho

though many of the adherents of the Drum and Midewiw n
live elsewhere on the reservation.

The Lac du Flambeau reservation is located in
extreme northern Wisconsin. It is a rectangular tract
of land about ten miles from east to west and about 12
miles from north to south.. It contains
and many smaller ones. Much of the area which is not
lake is swamp and the remainder is sandy soil on which

grow some hardwoods (including sugar maple), considerable

pine, and a large acreage of aSpen, or "popple."

Acculturative influences

 

Until the advent of a sawmill and the railroads
in 1893, Lac du Flambeau was an isolated reservation
with little contact with.white Americans. Since then it
has been increasingly drawn into the mainstream of
American life. There is still some logging of the pine
on the reservation anl the "popple" provides winter
employment for some men "cutting pulp" for the paper
industry to the south, but the major factors tending
to draw the Indians into the mainstream are the tourist
industry and a small factory.

The tourist industry got itS‘ start in the 19303
and in 1952 there were 26 resorts on the reservation.
These employ some Indians, both men and women, during
the tourist season -— June through August. The most
important effect of this business has probably been to
bring many permanent residents into the area with
consequent intensification of acculturative pressures.

Since lQhé, a small manufacturing company has
Operated a plant in what was once the Indian school in
the village. About 100 Indians, most of whom were
women, were employed there in 1952.

In the late lQhOs, the Indian school was closed
and the Indian children now attend the public school
with the white children of the area. (Boggs lQSh:

Ch 2) gives and excellent brief summary of Flambeau's

23
history and Hallowell (1955: Ch 18) covers some of the

earlier history of the Indian group who eventually

settled at Lac du Flambeau.

Polulation

 

There were 870 people on the tribal roll, as

Ho

corrected, in 1952. Th 3 figure includes resident
Ojibwa enrolled elsewhere and some additional changed
noted by local Indian service personnel and by me. This
list can be considered only approximately correct and
certainly includes some non-residents. Hallowell
estimated the number of resident Ojibwa in the summer

of lghé to be at least 831 (1955:3LO). The total
population of the school district in 1952 was estimated
by the principal to be 1,500, so that the year-around

residents of the reservation are about equally divided

between whites and Indians.

Subsisten e

 

Primary sources of income at Lac du Flambeau are
wage work for white employers and government payments
such as old age assistance, aid to dependent children,
public welfare, etc. Before the opening of the factory,
it was not unusual to have more than 50% of the families
on the reservation dependent on such payments for at

least part of the year. Since the Opening of the factory,

2h
this percentage has declined, but such payments are
still important to the existence of many families.
Ritzenthaler (1953) has described the efiects of the
opening of the factory on the community, perhaps,
somewhat optimistically.

Hunting and trapping have long been totally
inadequate as means of subsistence, but they do provide
supplementary food and emergency food occasionally.
Fishing is somewhat more important than hunting as a
means of obtaining food and provides some men with
considerable income in the summer through guiding
tourists. Successful guides also enjoy considerable
prestige among both Indians and white sportsmen.

These former subsistence activities are primarily
important as recreation.

A few families gather wild rice, two families make
maple sugar, and three or four tan hides and make moc-
casins. The motivation for these activities seems to

be largely sentimental attachment to the old ways.

Social Organization

*4.

marriage

The attenuated forms of cross-cousin marriage
found at Berens River do not exist at Lac du Flambeau.
Essentially similar reports to those from Berens

River can be given about the stability of marriage and

25
about attitudes toward infidelity and illegitmacy at
Lac du Flambeau. Marriages may be even less stable at
Lac du Flambeau than at Berens River, but this impression
may merely be due to longer residence at Lac du
Flambeau. In 1952, the county welfare department was
attempting to increase the stability of marriage through
forcing people into legal marriages by denying public
assistance to couples who were not married.

An important consequence of the greater avail-
ability of employment for women is the blurring of
distinctions in the roles of men and women. There
were several instances where the husbands of employed
women were essentially taking the role of housewife
during the working hours of their wives. They did not

perform this role with either enthusiasm or skill.

The Household

The household is the most important social unit
at Lac du Flambeau, as it arparently has been tradition-
ally for the Ojibwa. Hy notations on the tribal roll,
made as I came upon the information, indicate that a
minimum of thirty-six of the approximately 210 house-
holds were not nuclear family households; that is,
they contained three generations, were of the grand-

parent-grandchild type, or contained other relatives.

This is almost certainly a low estimate. Boggs reports

26
(l95h:h0) that such households constituted 30 percent

of a sample of M2 households best known to him.

The Extended Family

The extended family, as seen at Berens River,
is almost unknown at Lac du Flambeau. At Berens River,
a man and his sons can hunt and trap or fish commer-
cially or cut and haul wood together. At Lac du
Flambeau, Wiere most men work for wages, such cooperation
in production is not possible. There is sharing of
income and a common domicile with relatives, but not
the clustering of the houses of a man and his sons who
work together to produce a living for all of them.

Contacts with kin are maintained, but these
contacts do not seem to involve any particular group

of kin nor any particular kind of exchange of services.

Political organization at Lac du
considerably more complicated than it is at Berens
River. The number and variety of political units with
which the Indians must deal is much greater. These
political units include the tribal council and court,
The Federal Indian Service and Federal Courts, the

local school district, township, county, and state

27
governments. The school district and township govern-
ments arise as the result of the alienation of Indian
lands. About four-sevenths of the land of the reser—
vation is owned by whites (Hallowell 1955:339) who
organize and control the school board and township
government.

The school board is significant to the Indians
(whether they realize it or not) because the Indian
children attend the school for which the board sets
policy. The township government is significant to the
Indians because the local ordinances it enacts and
enforces apply to the Indians when they are in the
non-Indian four-sevenths of the reservation. The county

government is important to the Indians because it

p—u

adm nisters the various welfare programs on which many
of them are partially dependent. The policy of the county
welfare board is to discriminate against the Indians
in terms of the amounts of the benefits due them under
Federal and state law. The impact of the state govern-
ment is relatively indirect. The significance of the
Federal Government for the Indians lies in the fact
that the few laws which govern acts on Indian land are
laws of the Federal Government enforced in Federal
CourtS.

Tribal government is vested in an elected

Tribal Council of twelve members and a Tribal Court.

28
These are intended, under the Indian Teorganzzation
Act, to enact and enforce laws to govern Indian lands
which cannot, under the U.S. Constitution, be enacted by
Congress and enforced by the Toleral Government.
The Tribal Council nos enacted some ordiances requiring
children to attend school and others designed to in-
crease family stability. Under the prodding of local
white leaders and local Indian Service people, the
CourCil and Court have been successful in enforcing
school attendance, but not the other ordinances.

This failure is important since the internal
controls which reportedly functioned adequately under
ore-contact conditions and which are apparently ad-
equate at Berens River cannot be relied upon to the
same extent at Lac du Elambeau. Hallowell notes, from
his composite Rorschach profiles, that at Elambeau,
ther is ”...a weakening of the rigid control on which
the Inland Ojibwa depends, without any compensating
factor emerging," (1955:351). Thus external social
controls are more important at Lac du Elambeau than at
Berens River. Even so, 70% of the Elambeau Indians
tested by Hallowell and his students gave 50% or more
P resaonses (the criterion used for the statement about
rigid control quoted above). (Tetting and Hutchins 1966).

If it were not for this persistence of the Ojibwa

29
personality, the frequency of vandalism, theft, assault,
and homicide at Elambeau might well be considerably
higher than it is. Impressionistically these rates
seem to be high, but since no accurate figures are
available, it is difficult to judge.

The Tribal Council itself initiates little, if
anything, in the way of efforts to better the community.
Those activities which it does undertake are undertaken
almost exclusively on the urging of leaders who are
not members of the tribe. Lac du Elambeau has been
extremely fortunate to have had the services of an
intelligent, wise, and dedicated head of the local
Indian Service office, Dorothy Yellowcalf Stewart, her-
self a Pawnee. She is reSponsible for much of the
good that has been achieved by and for the Indians.

(In the summer of 1966, she was inducted into the tribe,
an unheard of honor for someone in the Indian Service,

0 1 o ’ 0 ’V’ 01 90 I
and given the name Ogimabinesi inwe, Chief-thunderbird-

 

woman.) Election to the Tribal Council confers little
prestige so far as most of the Indians are concerned.
Some of the Indians regard some of the members as "white
men." Some of these members also regard themselves as

white men.

Re ioious Behavior

#4

At Lac du Flambeau, unlike Berens River, most

30
of the Indian population is secular in orientation.

Boggs suggests that this may be due to the fact that

)

O

hristianity was introduced to Berens after long efforts
on the part of the Indians to secure the services of

a missionary and that the missions were identified as
Indian churches from the beginning. Whereas, at
Elambeau, there were no Christian churches until whites
became numerous and then the churches were identified

r”l

primarily with whites (1954: Ch 2).

The

H;

irst missionary arrived at Berens River in
1873 (Hallowell l955:l20), whereas whites did not beCOne
numerous at Lac du Elambeau until after the arrival
of the sawmill in l893. Thus Elambeau has a shorter
hsitory of contact with Christianity and one less
likely to secure converts. This may have something
to do with the survival of definitely Indian religious
ceremonies at Flambeau.

About hO people of all ages participate in the
Drum Religion and the "War Dance." Many of these also
carry on the Hidewiwin.

We were priviliged to partici

pate in many "song

. 'y . O I, 0’ . O O C .
serv1ces of the Drum Relioion (nimi idiWin) includino
u ’ u

a first fruits ceremony and a restoration of mourners,

’ r f . . O
to participate in th: "War Dance" (ogicida nimi 1d1w1n)

and to observe, without participation, some of the rites

31
of the Midewiwin. We also observed the wake and funeral
an old mide and a "medicine feast" (for the relief
OI asthma) in all of which the ceremonies were conducted
by several midE.
The ceremonies of the Drum Religion have been
described by Densmore (1913, 1929) and those of a

variant of the Drum Religion, the "Dream Dance,"

have been described by Barrett (1911). The Uidewiwin

 

has been described by Hoffman in a classic monograph
(188?). Although there were numerous variations from

the rituals and a genuine religious spirit seem to have
been preserved. The works cited suggest that there has
always been considerable variation in ritual from one
location to another and frequent innovations. Hallowell
describes the origin, in the Inland Group, of a Drum
Dance which seems to have the essential elements of the

Drum Religion and the Dream Dance although differing
in many elements of form (1955:160-163).

These essential elements seem to be communica-
tion with Hanido (the Great Spirit) by use of a

o 0 70V/ 0 o

sacred drum, which is referred to as Lisomiss (11t-
erally, "our grandfather") and by use of tobacco which
is generally smoked in a sacred pipe. Communication with

the Great Spirit is for the purpose of petitioning Him

/ I
for pimaee21W1n, life, in the sense of health,

 

2
longevity, and well-being for the participants.

*7

The range of variation in the rites of the Midrwiwin
is at least as great as in those of the Drum Religion.
Each Hide has his own songs and is usually unable to
sing those of another Hide, even those of his own
preceptor. The songs have a general equivalence of
form, but much variation of detail —- many of the
"words" are nonsense syllables. The paraphernalia of
the Midewiwin, drums, rattles, medicine bags, were as
described by Hoffman (1885:190-191). The Midewiwin,

like the Drum Religion and Dream Dance, is intended to

H0\

secure for its participants nimadez win.

 

Some of the changes in form are almost cer-
tainly decadent, resulting, in part at least, from the
small number of participants. When we returned in 1966,
we were surprised to find that there were still people
carrying on the Drum ieligian.

Although the bulk of the population at Lac du

Flambeau is not oriented to any church, it seems

justified to conclude that about 5% of the population

Ho

ta

Ho
3

ma ns a genuine traditional religious orientation.

 

Three things stand out from a brief visit to

Lac du Flambeau in August of 1966. First, an economic

93
.J

improvement. Second, what appear to be major political

(Y

developments. Third, tie surprising survival of the

Dr m mReligion.

Economic Developments

Lac du Flambeau has shared in the general
proSperity of the United States since 1960 and more of
tb e Indians seem to be adapting to the extent of;
becoming steady workers. Few Indians clerhed in the
stores in 1952 a; in 1966 many were working at such
jobs. Kore Indians seem to be engaged in tradifiional
crafts. The market for such objects had eXpanded
sufficiently to lead to more tanning of hides and
maeing moccasins and headed objects.

These changes suggest that expansion of ag-
gregate demand may indeed have results even in pockets

hard core poverty.

Political Developments

Even more impressive than the economic changes
are the political changes. The Tribal Council seems
to have become an effective body for the introduction
of change. People seem interested in what the Council
is doing and in its membership.

Three developments in public activity evidence

a new effectiveness for the Council. First a sewage

II I 11!! I." I lllll‘llll' ‘l‘llll l [I

 

 

31L
diSposal system; second, a water system; and, third,
8 public, low-cost housing project. These developments
came about as a result of the receipt of about $3OC,OOO
from the State of Wisconsin for some swamp land.
Mrs. Stewart (now retired from the Indian Service, but
still active in community affairs) apparently suggested
the projects. The Council not only managed to agree
on the value of the projects, but succeeded in getting
two-thirds of the enrolled members of the band to vote
in the referendum and carried on a successful campaign
for support. Although the housing project is partially
financed with Federal funds, it, too, required an ed-
ucational campaign to get it approved by the voters.
The "atomistic" Ojibwa have acted as a community.

These political developments do not seem to me
to indicate any basic change in the typical personality.
They seem, rather, to emphasize what can be done with an
appropriate political system.

The political system is modeled on the system
common to all levels of government in the United States.
There are two features which seem to be particularly
essential for the Ojibwa. The first is a device for
securing the assent of the people which does not require
them to interact intensively in a face-to-face situation.

The secret ballot used in all elections in this country

\1

35

seems to be nearly ideal for this purpose. The second
feature of the system is a leadership group which can
work together in fairly intensive interaction and achieve
consensus within the group. This leadership group must
also be able to convince the others of the desirability
of the proposed actions.

Clearly, the members of the leadership group,
unlike the rest of the population, must be able to
engage in intensive, face-to-face interaction. This seems
to require a personality different from that of the
typical Ojibwa. The members of this leadership group
at Lac du Flambeau do seem to be different in person-
ality. One family (the "Hunts") from this group is
included in the sample. This family is clearly differ-

eit from the others in the sample.

purvival of the drum religion

 

Most of the old people who were active partici-
pants in the Drum Religion in 1952 have died. Two of
the four former ogimae (head men) of drums are still
living, however, and the son (now in his forties) of

one of the surv1V1ng oc mag 18 now ogima of a drum

ail.

 

himself. His drum.was used for a song service which we
attended. There were fewer than half as many partici—

pants as there were in 1952, and, unless more younger

36
men become interested, there will soon be too few to
maintain even one drum at Lac du Flambeau. It was
reported to us that there are quite a few Indians from
other reservations who meet frequently to conduct Drum
services. Perhaps the automobile will permit the Drum

“eligion to survive longer than it otherwise would.

CHAPTER III

THE FAMILIES STUDIED

Introduction

 

The purpose of this chapter is to present
evidence concerning the extent to which the observed
people can be considered to be typical Ojibwa.

The first part of the chapter will be devoted
to the development of criteria by which to judge the
extent to which the people can be considered typical.
The remainder of the chapter will present descriptions
of the observed families in terms of these criteria.
Correlation of personality‘with cultural characteristics

The available evidence indicates (as one would
eXpect) that some Ojibwa conform more closely to the
typical personality pattern than others. This evidence
also suggests a correlation between the degree 0
tural conservatism and the degree of approximation to

the twaical personality.

V

71'

hallowell's Rorschach protocols (Micro-card

r+

Primary iecords in Culture and Personality) supper

fliree conclusions. First, in each of the three

( 3
’1
C)
U)

of Ojibwa, there is a wide range in the frequencies 0;
each of the Rorschach scores. Second, the differences

37

0
3:)

in the composite (or average) profiles among the
groups are due to differences in the number of people
whose scores approximate the scores of the typical
profile. Third, the number of people who have scores
close to the typical profile is greatest in the least
acculturated group and least in the most acculturated
group.

Working with the Menomini, uho are similar to
the Ojibwa both culturally and psychologically, Spindler
(1955) also found that composite Rorschach profiles
were related to cultural characteristics for five
groups on the Nenomini reservation and that the com-
posite profile of the "native-oriented group” most closely
resembled that of Hallowell's "Inland Group."

Boggs (195h;19v6), working with families from
two of Hallowell's groups, found interpersonal behavior
to be a function of level of acculturation. He also
found that members of the least acculturated families
were most likely to diSplay th. kind of interpersonal
behavior which he took to be typical.

All of these findings suggest that adults who
show typical Ojibwa behavior and children who are in
process of developing the Ojibwa typical personality
are most likely to be found in those families which are

culturally the most conservative.

39

Measures of cultural conservatism

 

Hallowell (1955:333- 33 h) and Boggs (l?5h:l78)
have used such measures for the Ojibwa and Spindler
(10 ms 107) for the Menomini. The measure used by

ach of these writers is specific to his problem and
to the particular groups or families with wh’e? he was
we r“ ‘

._ . q

\1

All three writers use religious behavior in
their measures, with Spindler making it his primary
criterion. Hallowell and Spindler both emphasize
language an means of dealing with illness. Boggs
emphasizes such social factors as participation in an
extended family, household composition, and the roles
of husband and wife. Boggs also stresses traditional
subsistence activities and Spindler finds possession

i A

of native objects sign1: Meant.

A measure of cultural conservatism

 

for this study, the following indices will be

1. Religious identification and participation.

T" °

nis criterion is not useful at Berens River
where everyone participates in Christian services, but

it is important criterion at Lac du Flambeau. At one

end of the relig eio us continuum are those who particila;e

in the Midewiwin, the Drum Dance, the war Dance, and

ho
related ceremonies. -t the other end are the fully
participating Christians. In the middle are those who

are secular or not clearly identified with either native

or Christian groups.

At the conservative end of this continuum are
those who do not use English at all. At the other end

are those who, while they may know some Ojibwa, SDBrk

(

q

bnglisn without Ojibwa mannerisms.

3. Source of income

At the conservative end of this continuuu are
those whose income is almost entirely derived from tra-
ditional sources such as hunting and sale of "medicines."
At the acculturated end of this continuum are those who
successfully participate in the economy of the white man.
Between these extremes are those who have a mixed sub-
sistence base, including those for whom the traditional
activities are followed for reasons of sentiment rather
than economics. Also included under this heading are

g

traeitional crafts, such as racing moccasins, whether for

.4...

own use or for sale.

L- Family type

At the conservative end of this continuum are

extended fa.ilies with clear, differentiated roles.

At the acculturated end are nuclear families with wide
departures from traditional roles of husband and wife
or roles disorganization. Toward the conservative end

families which, while not actually extended -am1lies,

(2')
C)

q

have strong ties with kin outside of the nuclear

'“)

r.)
.3
‘5

5. Treatment of illness
Since treatment of illness may have a ceremonial

V

aSpect, like religion, it does not differentiate among

f‘
.y
.L

the Ca adian families, but does differertiate among am-
ilies at Lac du Flambeau. At the less accult urated end
of the continuum are those who frequently use curing
ceremonies or participate in them, and who sometimes
also use the services of an U.U.. At the more accul-

tu acted end are those who use no native remedies and who

never participate in curing ceremonies.

6. Possession of native objects or practice of tra—
ditional behavior (in areas not covered above).

In the area of child care, for example, the
most traditional famil used the cradle board and its
associated wrappings, used the baby hammock, used

Sphaguum moss "diapers," and did not use the bottle.

A family toward the other end of this continuum used

h2
none of the traditional practices and the infant was
bottle-fed.
These criteria have not been worked into a

8

Ho
(‘1‘

genuine scale, but they do provide a basis which perm
the ranking of the observed families from least-
acculturated to most~acculturated, with only a couple of
cases where there is doubt about the placement of
families relative to each other.

The observed families are described below in
sequence from the most traditionally-oriented to the

least traditionally—oriented.

Berens River Households

 

The names used to designate these households

and their numbers are fictitious.

The "Wasse" household

This household is the least acculturated house-
hold of all those observed. It consists of Wasse (about
forty-five), his wife, a married daughter and her infant
son, a daughter about ten, and a son about four. wasse,
his wife and his unmarried children are summer visitors
from Little Grand Rapids, 110 miles up the river. The
married daughter with whom they are staying is a

regular resident and her husband is employed by one of

4.3
the fish companies. Further conformity to the extended
family pattern is indicated by the fact that this house-
hold is located next door to that of anather daughter
of wasse who is a pa manent resident of Berens

None of the people of this household could be
induced to use even a worh of bnglish and it seems
unlikely that any of them knew more than a few words.

A major source of income for wasse is the sale
of "medicines." Since all forms of religious and curing
ceremonials are forbidden at Berens, it is not known
whether or not he possesses any of the other shamanistic
skills. Hunting is also an important source of income
for this household. Wasse killed at least three moose
during the summer of 1952. Some of the meat was smoked
and the hides were tanned.

The married daughter is the only mother ob-
served who uses the cradle board. She also uses the
hammock for her baby (nearly universal at Berens) and
he is usually tightly swaddled in the manner used in
the cradle board. This infant is breast fed. His
mother and her married sister, who has a baby of
similar age, frequently sit and talk holding the in—
fants on their laps.

On three occasions the four year old was

observed "nursing” at his mother's undoubtedly dry

1
int

breast. These were occasions when he was unhappy,
once apparently because of the attention his sister's

baby was getting.

The Young Huhoman household

This household is related to the Wasse house-
hold through Young Muhoman's wife who is a daughter
of Wasse. Young theman and his wile are both about
twenty—eight years old. There are five children, a boy
of eight, and boy of six, a boy and a girl, both about
four (one probably adopted from another family), and a boy
of about six months.

Young Hukoman speaks English better than most
Berens River Indians, the eight year old boy also
Speaks it relatively well; and both are more at ease
with white strangers than are most of the Indians.

The other members of the household use no English.

Young Huhoman is a successful commercial fisher-
man and trapper and also gets considerable subsistence
from hunting. He and his "first cousin” went hunting and
returned the next day with two moose and a dozen ducks.
Game was also in evidence at other times. This house-
hold moves to the Fall commercial fishing area early ii

1 0

September. They live in a tent during the fall fisnino.

V

The baby is bottle-fed, and the hammdch is

L15
used. He is swaddled tightly in cradleboard wrappings,
but the cradleboard is not used. This family was
observed to use native herbal medicines in the treat-

ment of children's illnesses and also to use the facil—

In"

ties of the nursing station at the Q.C. mission.

There are occasional periods of a day or two
when large numbers of relatives stay at the Muhoman
house.

I

Toward the end of the summer, the gisip
household (to be described below) pitched a tent ad-
jacent to the Yaung Muhoman house. This followed the
marriage of the widowed Mrs. J 1 to "Old Muhoman,"

father 0f Young Muhoman.

The Muhhwa household

The household consists of Muhhwa, his wife, a
boy of six, 8 girl of four, and a baby girl of one
year. Muhhwa's widowed mother also lives with them,
sometimes occupying a tent pitched next to the house.

. . VTV.’

During the early part of the Summer, tne 8151p household
occupied a tent pitched next to the Mukhwa house.
T‘ 1 ’ I"? er< VT 1
r llOW1ng the marriage of mrs. 8151p to Young muxoman's
father, they pitched their tent next to Young Mukoman's
house.

'hwa Speaks English fairly well, his wife

M6
knows some English, and the boy knows a little.

Family income comes primarily from commercial

H)

ishing at which Muh wa is quite successful to judge
by his house and its furnishings. Dr. Nufeld's list
describes him as a "successful fisherman."

The baby is breast fed, it is swaddled, and the
hammock is used. Mrs. M kkwa and Mukkwa's mother were

frequently obsel erved mal ting and beading moose-hide

moccasins. They were not observed using native
medicines but fliey did avail themselves of the services
of the R mission nursing station.

y_/
S].

The Eis p - Old fiukoman household
This household occupied a tent for the entire
period of observation. For most of the Summer, the
tent was pitched near the Mu hhwa hous Later it
was pitched near the house of Young Hukoman after the
. . Vrv (

marriage (informal) of Abigail 813319 and Old Mukoman.

Abigail has eight children, but the older ones
are actually present in the household only occasionally.
(It is not unusual at either Berens or Flambeau for‘
adolescents to stay with kin, moving from one household
to another from time to time.) Occasionally her
unmarried daughter and the daughter's three-year-old

son are present. Her twenty-year-old son is also

present only occasionally and the children of fourteen,

in
twelve, and ten are sometimes absent for extended
periods.

Abigail's chief source of income until her
remarriage was the government issue of rations and the
“mother's allowance." Her twenty-year~old son fishes
(non-commercially) and hunts, providing some additional
subsistence.

Abigail and Old Hukoman both Speak some

English as do the school-age children.

Lac du Flambeau Households — lQSe

 

The Mandamin household

Old mendgmin is eighty-eight and his wife is
eighty-two. The other regular members of the household
are a boy of five and his sister who is nearly three.
Occasionally the mother of these children is present.
Although the children call the Mandamins "grandfather"
and"grandmother", the relationship of the children anl
their mother to the Mandamins is not at all clear.
The father of these two children is now living with
another woman who is legally married to a white man.
The mother is living with a white man off the reservation.
She returned to the mandamin household briefly after

flie birth of a baby fathered by her current ”husband.“

L8
Old Handamin is the senior member of the
Nidewiwin and a leader in all the Indian religious
ceremonies observed. He and his wife Speak English
only with considerable difficulty. The boy and girl who

live with them are fluent in the Ojibwa language -- a

‘3

rarity f r children at Lac du Flambeau. -heir mother
also sneaks Ojibwa fluently.
The chief source of income for this household

s rublic old age assistance. Old Mandamin attempts

l-‘0

”J

U) eke out this oittance by hunting and snaring rabbits.
This activity appears to be even more important to him
psychologically than it is economically.

Both of the old people regularly wear beaded
buckskin moccasins and the man owns and uses a "war
dance" drum and a water drum and rattles, (used in
curing ceremonies). When the children's mother brought
the new baby into the household, he was given a new
aair of beaded moccasins with the traditional holes

in the soles so he could not make the long journey to

e land of the Spirits. This was done on the insistence

5’1

the old people who also wanted the baby to be in a

O
H,

hammock. The baby is bottle fed and the hammock is

not used.

The Waswagen household

The regular members of this recently formed

'r

1'9

household are Waswagen, seventy-s'x, his wife ”Annie,"
fifty-nine, her brother's widow (addressed as "Auntie")
and the six-year-old daughter of Annie's son. The
adolescent adopted son of "Auntie" -- her deceased
sister's son -- is occasionally resident in the house-
hold. Other hin are also occasionally resident for
periods of two days to a month.

The six-year-old granddaughter is usually
addressed with a diminutive form of her Indian name
rather than with her Christian name. Her mother is a
Menomini, but she has been raised by her father's
parents since the age of one. Her father's father with
whom she and her grandmother lived until recently was
a conservative old man, active in Indian ceremonial
life. He is now bedridden in a state mental hospital
there his illness is diagnosed as senile psychosis.

Waswagen is a man of exceptional character,
whole-heartedly dedicated to maintaining Indian values.
He is the real leader of the Midewiwin and the Drum
Religion, although he defers to the older Kandamin in
matters of cermonial form. He leads the rites for the
dead, participates in curing ceremonies with water
drum and rattles, and prepares herbal medicines. Annie
also participates in all of these religious ceremonies --

she is an active member of the Midewiwin and the Drum

50

F

s o- doctors as

(‘3

he family uses the servic

d
r.)
H
Ho

(13
H.
O
:3
O

#3

well as native medicines and ceremonies for curing

Annie Speaks English well, though with Indian
mannerisms, and can read it well enough for most purposes.
Waswagen was raised off the reservation and never went
to school. His English gives the impression of a trans-
lation from thoughts in Ojibwa. ie cannot read or
write, but has recently eXpressed interest in learning
to read.

A

The family income is chiefly derived :rom various

forms of welfare, but is supplemented by earning from

-

Waswagen's sale of "mediCines” and of his services

as a midE. In addition, he and his wife "made rice,"
and made maple syrup -- activities which necessitated
their moving to the ricing area and the maple grove

for periods of about a week. Waswagen also hunts,
trans, saears suckers during the Spring run, and snares
rabbits in winter. As with Mandamin, these activities
are even more significant psychologically than econom—

ncome is also gained by tanning deer

Ho

30130

Ho
0
O)
H
[.4
“d
I
(

hides, from which Annie makes moccasins for sale as
well as for their own use. (Some of the designs for

beadworh are taken from Lyford's Ojibwa Crafts.)

 

Waswagen's preference for the old ways extends

51
to a taste for such foods as porcupine and skunn and
he exaressed his pleasure that Annie, in contrast
to his previous wife, was willing to prepare these

foods.

The Robinson household

(5)

This household was observed in 1952 and a ain

in 1966. Since there were major changes in the mem-

4

bershia of tne household and in the roles of the mem-

\

bers between these dates, it Will be described first

as it was in 1952 and then as it was in 1966.

"'D

The household consists (as o 1952) of Old Robinson
and his wife, both about sixty, their son, twenty-seven,
and the son’s wife, nineteen, and two daughters, twenty—

eight months and nine months.

3-»

Old Robinson and his wife are active partic
pants in the Midewiwin, the Drum Religion, and other

native ceremonies. The younger Robinsons are irequent

s i these activities

'3
{-1-
:3

a

L

spectators or marginal partici

L'-

A

and their two—year-old child is eiten present at the

dances. Old Robinson and his wife speak English with

(*1-
:3“

Oiibwa mannerisms and use it in talking with e

Vu
granddaughter, but both are more at ease Speaking Ojibwa.
The younger Robinsons are fluent in English, but re-

q

tain the Ojibwa manner of saeech. If they sneak Ojibwa,

\fl
[0

it is only in private.

The c inf source of support of this household
is the ”a no of the younger Robinsons. Both are
stead'ly emtloyod, he as a driver of fork lift trucks
and bulldozers and she as a worker in the local
electrical meter plant. This employment leaves the
care of the children to their grandparents during
moat of the day. The baby is sometimes left with her
mother's grandmother because the baby attacks her
older sister and frightens her (report of mother).

Old iobinson has arthritis and no longer hunts,

' and Spears

but his son hunts frequently, "makes rice,’
fish during the Spring "runs." These activities not

sfactien, they also add

t-‘o

only give him considerable sat
to the household's subsistence.

Old Hrs. Robinson makes beaded bucks in shirts
and moccasins and the older child has a pair of
moccasins which she wears. The baby is bottle-fed

but the hammock is used.

The Robinson household in 1966

This is still an extended f_mlly household,

Q.)

q

but there have been changes in its mcmbr snip. Old
Mrs. Robinson has died, and, in 1953, Young Robinson

and his wife separated and he married again. The two

5’3
daughters stayed with their father. The older daughter

is now married and living away from the reservation.

II

The daughter who was an infant in 19,2 is now a can—
able housekeeoer and the chief caretaker of her four
younger half-sisters in the Summer. She is paid for

her services. She is a good student in high school.

She also participates actively in the dances whiCh are
nut on for tourists and to a limited extent in the
ceremonies of the Drum Religion.

The children of the second Mrs. Young Robinson
are girls of eig t, seven, and six, and a baby girl of
three weeks.

Their mother is not Indian in physical char-
acteristics; her hair is light brown and her skin is
fair and she is not on the tribal roll. She sneaks
English with definite Indian mannerisms and considers
herself to be an Indian. The baby is bottle fed, but
the hammock is used and it is tightly swaddled in the
manner of infants who are out in a cradle-board.

Young Robinson still works for the same employer
and his second wife, like the first, is employed in
the meter nlant. He continues to hunt, Spear fish, and
gather wild rice. Desoite the clock-pacing necessitated
by this employment of the adults, the children still

have much of the freedom of eating times characteristic

51!—
of the Berens River Indians. That is, there is almost
always some food available so that they eat when they
feel like it, rather than on a schedule. For instance,
the children were observed to eat cold pancakes when
they got up, much as the children at Berens River

‘!

at~ bannocks whenever they felt line it.

(

Since 1952, the Robinsons have added running
water, a bottle—gas cook stove, a television, and a
teleoone, but these somehow blend into the Indian
style of furnishing and housekeeping and remain un-
obtrusive..The interior of the house seems much the
same as it was in 1952 desnite these changes. One
change adds to the "Indian-ness.” The baby's hammock
is now in the "living room" whereas it was in the second-
floor room in 1952.

Three of the large drums used in the ceremonies
of the Drum Religion are now kept in the nobinson house.
A song service was held in the Robinson's living room
with the drum and the singers in front of the television

vhile one of the children watched the service from

nobinson's narticipation in the Drum

erigion was only nerieneral in 1952, but he is now

a drum -- the drum.which was

)4
l 7
I)
Q.
Ho

n the song service mentioned above.

\f'L
\R

The Uarshall household

This household, like the Waswagen household,
is rather newly formed. Karshall returned to the
reservation about three years ago after living 6133'
where for thirty years. His wife has five children by
two previous marriages. The two oldest, girls of eighteen
and seventeen, are not often at home. A girl of nine and
a boy of seven are regular members of the household.
The youngest child, a boy of five, is in a tuberculosis

sanatarium. Unlike the nreviously described Lac du

"tr i

Flambeau families, the Marshalls live in the new Village.’

Both of the Harshall adults are deeply involved

n the Midewiwin, the Drum Religion, and the War Dance.

Ho

Marshall was installed as "heater" of a drum at one
of the ceremonies which we observed. They were also

observed oarticipating in curing ceremonies and the

b3

ites for a mide who died. The younger children
oarticipate in the Observances of the Drum religion
and are eresent at other ceremonie .

The Karshalls speak Ojibwa fluently, but file

children are not learning more than a smattering of

heir subsistence is derived from welfare
eayments, from Mrs. Marshall's work as a cook at a

A

resort during the tourist season, and irom odd jobs.

Marshall also does some fisning as a means of adding

to their diet.

The Pitch household

Betty Fitch has had six children by a nrevious
marriage and has two by this one. The three oldest
children are not living with her and her present
husband. The oldest of the children living at home is
a girl of ten; the youngest, a girl of 1h months.

The Pitches do not participate in native
religious ceremonies nor in either of the Christian
churches. Jack Fitch has had an extensive exoosure
to native religious practices —— his mother is a member
of the Midewiwin and an active participant in the Drum
Religion. Betty's ten year 01d daughter has an Indian
name. She is addressed by a diminutive form of the
name, but the whole name is not revealed. This suggests
more native orientation than is revealed by the other
characteristics of this household.

Both Pitches, but esnecially Jack, sneak English
well. Betty has more Ojibwa mannerisms in her Speech.

The income of the household is derived from
Betty's employment in the electronics plant. This
leaves Jack in charge of the house and the children,

a role in which he is neither happy nor s illf l.

5;.

No native artifacts were observed in the pos-

S7
session of the members of this household.

he baby is bottle-fed and snends most of her

This is a disorganized household. Betty and
Jack have frequent drinking bouts in which he sometimes
beats her. Neither the children nor the house are
well-cared-for, although they do get improved attention

on weekends.

The Pitch household in 1966
In 1966, the Fitch household consists of Betty

"boarders" -— one a young man of about eighteen

and two
and the other, Ben Ellis, a man of thirty-five. This
household is included because Betty regularly cares
for the five-year-old son and two-year-old daughter
of her daughter, Xary St. Louis, while Hary and her
husband are employed in the electronics factory.

This household has more of an Indian flavor in
1966 than it did in 1952. There is a tanning frame
built into the trees in the front yard and Betty was
observed making and beading buckskin moccaSins and other
buckskin objects for sale to tourists. Betty also

participates regularly in the dances put on for tourists.

In the dances she wears a beaded dress and moccasins.

The St. Louis household

This household consists of Jim St. Louis, thirty,
his wife Hary, twenty-four, and their two children,
Billy, five, and Alice, two. Kary is the daughter of
Betty Titch. She was ten years old when the Fitch family
was observed in 1952.

Their house is located near the house of Jim's
widowed mother (who is white). Jim and Mary are both
employed at the electronics plant, and the children
are usually left with Mary's mother, Betty Fitch,
during working hours. Host of the observations of the
St. Louis children.were made at Betty's house. Since
Jim spent most of his non-working hours fishing, it
was not nossible to observe his interaction with the
children.

The family seems to be secular rather than
religious in orientation. Their identification is,
however, clearly Indian, and Mary is known by a dimin-
utive form of her Indian name. Hary's Speech shows
many Indian mannerisms, but she knows only a few
Ojibwa words. She does beadworh for sale to tourists,
but otherwise there are no traditional objects in the

house.

The Peed household

This household consists of Pauline and her five

children, ages four to twelve. Their house is located
very close to that of Pauline's mother and the fam ly
seems to live at her mother's house as much as at
Pauline's. Pauline's father is dead and her mother does
not have a man. Pauline's sister is sometimes present
and occasionally a boy of about sixteen, to whom Pauline
refers as a nenhew. The father of Pauline's children

18 also the father of the two children who were living
With the Yandamins in 1952. He was living with Pauline
in 1952, but not in 1966.

heligiously, this household is quite secular

and seems to have little contact with anyone with

either a native or Christian religious identification.

7:)

lauline knows a few Ojibwa words, but can not sneak
the language. All of the family sneak English with
Indian mannerisms.

The income of this household seems to be derived
primarily from the emoloyment of Pauline's mother and
:rom the occasional emoloyment of Pauline at a resort.

This household, at least in the summer of 1966,
seemed to be nart of a matrilocal extended fanily, out
the absence of any adult males certainly made it an
atynical family.

No traditional objects or practices were ob-

served.

f“.

-ne Clauson household

»4

his household consists of Roger Clauson, his
wife, Agnes, Agnes' aged and blind mother, and their
grandson, aged seven. Two other grandsons are frequently
“resent. All observations on these children were made

in this household.

Roger is white and Agnes is Indian. The grand-
sons are children of two of the sons of Roger and Agnes.
All the children of Roger and Agnes are on the tribal
roll. The boys' mothers are Indian and are no longer
married to the boys' fathers. The oresent wives of the
boys' fathers (the boys' sten—mothers) are also Indian.

Roger is a relatively successful small business-

I‘

ran.and their house is much the ovst o; the houses of
any of the families observed. It gets some Indian flavor
because Agnes' mother makes birch-bark baskets for sale
and one room is full of her sunalies and her inventory

of finished baskets.

Roger does not share the Indian emotional
restraint. Onr one occasion, his wife had calmly told
the resident grandson that she was angry. Roger said
to her, "You're mad. Why don't you say you're mad’l‘i

!

Agnes did say that she was ’mad”, ut she did't sound

much more emotional than she had the first time.

51

usehold is secular in orientation and

"7
if
FA
U)
.‘7‘
C.)

there is little contact with those Indians who maintain
traditional ceremonial activities. Evidences of tradit-

oral culture are slight, except that Agnes and the

H-

grandsons Sneak English with Ojibwa man rrisms.

The Hunt household in 1952

The Hunts regard themselves as Indians and they
have considerable knowledge of native customs, but
their acceotance of the values of the white middle
class is quite genuine and fairly complete. John and‘
Ann are both about thirty. They have three children --
two boys, seven and five, and a girl of four.

They attend the Roman Catholic Church regularly
and are devout in their observance of the tenets of
the faith.

Their income is primarily derived from John's
employment in the electronics plant where he holds a
position of considerable resronsibility. Their sub-
sistence is considerably improved by John's success
as a hunter and fisherman and by their use of a freezer.
They also harvest wild rice.

The interior of their house is finsihed and
painted (both unusual). The living room is furniShed
with a table, chairs, and a sofa; it is not used for

a bedroom and nothing is stored along the walls or

62

'1‘?“

under the sofa or table. lacy have a bottle—gas
cooking stove, electricity, and running water. The
housekeening is such that one who has been raised on
our standards of cleanliness could accept an invitation
to a meal without reluctance.

‘A row of evergreens has been alanted along the

road in front of the house and flowers have been filanted

near the front entrance.

The Hunt household in 1966
The three children who were observed in 1952
have left the reservation, the two boys for em;loyment

anngpgirl through marriage. There are two children

\fl

born since 19 2; a girl of thirteen and a girl of six.
John has achieved a position of even greater
resaonsibility'with the electronics company which has
grown considerably. He is active in the Knights of
Columbus and the Lions Club and has been elected to
the Tribal Countil.
Their house has been further improved by the

addition of a complete bathroom and numerous adpliances

and new furniture.

fish and "ma“e rice",

John continues to hunt and
but this is clearly a household which largely adheres to

the values of the white middle—class, deSpite the identi-

63

R

*ication of its members as Indians.

’_

Summary

The samsle consists of thirteen households,
four of them from Berens River and nine from Lac du
Flambeau. As judged by the criteria developed in the first
nart of this chapter, the four Canadian families and
the first three of the Wisconsin families (Mandamin,
Waswagen, and Robinson) can be considered quite con-
servative. The next four of the Wisconsin families
(Xarshall, Fitch, St. Louis, and Reed) are clearly less
conservative, although the Marshall's participation
in native religious activities luts them relatively
closer to the conservative group. The Clausons and
esnecially the Hunts are clearly distinct from all of
the others. The Hunt's fuller acceptance of the values
of the white middle—class sets them apart from the

Clausons. This samnle has a definite conservative

u)
’1)
’ 5
Q.
r,-
,‘3‘
(0

bias desaite the inclusion of the Clauser.

J

Hunts. Four of the nine Lac du T (nbrau families are
from the five aercent of the nonulation wlich carries

on native ceremonial observances.

CHAPTbR IV

THEliIES OF The NOTIVATION.AND UbvblOlfiENT

OF OJIBWA RLSTRAIVT

The literature provides two theories of the way
in which this type of personality functions, and two
theories of the learning process through which this
personality is developel. The two theories of the
functioning of Ojibwa typical personality'will be con-

sidered first.

Theories of the Hotivation of Restraint

Perhaps an example from my field notes will give
some body to the reports of Ojibwa restraint. The
nature and extent of this restraint can best be illus-
trated by an extreme example. The incident which is
described below is the most extreme example of non-
aggressive reaction to provocation in my field eXper-
iences. The incident occurred at an evening ceremony
of the Drum Relieion held in one of the houses in the
Old Village at Lac du Flambeau.

The description of the incident must be are-
faced by some information about the Drum Relieion which

\J

6h

1

-‘

is needed to understand the erovocation. The drum, itsr_.

3

a;
is sacred. It is addressed as "Our Grandfather"
(nisomiss)- it may be touched only be certain designated
officials of the group; if the dr~n is moved from one

II.

housw to another, dr if a mistake is made in its
hardling er ritual, a dance must be held "to apologize
to the drum"; only sacred songs may be played upon it;

ani, finally, no one who has been drinking should

aaproach the drum (from field rotes 195?, see also

Barrett 1911; Uensmore 1999). With these nreseriptions
an1 prohibitions in mind, the significance of this in-
cident becomes clearer.

Dancing had been going on for some time when an
Indian, who was obviously very drunk entered. At
first, he created only a mild disturbance by trying
to get rCSponses from the children. The drumming
and singing went on, but the adults who were not
drumming became quite tense. when none of the
children were willing to reSpond to his overtures,
he took a drumstick and sat down at the drum. He
began to play a "Squaw dance" song. (”Squaw
dances" are golely for entertainment, the music is
therefore profane.) He also yelled such things
as, ”Come on, let's dance, let's have some fun!"
All of the Indians became tense and silent and the
singers left the drum. No one said anything to the
drunken man or made any effort to prevent further
sacrilege. Eventually he tired of trying to get
narticipation from the group, sat on a bench along
the wall, and went to sleep.

As 5001 as the drunk left the drum, the au-
pointed officials hastily put the drum away.
Then,one of the old men came to me and asked me
to tell the Indian nolicemen to come and take the
drunk away because the old aeoale who lived in
the house did't want him staying there all night.

66
The provocation here is certainly extreme, the
drunk has committed and is continuing to commit
multiple sacrilege (in addition to having inter-
rupted the “song service"), yet no one makes any

effort to stop him or to punish him for his sacri-

legious behavior.

Introduction

 

Ho

The first of these thsor es is so widely ac-

nt that it is often

l-‘o

ceated as an eXplanation for restra
taken for granted as the only possible explanation.
This theory has been advanced by Hallowell to exnlain
Ojibwa behavior. The central notion of this theory is
that hostile eeople avoid all behavior which is likely
to annoy some other nerson because they fear that the

other nerson will retaliate and inJure them severely.

”fear of others theory.”

I shall call this theory the

The second theory of Ojibwa functioning is de-
rived from the writings of Freud, Erikson, and Bettel-
heim. The central notion of this theory is that hostile
neOple avoid all aggressive actions because they fear
that such actions will do more damage to the other

nerson than they wish to do. I shall refer to this as

U18 "fear of self theory."

Hallowell's statement or the ”fear of others" theo,y
of Ojibwa behavior

 

Hallowell does not concisely state fliis theory

#40

in one location but t is all to be found in Cheaters 6

and 15 of Culture and Fxoerience. Briefly stated,

 

this theory says that, desoite a tendency to be hostile,
the Ojibwa suopress "...any impulse to tell someone

else what to do..." (1955:135) and refrain from "overt
aggre551on in face-to face situations..." (1955:278)
because, "...there is always the possibility that another

m
i

man's nower may be greater than my'own. lo provoke him

to exercise it by offending him, is always a gamble,"
(10552290) and ”... the best defense is to avoid offense..."

’0

(19§§:290). Tnis forbearance is found even ‘...in all

ace relations with others that in—
evitably must have aroused emotions of annoyance,

anger, or a desire to criticize or correct..." (195%:

137).

Restatement of the “fear of others“ theory

In short, this theory asserts fllat when annoyed,
the tyoical Ojibwa would "like to" choose a course of
action which would be likely to produce the outcomes
of stooning the annoyance and punishing the person who

caused the annoyance. But he refrains from actions

6n
K.)

which are likely to produce these outcomes because he
believes these actions would be very unlikely to pro-

duce another outcome which he wants -- avoidance of

severe injury to himself through the other's retaliation.

Some imolicit assumntions of this theory

 

Hallowell seems to have used (probably without
.1. J.

full awareness) the idea that neonle unconsciously cal-

culate tne exnected value of oossible actions as a

 

basis or deciding which action to choose. His argu-

ment (presented above) that the Indians avoid offending
others because of the "possibility that another man's
oower may be greater...” suggests that Hallowell con-
ceives of the Indians' estimating this probability. That
he sees them giving un all other outcomes which might

be secured by an offensive action because of a small

a o

probability of suffering severe retaliation indicates

that Hallowell believes that the Indians value the
outcome "not being seriously injured" more than they value

.3
H)

any combination 0 other outcomes. This version of

the theory can easily be stated within the eXpected
value frame work developed in contemporary decision

theory.

69

ue Franeworh for oredicting_the choices

 

 

A

For any course 0: action which an individual
might choose, there are at least two (and usually more)
possible results, or outcomes, of choosing that course
of action. Usually, the individual ureters some of these
outcomes to others. That is, each outcome has a value
for him and the values are not all the same.

H I"
~

ror each course or

In

action, there is a certa-n

objective nrobability that eacq o; the outcomes will

 

occur.

The individual forms estimates of these prob-
abilities based on his exoerience -- estimates which may
diverge quite widely from the objective probabilities.
These estimates are called the individual's subjective
probabilities.

 

4

Thus, for eacn course 0? action, each outcome
has certain value to the individual and also a certain
probability of occurrence. In choosing a course of action,
the individual must take into account both the values
of the outcomes and the probabilities of their occurrence
for each course of action. This is tak'n to be done through

"calculation" of the exoectcd value of all courses of

 

action. The subjective exoected value (to an individual)

of a oiven course of action is the sum 0 the nroducts

\./ _

70
obtained by multiplying the value of each outcome by

its subiective probability of occurrence. The notion

 

n

involved here is exemoliiied bythe old adage, "A bird
in the hand is worth two in the bush."
It is not asserted that the individual whose

behavior we are try to predict actually goes through

Ho
:3
(.0

any such ire c<:ss of calcula ti 1ng expected value. What
‘3 asserted is that such a process of calculation per-
mits the observer to predict the subject's probable

choices with increased accuracy.

 

In ormal statement of the "fear 0; others" theory in
the evoected value fr amrvo.l.

 

In a situation in which some other person is

being annoying, the Ojibwa are pictured as valueing

l-g—b

three outcomes: ( ) stopping the annoyance, (2)

juring the annoyer, and (3) avoiding being seriously

t-ao
:3

injured themselves.

These three outcomes and their negates can occur
in eight possible combinations. Hallowell indicates
that the Ojibwa value any of the four combinations
uhich include avoidance of injury to themselves much
more highly than they value any of the four combinations
which involve such injury. Their most highly valued

combination of outcomes is indicated to be that in

which the annoyance is stopped, the annoyer is

71
punished, and the self is uninjured.

To produce this outcome seems to require an
aggressive action -- certainly the annoyer can not be
punished otherwise. The Ojibwa, however, rarely choose
aggressive actions and the usual outcome achieved is:
the annoyance is not stopped, the annoyer is not punished,
and the self is not injured.

Hallowell explains this preference for non-
offensive actions by asserting, in effect, that the
Ojibwa believe that any action which is likely to
produce the stopping of the annoyance anl the punishment
of the annoyer also has a high enough probability of
producing retaliation with severe injury so that they
do not choose offensive actions in this situation.

This amounts to saying that, given these values
of the Ojibwa and their subjective probabilities, the
expected value, to the Ojibwa, of non-offensive actions
is enough greater than the expected value of offensive
actions so that they almost always choose non-offensive

actions. This places their subjective probability estimate

AL

of retaliation at the center of the exnlanation of the r

 

choi

O
O

.,8.
Since the subjective probability of retaliation
plays such a central role in this theory in accounting

for Ojibwa choice of non-offensive courses of action,

72
then the behavior of Ojibwa toward persons who are
believed to be incapable of retaliation shduld provide
a clear test of this theory. The reactions of Ojibwa
adults to offending children should provide such a
test since the Ojibwa believe that no child possesses
any supernatural powers. They are eXQlicit that children

are extremely vulnerable as a result of this lack

This version of the "fear of others" theory can
be partially formalized by assigning numerical values
to the values of the outcomes and to the subjective
probability estimates of the Ojibwa which are consistent
with Hallowell's statements. Given some other reasonable
assumptions, we can determine the levels of the values
of the C ibwa and the subjective probabilities re-
quired to predict, within thiS‘ framework, their very
high probability of choosing non—offensive actions when

annoyed. Such a formalization of this theory is

presented in the appendix-

The "fear of self" theory in the literature.

This theory appears in Freud's 1909 "Notes on
a Case of Obsessional Neurosis" although not expressed
clearly and concisely. One sentence will serve to

' v .

indicate Freud's use of this theory: "Tram the: 1:10

0
’1
I

73
ward he was a coward -- out of fear of the violence of
his own rage” (7F6ud 1963363).
Bettelheim has also employed this theory in

the interpretation of extreme inhibition:

In order to protect herself anl those around her
from her own violence, whose consequences she ex-
aggerated megalomanically, she inhibited herself
in every direction: intellectually, physically,
an%)socially she was unable to budge (1950:207-
20‘ .

Erihson, too, follows this theory in interpreting a
case of severe inhibition of the use of the hands
in a little girl (1963::95—208).

Each of these three children seems to have
formed a self-image in which the self is conceptualized
as extremely powerful and dangerous to others-~even
adults. The little girl described by Erikson indicated
through her drawings that she conceptualized her
hands as huge and powerful. In each case, the child’s
inhibition of aggressive actions rests in part on the

image of the self as extraordinarily"powerful.

§tat§ment pf the "fear of self” theory

 

This theory asserts that, when annoyed, the
typical Ojibwa would "like to" choose a course of action

which would be likely to produce the outcomes of stop—

7h
ping the annoyance and punishing the person who caused
the annoyance. He refrains from actionsxvhich are
likely to produce these outcomes because he believes
that due to his great powers, any action he might

choose which is likely to produce these outcom s is

('0

very likely to produce the outcome of the very serious

IHJUTy or death of the other person -- an outcome which

U)

he wishes to avoid, since he also has warm feeling
toward the other. The person believes that there is
a high probability of doing serious damage to someone
else if he takes any offensive action at all when he
is angry.

" .p

Informal statement pf the pear of sol?" theory in the

—_

 

_i fi<— a
expected txflwie irameworn

 

In a situation in which some other person is

bein

(O

annoying, the Ojibwa are pictured as valueing three
outcomes, two of which are the same as those involved
in the "fear of others" theory: (1) stoiping the an-

noyance, nd, (2) punishing the annoyer. The third

CL)

outcome, however, is quite different from that in the
"Fear of others" theory. The third outcome involved
in this theory is that of serious injury to the annoyer.

As in the other theory, those four combinations

H)

0 outcomes which include the third outcome, in thjj

 

case "no serious injury to the annoyer," are valued

('1-

much more highly han any 0; the four outcomes thich

njury to the annoyer." In this theory,

In"
:13
O
H
[:2
{3...
(‘3
U)
0
”‘3
H.
O
{I
U)
H-

tn most highly valued combination of outcomes is
considered to be that in which the annoyance is stopped,
the annoyer is punished, and the annoyer is ngt seriously
injured.

It would seem that this outcome could be pro-

sivr -ction which would punish

3
('3
C.)

(.0

duced bj*a limited aocre

(

..J
(0

the annoyer without doing hi: .crious injury. The
Ojibwa seldom attempt such actions and the most frecuent
outcome actually produced is: the annoyance is not
stonned, the annoyer is not punished, and the annoyer

is not seriously injured.

the Ojibwa for non-offensive
actions is exalained within this flieory by the as-
sertion that they believe that any action they might

choose which is lihely to stop the annoyance and

punish the annoyer also has a high probability of doing

lv—Jo

serious njury to the annoyer.

.

This can be roahrased to say that, given these
values of the Ojibwa, and the subjective probabilities,

ctions is so low that

D.)

the expected value of offensive

 

’J

ca—eJ it is their su‘iective rooability estimate oi

U

 

76
serious injury to the annoyer which is central to

the exnlanation of the Oiibwa's choices.

 

As in the case of the "fear of others” theory,
this theory will be tested by observing adult behavior

Ho

toward offend n

O

hildren. In this case, the critical

issue is not whether or not children posses retaliatory
capabilities, those are irrelevant. The important
question is the extent to which adults believe that
directing or attahing a child will injure the child.
This theory, too, is presented with the addit-

V

ional precision of partial formalization in the appendix.

Theories of the Development of Ojibwa Restraint

The central sroblem of theoriesgof develgpment

# A.—
*—

 

As they were stated above, the two theories

of the way in which the Ojibwa type of

’(j

ersonality
functions exalain the individual's preference for non-
aggressive behavior in terms of the individual's belieig
regarding the consequences of aggressive and non—ag-
gressive behavior. That is, they explain the individual's
behavior in terms of his subjective probabil'ty

estimates. One theory asserts that the individual
estimates that, lollowing aggressive behavior on his

hart, the arobability that he will suffer severe in-

77
jury to himself is unacceptably high. The otler theory
asserts that it is the estimated probability of serious
injury to someone else following aggressive behavior
which is unacceptably high. Within this frame work,
then, a theory of development must account for the
formation of estimates of the probability of these
outcomes.

Although it is not usually stated in these terms,
this seems to be the kind of problem with which learning
theorists have long been concerned. It seems aporo-
priate, there ore, to consider the problem of the devel-
opment of Ojibwa personality as a problem in learning

and to use observations of the Ojibwa to test two dif-

ferent conceotions of the learning orocess.

Two conc/otions of the learning process

 

These two conceatio.s of the learning process

,3

differ primarily in the extent to which reasoning

on the part of the learner is considered to be involved.
That group of theories in which reasoning or inference
is considered to play a very small part in the learning
process will be called "Group I." The group of theories
in.uhich reasoning or inference is considered to play

a larger role in the learning process will be called

H

"Group II." Both theories apply equally to conscious"

752

" learning. The behavior

learning and to "unconscious
with which we are concerned here is largely unconscious.
That is, the people would find it difficult to verbalize

their values and beliefs about interpersonal behavior.

 

Group I theories are alike in containing the
assumption (usually not exolicit) that an individual
will come to expect outcome 01 to follow course of

action Cl if and only if he has chosen course of action

 

Cl and it has been followed by outcome 01.

Some examples may help to clarify the way in
which this assumption differentiates Group I learning
theories from those of Group II. (Group II theories
vdll also be exemplified belmv.)

A child (whom we will call ”Joe”) will come to

exoect that punishment (01) will follow if he annoys (Cl)

 

an adult only if .e actually annoys an adult and

punishment follows. This statement also says that,
1: Joe does not annoy an adult, he will not come to
exnect that punishment will follow his annoying an

adult.

n

Ho

4'
all

3

Joe will come to expect that

(D

an BHCTY

attack (Cl) on an adult will seriously injure (01) the

 

adult only if Joe actually makes an angry attack and

79
the adult is injured. Further, if Joe never makes
an angry attack on an adult, then he will never come
to expect that an angry attack will result in serious

injury to an adult.

Group_II le_rnino theories characterized
Group II theories also assume that the indi-
vidual will come to expect outcome 01 to follow course

of action Cl 3: the individual chooses course of action

Cl and it is followed by outcome 01. But, Group II

 

theories assert in addition, that the individual will,
under some conditions, come to expect that outcome 01
will follow course of action Cl ii outcome 02 follows
course of action C2, even though the individual has
never chosen Cl. The conditions which make this pos-
sible are that the individual sees a relationship
between C2 and Cl and between 2 and 01.

Group II theories, like Group I theories,
hold that, if Joe annoys an adult and gets punished,
he will come to exnect punishment for annoying adults
and Group II theories also hold that, if Joe angrily
attacks ar adult and the adult is injured, he will make
the connection between his behavior and the injury to

the adult. In addition, Group II theories assert diet

 

Joe will come to expect that serious injury to an ad“*t

-AL—v

 

f‘n

~\”

(01) will follow an aggryge tack (Cl) by Joe even
though Joe never makes an angry,attack on an adult_£:
Joe makes a very mild, non-angry attack, such as a
disturbing noise, (C2) and the adult accepts the injury
(disturbance) without either retaliation or_controlling
Joe's attack_(p?). In other words, Joe reasons that,
since the adult doesn't prevent Joe from annoying him,
he would not (or could not) prevent Joe from injuring
him severely if Joe were to make an attack.

These examples oversimplify the case, since
both theories will be stated in terms of the probabil-
ities that the outcomes will follow the Specified

causes of action. Newrer has stated a Group I learning

theory in terms of probabilities.

H

Mowrer's_partial statement of
in terms of subjective nrobab

aCheup
ility est

earning theory
tes

L

Q) r--

H.

m

 

 

. . . we must assume that probability has, first

of all, a basis in the real world which, by means
of learning can be gradually reporduced in the
nervous system so as to give a sense of conviction,
or set of exoectations, which corresnonds fairly
precisely (one would hope) to prevailine external
reality. Thus, of necessity, we assume an external
reality'qpbjective probability) and an internal
(personalistic) one; and learning is the process
whereby the latter comes to match the former
(1960bz3h7) (italics in original).

A more complete statement of Mgwrer’s position

In the context of the rest of Mowrer's discussion

m

\l

of learning, a fuller statement of this Group I theory
in terms of subjective probabilities would be: If

(I) in any given class of situations, in which
several courses of action and several outcomes are
available,

(2) the individual chooses each of the courses
of action available in that situation many times, then

(3) the individual's estimates, for each of
these courses of action, that it will produce each of
the outcomes, will corresoond "fairly precisely"
to "prevailing external reality" or the "best estimates"

oi these probabilities.

Implications of Group_I theory;§gr the jSbwa

If this learning theory is correct, and if it
is correct that the Ojibwa do not choose offensive
actions because they believe that such actions have
high probability of producing severe retaliation, then
the objective probability of such retaliation shduld
correSpond with the subjective probabilities. The
probability that an offended person will take direct
retaliatory action toward an offender is very low.
It is this observation from which the ”emotional

restraint" of the Ojibwa is inferred. It is also this

observation which makes it necessary to assume that any

82

effort at retaliation must be by the ind1rect route

of sorcery. Eat the obiective or robabili 1ty that someone

 

O A

will be injured through sorcery is zero —- 11 one accepts
contempora1ry scientific assumptions. There is, therefore
a large discrepancy between the subjective probability
of retaliation and the obje ective probability of re—

taliation.

Resolution of failure of subjective probabilities to
correSpond with objective probabilities

Such failures of correspondence between subjective
and objective probabilities are not rare and they pose
a problem for Group I theories. (See Mowrer 1950: ch 18
for a discussion of this "neurotic paradox.” ) One
method, (the "fixation ap oproach" , of resolving this
problem is to consider the present subjective prob-

abilities as corresoonding to the objective probabil-

 

ities with which outcomes actually occurred in largely
(but not completely) similar situations of the past.

This view is adopted by Buss:

The way to insure a generalized 1n nhib1tion of
aggression is to punish it in a variety f situations
so that the individual learns that aggression is
associated with so many stimuli that it is simply
not safe to aggress (1961:59).

“a
(V?)

H

Dellard, et al also adopt this view: ...those actions

cease to occur which, in the past, have been followed

by punishment" (1939:33). Another method of dealing with
this problem the "identification approach", holds
that the child "identifies with” the rewarding and punish—

ing parent —- a mechanism by means of which the child
can reward himself for behavior previously rewarded

by the parent (Sears, Rau, and Alpert 196S:S). Pre-
sumably, this also leads the child to punish himself for
acts previously punished by the parents. Similar

views are expressed by Bandura and Walters (1963:92-100),
Cameron (19h7z99-102), uarsons (1961: l7h-18h), Spiro

(1961:116-121), Whiting and Child (

'-—l

953:2uo-aua)

‘0

v1

and Berkowitz (lQéL

Both aoproaches to dealing with this problem
assert that the subjective probabilities which do not
correSpond with present ”reality” do correSpond with

i H

a 'reality which was CXperienced in the past when the

subject was a child.

Summary statenrn; o' implications for the Oiibwa

...- -.-—~—-.1.-.._

 

 

If either the ”“iration approach" or the

identification approach" is correct, the objective

 

 

 

probability of severe punishment for offensive acts

by Ojibwa Efljlgggg should correSpond with the 23?:SC-

0
(11'

L
322$ aroba bilities necessary to explain'the behavior

e? Ojibwa adults within the fee r of others theory.
?ormal analysis of this theory reveals that this prob-
ability must be at lea a .hS in order to account for the
behavior at Ojibwa adults (see appendix).

It is this implication of Group I theories

which is to be tested.

Group II learninnfl theories in the literature.

 

 

Concise anl explicit statements of Group II

t rature.

Ho

learning theories are not abundant in the l

’7)

(1‘

Kellv aroues that all people learn hrough an informal
version of scientific hypothesis-testing and exolicitiy'

introduces the notion of indirect eV1deence:

 

A good scie entist tries to bring his constructs
up for test as soon as possible. But he tries
hem out initially in to st- tube pro oportions.
I hazards appear to be great, he will first seek
some indirect evidence on the probable outcome.
or his trialsfiTT§g§7l§ emp.asis supplied).

Indirect learning in the literature

 

This notion of indirect testing of hypotheses
is an important feature implicit in explanations of the
inhibition of aggressive behavior in some psycho—

analytic case histories.

0—!

U)

rs

rreud's 1909 work, ”Notes on a Case of

H
(in
”'3
m
0
C)
U)
{—r
5‘
O
D‘

Neurosis” (1963:62-63) eeinning

”u

Obsessions
of the oatient's inhibition of agoressive behavior to
an incident in which the patient, at the age of three,

concluded that his father was afraid of the boy's rage:

...he “an done something naughty, for which his
father had fiven him a beating. The little boy
had flown into a terrible rage and had hurled
abuse at his father even while he was under

his blow“ . . . His father, shaken by such an
outburst of elemental fury, had stopped beating
hih . . . (7063:32-53)

The boy apparently continued to test the hypothesis

that his father was afraid of him since he noted that

the father never ‘eat him a
A similar use of the notion of indirect hypo-

of the t

"3

eat—

her

U)
H.
,J
.1
l-N
U)
C)
3
u.
U)
C)
O
O
O
:3
:3
r!-

thesis testing occur

1- 3

<- t‘
3‘
CD
S12
U)
C O
O

ment of a child with severe inhibition 0

girl's inhibition was due to her belief that her hands

were daigerous to other people. This notion received

indirect confirmation for her when her handling of a

I"

light plug was followed by a blown luse and extinguisned

as

A more exnlicit statement of Groue II thgory in an—
glication to the ‘evelopment of restraint

 

 

The general notion of this theory is that the
individual draws conclusions about what would haeaen

n a critical

Ho

action

0
C
5...
C)
"3

'f he were to choose a narti

situation from.what does nan en in a different situation

 

lS angry at another individual (that is, the first
individual puts fairly high value on the outcome 01
moderate injury to the second irdividual) but he places
still higher value on 23: seriously injurine the second

individual. From his experiences in another situation,

on he might choose which has

Ho

he concludes that any act

0)

high arobability of aroducing moderate injury also has
a hieh probability of aroducing serious injury. He comes

Q

to this conclusion desnite the fact that he "never"

l—N

chooses an action wit h gh probability of producing Egg:
crate injury when he is in this situation. He does not
choose this action because he hypothesizes that it might
aroduce serious injury.

The first individual concludes that his hypo-
thesis that a course of action with a high probability
of producing moderate injury also has a high probability

of producing severe injury to the second individual

is correct because, in the second situation, in which

87

he is n33 angry at the second individual, an action
which has high probability of producing slight injury
to the second person has a high probability of pro-
ducing an acquiescent reaponse from the second person.
(An acquiescent resoonse is an action which involves
no counter-offensive actions sufficient to produce a
change to non-offensive behavior on the part of the
first individual.)

The individual assumes that the probability
that the hypothetical offensive action in the critical
situation would produce severe injury to the other
person is equal to the frequency with which his
slightly offensive action in the second situation does
produce an acquiescent reSponse from the other.

For example, a child who is naughty, who offends
an adult by being noisy, finds that, although the adult
doesn't like his behavior the adult does not make him
change to acceptable behavior. Since the adult does not
control the child, the child concludes that the adult
can not control him. Hence, the child infers that, if
he made an angry attack on the adult, the adult would

be seriously injured.

87
Testing the Theories of Functioning and of Learning

The two learning theories can be tested adequately
only if the subjective probability estimates necessary
to account for the typical behavior of the Ojibwa
are known. Estimates of minimum values of these sub-
jective probabilities have been obtained from the partial
formalizations of the two theories of the way in which
this type of personality functions. Since the tests
of the learning theories are dependent on the theories
of functioning, tests of the learning theories also
constitute tests of the functioning theories. The
theories of adult functioning and the theories of learning

can be tested with the dame data.

Analytical categories

We are interested in classes of outcomes of
situations in which a child's behavior has been offensive
to an adult. These outcomes are defined in terms of
the behavior of the adult in reSponse to the offensive
behavior of the child and whether the child's behavior
remains offensive or becomes acceptable following the

adult’s reSponse.

Offensive Behavior of a child

Offensive behavior of a child is any behavior

88
which might offend (or irritate or annoy) an adult.
Basically such behavior is aggressive, but, because

we are concerned here with extremely mild aggression

 

in most cases, I have used the term_9ffensive rather

than the term aggressive, which seems to connote stronger

 

actions. The term offensive should not be construed in

 

the sense to which defensive is an antonym. A child

 

who has been asked to do a chore and who withdraws
3

before doing the chore is behaving offensively in the

 

sense intended here. This correSponds fairly closely

 

with the sense of "changeworthy behavior" as used by

Sears (Sears, Maccoby, and Levin 1957).

Adult ReSponses to offensive behavior

The possible reSponses of the adult which are
relevant to the theories of concern are: the adult
may (1) neither tell the child what to do nor punish
him. (2) tell the child what to do but not punish him,
(3) not tell the child what to do but punish him, and

(h) both tell the child what to do and punish him.

Outcomes of the situations
With each of these four adult reSponses, the
child's behavior may remain offensive or it may become

acceptable (non—offensive). All together, there are

89
then eight possible outcomes. The theories imply that
different relative frequencies of these outcomes should
be observed for the class of situations in which the

child's behavior has been offensive to the adult.

§pecification of the operational meanings of these terms

Offensive Behavior is any behavior which makes

 

it more difficult for another to produce outcomes which
he values. Especially for children, offensive behavior
includes:

(a) quarreling, verbal or physical;

(b) disturbing adults by requests, or by
noise or commotion in the vicinity of the adult;

(c) acting in a manner potentially damaging
to property, including touching adult tools or work
in process;

(d). behaving in a manner potentially dangerous
to the self;

(e) failing to comply with any indication,
however subtle, of the adult's wishes;

(f) failing to conform to household routine;

(g) violating ritual prohibitions.

Behavior of an adult in reSponse to offensive behavior
of a child.

Directing is a brief eXpression for "telling

 

90

another what to do." In general directing is less

 

intensely offensive that attacking. The term will

 

be used to refer to any behavior on the part of one
person which is intended to influence another individ-

ual's choice of a course of action. Directing Specifically

 

includes:

(a) 'using imperative sentences addressed to
another;

(b) requesting that another person perform some
act;

(c) warning of the possible outcomes of choosing
a course of action, such as telling a climbing child,
"You'll fall;"

(d) asking a "question" which conveys intentions
to have the other change his behavior, such as a mother
asking a child who is sitting in her lap, "Do you
want to get down?"

(e) physically controlling the behavior of
another, such as graSping and holding a running child;
(This kind of "directing" may sometimes be so strongly
offensive as to constitute an attack.)

In general, attacking is offensive to a greater

 

degree than is directing. Attacking includes those

 

actions which are primarily intended to produce physical

or psychological pain. Specifically included in attackigg

 

are:

91

(a) all actions, such as striking, which may
produce tissue damage and physical pain;

(b) scolding, telling another that he is
unacceptable or "bad;"

(c) ridiculing another;

(d) restricting the interaction of the indi-
vidual with others;

(e) depriving the individual of a privilege.

Attacking behavior which is a response to

 

offensive behavior of another is called puni§hing.
For the purposes of this study, any punishment in the

above sense will be considered "severe" punishment.

Some problems in determining outcomes

 

It is relatively easy to determine which of the
four courses of action an adult has chosen. It is
sometimes more difficult to determine whether a child's
behavior should be considered to be acceptable or
offensive. There are two problems here.

The first problem concerns cases of "pseudo-
compliance" which may either initiate or continue
offensive behavior. These cases involve compliance with
the "letter” of the instructions, but not with the
"Spirit." Perhaps some examples will make this dis-

tinction clear. The first is from Berens River.

92
A man is building a fire. He calls his six year old
son who comes to him. The man tells the boy to get
some firewood. The boy gets a small stick on which

he has been whittling and puts in in the fire, then
returns to his play.

This is regarded as offensive behavior on the part of the
boy, since the man's intention was clearly to have the
boy collect a pile of wood.

The second case involves a six—year-old Lac du
Flambeau girl who lives with her grandparents. These
incidents take place at a ceremonial dance held in a

house in "The Old Village."

The girl is playing with a boy and his sister. She
becomes ver noisy. Three women (not including her
grandmother) tell her to be quiet. After each
admonition she is quiet for a few seconds, then
becomes noisy again immediately.

This is taken to be a continuation of the offensive be-
havior, since the intention of the adults was for her to
remain quiet and she did not.

Another problem arises when there is more than one
adult within range of the child's offensive behavior and
all ignore it. When this occurs, it is scored only for
the adult who is primarily reSponsible for the child.

The second problem concerns cases in which,

following some adult direction or punishment, the child's

93

offensive behavior clearly continues. Since the child
must eventually choose some acceptable action, under
what conditions should this be considered a situation
with an outcome involving offensive behavior? The decis-
ion concerning Uiis problem was often made in the field
and the decision to consider that the outcome involved
offensive behavior depended on several factors. In
those cases where the adult directed or punished, one
important factor was the length of time the offensive
behavior persisted without further directing or punishing
by the adult. There must have been sufficient time for
the adult to become aware that the child was contin-
uing his offensive behavior and to direct or punish if
he chooses. This time varies with details of the sit-
uation such as the nature of the offensive behavior and
the case of adult observation of the continued offense.

More important than the lapse of time is the
tendency of the adult to shift his attention from the
child who continues to misbehave. This attention shift
gives the impression of a deliberate (but not necessarily
conscious) decision to ignore the child's continuing
offense.

In case where the adult neither directs nor
punishes, and the child's behavior eventually becomes

acceptable (perhaps because the adult has made an adjust-

, 9h
ment), the situation will be considered to have ended with
the child's behavior still offensive. Group I theories
would seem to hold that adult adjustment to the child's
offensive behavior which led to acceptable behavior would
constitute reward of the offensive behavior and hence
would be likely to lead to higher, rather than lower,

frequency of choice of offensive behavior.

Analypis of sample incidents

As indicated above ( p 88), eight possible out-
comes of a situation on which the child's behavior has
been offensive have been distinguished on the basis of the
implications of the theories being tested. The primi-

tive outcomes are:

oa = child is directed by adult = "directed"

ob = child is punished by adult = "punished"

oc = child's behavior is offensive = "offensive"
("oc'" = "not offensive" = "acceptable")

The eight combinations of these are:

Oa = oa'ob'oc' = not directed, not punished, accep-
table ‘ ‘

Ob = oa ob'oc' = directed, not punished, acceptable

Oc = oa'ob oc' = not directed, punished, acceptable

Od = oa ob 00' = directed, punished, acceptable

Oe = oa’ob'oc = not directed, not punished, offensive

Of = oa ob'oc = directed, not punished, offensive

Og = oa’ob 00 = not directed, punished, offensive

Oh = oa ob 00 = directed, punished, offensive

9S

Implications of the Grogp_l learning theogy

Group I theory implies that Co, Od, Og, and Oh
(the four outcomes which include severe punishment)
should occur for each child with a combined relative
frequency of at least .h5. That is, on at least h5%
of the occasions in which a child is offensive, he
should be punished to account for the high degree of
restraint. (This frequency estimate is derived from
the formal analysis of the fear of others theory.

Appendix).

Group II theory implies that the adult should
show an acquiescent reSponse to the child's offensive
behavior with a relative frequency of at least .75.

An acquiescent reSponse was defined as one in which there
occur no offensive actions sufficient to produce a change
to acceptable behavior on the part of the child (p.57 .)
That is, on at least 75% of the occasions when a child
offends an adult, that adult does nothing which actually
results in the child's changing to acceptable behavior.

Oe, Of, Og and Oh are outcomes involving an
acquiescent reSponse by the adult. The combined relative
frequency of these four outcomes should be at least .75
according to the formal analysis of the fear of self theory

(Appendix).

6

\O

The analysis of some incidents of adult—chil

h
..
I
..

interaction which allows will indicate the use of these

outcome categories.

Behavior samples analyzed

 

The following behavior samples involve two
siblings o: the St. Louis family, Alice, two, and Bill,
five. lhey are in interaction with their maternal
grandmother, Betty, while their mother, Mary, is away at
he first incident described is the first ob-
servation of this family on this field trip. Betty and
Hary are known to the field workers from previous
field trips. The following incidents took place outside

the front door of Betty's house.

(1) After initial greetings, Bett‘ said, ”This is my
little granddaughter," pointing to Alice. I
asked Betty for name of granddaughter. Betty
said to child, "Tell him your name.” Child made
no reSponse and Betty said nothing further.

Offensive behavior on the part of Alice begins with her

A O

iailure to comply with the adult's d

H.
'1
O
O
(4'
H.
O
:3
0

Betty
gave no further directions and did n t punish the child
and the child never complied with the original direction.

In terms of the outcome categories, once the child's

behavior became offensive, the adult chose the action

'97
"neither direct nor punish" and the child's behavior
remained offensive. The outcome of this situation is

therefore classed as "Oe."

(2) Alice went to the door of the house and tried to
open it. Betty jumped up and opened it for Alice
and followed her inside. Alice shortly emerged
eating a soda cracker and followed by Betty.

Alice's attempt to enter the house is regarded as offensive
behavior on two counts -- it seems to be a non-verbal
request which disturbs her grandmother and Alice's going
alone after something in the house is a threat to prop-
erty. Betty's reaction is considered to be "neither dir-
ecting nor attacking" since she does not attempt to step
the child from disturbing her nor from getting into the
house and getting the cracker -- Betty aids her, perhaps
as a device to limit the possible damage to the cracker
supply. This incident is considered to have ended with
the child's behavior still offensive since it became ac—
ceptable through an adjustment by the adult to the child's
demands. The outcome of this situation thus falls in the
class Oe. Two more incidents, essentially similar to this
one occurred, adding two more outcomes to the Oe class.

(3,h)

98
(5) Alice again went to the door and tried to open it.
This time Betty held the door shut and said,’"Uh,
Uh. Later1"’ Betty also remarked to my wife,
"She's really got me running!" Alice desisted

from her efforts to get in for the remainder of
this observation period.

This time, Betty's reSponse to the offensive
behavior is directing and not punishing and the child's
behavior becomes acceptable, hence this outcome falls

in the class Ob.

(6) Bill and some other boys were throwing pebbles
at each other in the area around the house.
This was noticed by Betty who ignored it.

Throwing stones at other children and being a
target for their throws is offensive behavior since it
is potentially dangerous to the child and to property.
Since Betty noticed the offensive behavior and neither
directed nor punished the boys who continued their ac-
tivities, the outcome of this situation fails in class Oe.

Later, there were three incidents separated from
each other by several minutes, in which Betty directed
the boys to stOp throwing stones (7,8,9). In each case,
the boys continued and Betty shifted her attention to
other matters. The outcomes of these three situations

all fall in the Class Of.

99

(10) Billy told me a story about having driven a
snowmobile. Betty listened awhile and then
said, "You lie, Now I know you lie! Your dad'

drove it. You just rode." Billy looked down,
smiling faintly, and said no more about snow-
mobiles.

This incident is considered to have the outcome
Oc, since Betty told Billy that he was bad, but did not
tell him eXplicitly to change, and Billy's behavior

became acceptable.
Methods of testing the Hypotheses

Sglection of aggrgprigte_tests

Each of the theories purporting to explain Ojibwa
emotional restraint implies a hypothesis for each adult,
that, when the adult is offended by a child, the adult
should be observed to choose each of four courses of ac-
tion with a Specified relative frequency. The theories
differ in the relative frequencies of choice which they
imply. Similarly, each of the learning theories implies
a hypothesis for each Child. Each child should be
observed to receive Specified frequencies of either
punitive or acquiescent resoonses when he offends an
adult. These theories also differ in th‘ relative
frequencies which they imply.

In all of these cases, the question is one of

lOO
whether or not observed frequencies are consistent with
the theoretical expected frequencies. Therefore, the
appropriate tests are one-sample tests of goodness of
fit. Since none of the data involve measurement in an
interval or ratio scale, only non-parametric tests are

appropriate (318991 1956:29"33)°

A test for the theories of emotional restraint

In testing the theories which purport to explain
Ojibwa restraint, the concern is with the goodness of
fit of observed frequencies with eXpected frequencies in
four categories. Theng test must be ruled out because
some of the expected frequencies are smaller than 5.
The Kolmogorov-Smirnov test not only lacks this ob-
jection, but may be in all cases a more powerful test
than the X2 test (Siegel 1956:51). The Kolmogorov-
Smirnov test will therefore be used for testing the hypo-
theses which puroorts to eXplain the individual behavior

of the adults.

A test for the learning theories

For tests of the learning theories, where the
data are dichotomous, either the binomial test or the
7? test would be apprOpriate. Since expected frequencies

here will also often be less than 5, theTK? test will

101

not be usable and the binomial test will be employed.

Statistical hypotheses from the theories of restraint

 

Statistical hypothesis from the "fear of others" theory
Formal Analysis of the "fear of others" theory
yeilds two sets (depending on the assumptions) of prob-
abilities of coice of courses of action by the adult.
(Matrix Bl, p.216and Matrix B2, p.221). According to
Matrix Bl, the adults should choose Cl with a relative
frequency of .11, C2 with relative frequency of .22,
C3 with relative frequency of .35 and Ch with relative
frequency of .32. The outcomes which involve adult choice
of Cl are Oa and Oe, the outcomes which involve C2 and
Ob and Of, outcomes which involve C3 are Oc and Og, and
outcomes which involve Ch are Od and Oh. Thus, according
to this hypothesis: p(Oa+Oe) = .11, p(Ob+Of) = .22,

p(Oc+Og) = .35, and p(Od+Oh) = .32. The cumulative

 

distribution of these expected frequencies is used in
the Kolmogorov-Simirnov test (Siegel l956:h7-52).
The theoretical cumulative distribution of outcomes is

then as follows:

Course of action Ch C3 C2 Cl

Theoretical cumulative .32 .67 .89 1.00
frequencies

102
(The order of listing of the outcomes has been reversed

to facilitate computation.)

Statistical hypothesis from the "fear of self" theory

As formalized (Matrix Cl, p.223) the "fear of
self" theory implies that adults who are offended by a
child should choose Cl with a relative frequency of .91,
C2 with a relative frequency of .039, C3 with a relative
frequency less than .029, and Ch with a relative frequency
of .021. The outcomes which involve Cl are Oa and Oe,
those which involve C2 are Oh and Of, those which involve
C3 are Oc and Og, and the outcomes which involve Ch
are Od and Oh. The theoretical cumulative frequencies

of these outcomes are as in the table below:

Course of action Ch C3 C2 C1

Theoretical cumulative .02 .03 .09 1.00
frequencies

Both of these theoretical distributions will be
tested by comparing the observed distributions with
the theoretical distributions for goodness of fit using

the Kolmogarov-Smirnov test.

Example of K-S test

As an example of the use of this test, the hypo-

133

"fear of others" theory is tested

thesis deduced from the
below for one adult using the incidents presented as

samples of the field data.

TABUSlpl

SAF'UDLE AUPLICATIOI:A\.A1:—C TBS To 1
CHILD I‘IT293CTIOM (KUER A NYNVWMCI

(YET- t‘fuc 1! PM- Frnwx
V‘lL;.\\) LAA‘M \ .

IEII”CIBEUK3.J ADU'i-
S EROLETTE? "EEARO

Course of action Ch. C3 C2 Cl
f = observed frequencies 0 l h 5
Eo(X) = theoretical cumulative .32 .67 .89 1.00

distribution under ho

Sn(x) = cumulative distribution .00 .10 .50 1.00
of observed choices

V
5)

he K-S test concentrates on the largest difference

 

 

of

"3
Po
b4
C
(.Q.
H.
O
:3

between the Uieoretical cumulative dist
outcomes and the cumulative distribution of observed
outcomes and indicates the probability of a difference
as large as that observed if the observations are a
random sample from the theoretical distribution. In
this example, the largest value of this difference

(denoted "3") S .57. The table of critical values

i-‘O

fl

of D indicates that the probability of u as large as
.57 with N = 10 is less than .01. Therefore it seems

unlikely, on the basis of this example, that the "fear

10h
of others" theory is a satZSfactory explanation of the

behavior of Betty in these instances.

Derivation of statistical hypotheses from the theories
of learning statistical hypothesis from the Group I theory
The Group I theory asserts (in essence) that the
subjective probability of retaliation for offensive
behavior must be equal to an objective probability of
retaliation eXperienced earlier. The formalizations of
the "fear of others" theory indicate that the minimum
subjective probability of retaliation sufficient to ac-
count for the high probability of choice of Cl is .h5
(Matrix Bl p.216). The outcomes whiCh involve retaliation,
or punishment, are Oc, Od, Og, and Oh. Thus the sta-

tistical hypothesis is:
Ho:p(Oc + Od + Og + Oh) _>_ .h5

Statistical hypothesis from the Group II theory

The Group II theory asserts (in essence) that
the subjective probability of doing serious injury to
another by an aggressive act is equal to the objec-
tive probability (eXperienced earlier) of getting an
acquiescent reSponse to an offensive act. The formal-

ization of the "fear of self" theory (Matrix C1, p. 223)

105
indicates that the subjective probability of doing
serious injury to another must be about .75 in order to
account for the very high probability of choice of C1.
Therefore the probability that an adult will reSpond
acquiescently to the offensive behavior of a child must
be about .75. An adult reSponds acquiescently if he
chooses a course of action which does not produce an
outcome involving acceptable behavior on the part of
the child. The four outcomes which involve continued
offensive behavior on the part of the child are Oe, Of,
Og, and Oh. The occurrence of any of these outcomes
indicates an acquiescent reSponse on the part of the

adult. The statistical hypothesis is therefore:
Ho:a(Oe + or + 09 + Oh) 2; .75.

Both of these hypotheses will be tested by use

of the binomial test of goodness of fit.

An example of the use of the binomial test

As an example of use of the Binomial test for
the Group I hypothesis, the five incidents (previously
described) of interaction between Alice and her grand-
md her, Betty, resulted in four outcomes classed as Oe

and one outcome classed Ob. That is, not one of these

106
five observations fell into the class predicted by Ho.
The binomial test gives the probability of observing X
or fewer instances of the Specified class. Since, in
this case, x = O, the test will give the probability
of exactly 0 observations if the true frequency is .h5.

This probability is calculated from the formula:

N N - x
p(x) =( ) pXQ-- (Siegel 1956:37)
X

Substituting in this formula we get:

5
p<o) 2(0) (.u5)0(.55)5
p(O) = l X l X .0503
p(O) = .05

(In the actual application of this test in Chapter V,
the probabilities were obtained by use of the computer
or from tables.)

Therefore, if Ho is correct, the observation of
no cases which are consistent with Ho when N = S is
improbable -- i.e. the hypothesis can almost be rejected
at the .05 level for this child on the basis of this

example.

107

The tests described and illustrated above make it
possible to decide for each adult which theory (if
either) better predicts his behavior and, for each
child, which theory (if either) better predicts the kind
of responses to his offensive behavior which he experiences.
The acceptance or rejection of any of these theories
as an explanation of the functioning or development of
Ojibwa typical personality depends on the proportion
of the sample which the theory eXplains.

If it were certain that the sample of people
were random and if the true proportion of typical per—
sonalities in the pOpulation were known, it would be
eXpected that the proportion in the sample would not
differ significantly from the proportion in the_p0pulation.
Unfortunately, the sample is not random and information
about the prOportion of the population whose person-
alities fit the typical pattern is limited and is based
on non-random samples. The information available about
the sample and the population does permit some rough

estimates as to the proportions to be expected.

Relative frequency of typical personalities in the
pppulations studied

 

There is little published information on which
to base a conclusion concerning the actual proportions

of the Ojibwa who diSplay the hostility and restraint

108
which are taken to constitute the ”core" of the typical
personality. Lowenfels' analysis of children's drawings
from Lac du Flambeau led her to conclude that all of
the children showed "constriction" or restraint and that.
63% of these children were "severely overcontrolled"
(Quoted in Barnouw 1963:282). This suggests that the
prOportion showing the typical personalfiLy should be
near 60% to 65%. An analysis of Hallowell's primary
records of Rorschach protocols from the two locations
found that 70% of the Lac du Flambeau subjects gave
50% or more pure form reSponses, which is one accepted
measure of overcontrol. Eighty-seven percent of the
subjects from the Lakeside group gave 50% or more pure
form reSponses (Tetting and Hutchins 1966). The two
studies, taken together, suggest that 60% or more of the
adult population in each place should be typical Ojibwa
and that 60% or more of the children may be expected
to be in the process of developing the typical person-
ality. ‘This expectation about the children is based
not only on Lowenfels analysis of children's drawings,
but on Hallowell's conclusion that the typical person-
ality has persisted through many generations of contact

with people of western culture.

109

Characteristics of the sample

A separate chapter has been devoted to the
characteristics of the families which are included in
the sample. This chapter seems to justify the con-
clusion that the sample is probably biased toward families
with a relatively traditional orientation. If the
assumption is made that those families with the rel-
atively more traditional orientation (as judged by cer-
tain social characteristics) are more likely to behave
in the typical fashion, then it seems justified to
conclude that the sample does not under-represent the
relative frequency of typical personalities. Boggs
has presented considerable evidence that behavior varies
with these social characteristics. (195h: Ch 2). It
will therefore be assumed that, if the behavior of
h0% or more of the sample is inconsiétent with the
hypothesis derived from a particular theory, the hypo-
thesis can be rejected and the theory from which it was

deduced questioned.

CHAPTER V

RESULTS

Theories of the Motivation of Restraint

The :fear of others" theory

 

We will begin with the two theories of the mo-
tivation of Ojibwa restraint. The first of these two,
which was derived from Hallowell's (and the Ojibwas')
attempts to explain.this restraint, has been called the
"fear of others" theory. Briefly and informally stated,
this theory asserts that the Ojibwa refrain from directing

or attacking another Ojibwa who is annoying them be-
cause they fear that doing so will lead the other to
retaliate by means of sorcery. Since the Ojibwa are
also hostile, if this theory is correct, then an Ojibwa
who has been annoyed by someone who lacks magical powers
should readily attack a defenseless annoyer. Since
children are held to lack magical powers, it should
be expected that Ojibwa adults would attack (punish)
annoying children with high frequency.

A formal decision model of the typical Ojibwa
as this very restricted theory assumes him to be was

construcua1(see appendix). From this model, it was

110

111
calculated that the typical Ojibwa, when annoyed by a
child, would choose Cl (neither direct nor punish)
with a relative frequency of .11; that he would choose
C2 (direct but not punish) with a relative frequency of
.22; that he would choose C3 (not direct but punish)
with a relative frequency of .35; and that he would
choose Ch (both direct and punish) with a relative

frequency of .32.

Observed reSponses: composite treatment

Three or more reSponses to annoying behavior of
children were observed for 29 adults. The total number
of reSponses for these 29 adults is 671. Of these
reSponses, 350 (.52) were Cl, 298 (.MM) were C2, 13
(.02) were C3, and 10 (.02) were Ch. Table 5.1 com-
pares this observed distribution with the theoretical dis—

tribution deduced from the "fear of others” theory.

TABLE 5.1

DISTRIBUTION OF OBSERVED RESPONSES OF TWENTYFNDESOJIBWA
ADULTS COMPAIED WITH THE THEORETICAL DISTRIBUTION FROM
THE "FEAR OF OTHERS" THEORY

Courses of Action%
Ch C3 C2 C1

theoretical distribution .32 .35 .22 .11
observed distribution .02 .02 .hh .55

 

XThe order in which the courses of action are
listed has been reversed to facilitate computation in the
K-8 statistical test.

112
InSpection of Table 5.1 indicates that there
is little correSpondence between the observed relative
frequencies of choice of the four courses of action
and the theoretical relative frequencies. Where the
theoretical frequencies are high, the observed fre-

quencies are low and vice versa.

Obserypd FCS:pflSCS§ individual treatment

The composite observed frequencies, of course,
present an average picture of the reSponses of all
twenty-nine adults, whereas the theory purports to
predict individual behavior. Table 5.2 presents the
data individually for each of these twenty-nine adults.
Table 5.2 includes a Kolmogorov-Smirnov test of the
significance of the observed deviation from the

theoretical frequency for each adult.

113
TABLE 5.2

OBSERVED RESPONSES OE ADULTS TO OEEENDING CHILDREN AND
TEST OF HYPOTHESIS FROM "FEAR OF OTHERS" THEORY FOR
EACH OE TWENTY-NINE ADULTS

f = observed frequency of choice of each course of action

 

 

 

 

Fo(X) = theoretical cumulative distribution of choices

under Ho from the "fear of others” theory.
Sn(X) = cumulative distribution of observed choices.
D = maximum /ro(x) — Sn(x)/
p = probability associated with D under Ho
Adult No. 1 John Hunt L du F 1952
Course of action Ch C3 C2 C1
f O O 7 11 6 N = 17
ro(X) .32 .67 .89 1.00 D = .67
Sn(X) .00 ,_.00 .6“ 1.00 p < .01
/roC<)-Snix)/ .32 .67 .2h .00
Adult No. 2 Ann Hunt L du F 1952
Course of action Ch C3 C2 C1
f l O 17 8 N = 26
FO(X) .32 067 089 1.00 D = 063
Sn(X) .04 .ogf .69 1.00 p < .01
/ro(x) - Sn(x)/ .28‘ .63 .20 .00

/
Adult No. 5 Old Mandamin L du F 1952
Course of action Ch C3 C2 C1
r 0 2 6 h N = 12
Fo(X) .32 .67 .89 1.00 D = .50
Sn(X) .00 .11, .6Z__ 1.00 p < .01
/ro(X) — Sn(x)/ .32 .50 .22 .00
/

Adult No. 6 Old Annie Mandamin L du F 1952
Course of action Ch C3 C2 C1
f 2 O 15 11 N = 28
F0(X) .32 .67 .89 1.00 D = .60
Sn(X) .07 .61 1.00 p < .01

 

.07
/ro(x) - Sn(X)/ '723’ .60 .28 .00

lllt

TABLE 5.2 -- continued

 

 

 

 

 

 

 

Adult No. 7 Rose Decker L du F 1952
Course of action Ch C3 C2 C1

f O .'O 2 5 N: 7
E0(X) .32 .67 .89 1.00 D = .67
Sn(k ) 0 0 .29_, 1.00 p 4 .01
/ro( (x) _ Sn(> <)/ .32 .67 .60 .00

Adult No. 15 Pauline Reed L du F 1966

Course of action Ch C3 C2 C1

f 0 0 37 8 N = h5
FO(X) 032 .67 .89 1.00 D = .67
Sn(X) .00 .00_, .82 1.00 p < .01
/ro(X) - Sn(X)/ .32 .67 .07 .00

Adult No. 20 Old Robinson L du E 1952,1966
Course of action Ch C3 C2 C1

f 0 25 1h N = ho
ro(X) .32 .67 .89 1.00 D = .6h
Sn(X) .03_ .65 1.00 p < .01
/ro(x) - Sn(X)/ .32 .6h .2h .00

Adult No. 21 Old Mrs. Robinson L du F 1952

Course of action Ch C3 C2 C1

r 0 8 23 N = 32
ro(X) .32 .67 .89 1.00 D = .6h
Sn(x ) 0 .93 .28 1.00 p < .01
/FO(X) "' Sn(/<)/ 032 .65 06]: 0

Adult No. 22 Young Robinson L du F 1952

Course of action Ch C3 C2 C1

f O O 1 5 N: 6
Fo(X) .32 .67 .89 1.00 D = .72
Sn(X) 0 Q__ .17 1.00 p < .01
/ro(X) - Sn(X)/ .32 .67 .72 .00

Adult No. 23 Eunice Robinson L du F 1952

Course of action Ch C3 C2 C1

f O 2 7 N: 9
ro(X) .32 .67 .89 1.00 D = .67
Sn(X) 0 _33 0 .22 1.00 p < .01
/.oc<)-5n(x)/ .32 .67 .67 .00

Adult No. 2h Arlene Robinson L du F 1966

Course of action Ch C3 C2 C1

r 0 0 3 h7 N = 50
Fo(X) .32 .67 .89 1.00 D = .83
Sn(X) 0 0 .06 1.00 p < .01
/ro(x) — Sn(x)/ .32 .67 .83 .00

115

 

 

 

 

 

 

 

 

TABLE 5.2 -— continued
Adult No. 25 Jean Robinson L du E 1966
Course of action Ch C3 C2 C1
f O O 7 22 N = 29
ro(> <) .32 .67 .89 1.00 D = .67
Sn(} <) 0 0 .2h 1.00 p < .01
/F0( (A) - Sn(/x)/ .32 .67 0'6; .00
Adult No. 31 Annie Waswagen L du F 1952
Course of action Ch C3 C2 C1
in" 1 O 3 8 N = 12
POLY) .32 067 089 1000 D = 05C 1
Sn(X) .08 .08 .33 1.00 p < .01
/ro(X) - Sn(X)/ .2h .59 .56 .00
Adult No. hO Jack Fitch L du F 1952
Course of action Ch C3 C2 C1
f 1 1h h2 N = 61
E0 (X) .32 .67 .89 1.00 D = .59
/ro(x) — Sn(X)/ .30 .59 .58 .00
Adult No. hl Betty Fitch L du E 1952,1966
Course of action Ch C3 C2 C1
f ' O 113 25 N = 70
ro(X) .32 .67 .89 1.00 D = .6h
Sn(X) 0 .6h 1.00 p < .01
/ro(>. ) - Sn(><)/ .32 .6h .25*’ .00
Adult No. h2 Old Mary Fitch L du F 1952
Course of action Ch C3 C2 C1
r 0 0 0 3 N =
Fo(X) .32 .67 .89 1.00 D = .89
Sn(X) 0 0 0 1.00 p < .01
,6e(x).-Suhx)/ .32 .67 .89 .00
Adult No. h3 Ben Ellis L du F 1966
Course of action Ch C3 C2 C1
f 1 8 h N = 1h
ro(X) .32 .67 .89 1.00 D = .53
SH(X) OO'Z olIL .71 _._1.00 p < 001
/ro(X) — Sn(X)/ .25' .53 .f8 .00
Adult No. 51 Marion Marshall L du E 1952
Course of action Ch C3 2 C1
r 1 1 1h- 19 N = 65
Fo(X) .32 .67 .89 1.00 D 2 .6h-
Sn(X ) .02 .03 .71 1.09 < .01
/Fo(} ) - Sn(><)/ .30 .611, .18 .09

 

116

TABLE 5.2 -- continued

 

 

 

 

 

 

 

 

Adult No. 63 Frank Mukkwa B.R. 1952
Course of action Ch C3 C2 C1
f 0 5 7 N
Fo(X) .32 .67 .89 1.00 D
Sn(X) O Ow .h-2__A1.QO p
/Do(X) — Sn(X)/ .32 .67 :h7 .00
Adult No. 6h Maggie Mukkwa B.R. 1952
Course of action Ch C3 C2 C1
1“ O O .11 19 N
Fo(X) .32 .67 .89 1.00 D
Sn(X) 0 _* 0 .17 1.00 p
/Fo(X) - Sn(X)/ .32 .67 .72 .00
Adult No. 70 Young Mukoman B.R. 1952
Course of action Ch C3 C2 Cl
1‘ O 1 3 1 N
Fo(X) .32 .67 .89 1.00 D
Sn(X) 0 .20 .80 l.OQ__p
/Fo(X) — Sn(X)/ .32 .L7 .09 .00
Adult No. 71 Mary Mukoman B.R. 1952
Course of action Ch C3 C2 C1
f 1 0 0 13 N
m( <) .32 .67 .89 1.00 D
n(X ) .07_ .07 .073 1.00 p
/ro(X) — Sn(X)/ .25“ .60 .82 .00
V/ I
Adultiku 76 lungaii Sign) B.R. 1952
Course of action Ch C3 C2 C1
f O O 6 1h N
Fo(X) .3 .67 .89 1.00 D
Sn(X) 0 0 ._30 1.00 p
/Fo(X) —'Sn(X)/ .32 .67 .59 .00
Adult No. 80 Wasse B.R. ,1952
Course of action Ch C3 C2 C1
1‘ O O 1 2 N
Fo(X) ' .32 .67 .89 1.00 D
Sn(X) O __ O 433 1.00 p
/Fo(X) - Sn(X)/ .32 .67 .56 .00
Adult No. 81 Mrs. Wasse B.R. 1952
Course of action Ch C3 C2 Cl
1‘ O ’O 2 11 N
Fo(X) .32 .67 .30 1.00 D
Sn (X) 0 ' O .1:_ 1.00 p
/FO( ()<) "' Sn(X)/ .32. 067 0?.“ .00

All || A II II A II II M H II A H II A ll ll
LU

All ll

12
.67
.01

23
.72
.01

U1

0 0
H1

ova

1h
82
.01

2O
.67
.01

.67
.10

13
'9‘ 71!
. 01

117

TABLE 5.2 -- continued

 

 

 

 

 

Adult No. 101 Agnes Clauson L du F 1966
Course of action Ch C3 C2 C1

f 2 O 10 8 N = 20
ro(x) .32 .67 .89 1.00 D = .57
Sn(X) .10 .1o .60 1.00 p < .01
/ro(x) - Sn(x)/ .22 .37 .29 .00

Adult No. 121 Mary St. Louis L du F 1966
Course of action Ch C3 C2 C1

r o o 11 1h N = 25
ro(x) .32 .67 .89 1.00 D = .67
Sn(X) O O .hfll 1.00 p < .01
,fl%mx) -EhMX)/ .32 .67 .h5 .00

Adult No. 990 Mrs. Schmidt L du F 1952
Course of action Ch C3 C2 C1

f O O S O N = 5
Fo(X) .32 .67 .89 1.00 D = .67
Sn(X) _;_o ' O 1.00 1.00 p < .01
/F0 (X) " 8n(>()/ 032 067 all .60

Adult No. 992 Mrs. Johnson L du F 1952
Course of action Ch C3 C2 C1

f O O S O N = 5
ro(x) .32 .67 .89 1.00 D = .67
Sn(X) o o 1.00 1.00 p < .01
/¢n(x)-Sn(x)/ .32 .67 .11 .00

Table 5.2 confirms that inference made from the
composite data presented in Table 5.1. The null hypo-
thesis derived from the "fear of others" theory can be
rejected at the .01 level for twenty-seven of the
twenty-nine adults. For the other two adults, the
probability that the null hypothesis is correct is equal

to .15 in one case and less than .10 in the other.

Types of ounishment observed

The punishments which were administered tended to

118
be mild as well as infrequent. The following incidents
reoresent fiveof‘the first six incidents of punishment
(in order of coding, not of observation) in my notes.
The remaining incident of the first six is quoted in

the scoring examnles in Chaoter IV.

D is a girl of six, Annie is D's grandmother.

Annie Waswagen at treadle-operated sewing maching.

D began to move the belt of the sewing machine up
and down about a half inch. Annie said, "Get away
from there," in a rather loud growl, turning her
head toward D rapidly as she growled. D drew back,
startled, but did not move away from sewing machine.
A little later she began to touch the belt lightly
with her forefinger. No reSponse from.Annie.
(Flambeau 52) ~

Since the growl was enough to startle D, it was

considered to be a ounishment.

B is a boy of six. G is a olaymate. Marion Marshall
is B's mother. Marion baking bread. The two boys
came into the house. B said G wanted a drink.

Marion launched into a tirade about "...Didn't he
have any water at his own houseY..." Both boys
listened without saying anything or moving and with
blank faces. B went and got himself a drinl. G
remained standing near door. (Flambeau 52)

D and B are brothers. D is eight and E is four. H

is their sister, five. Maggie Mukoman is their mother.
D and E sitting on ground near path. I sat down

near them, we talked a little. H came out of

house and screamed at them to do something I couldn't
understand. D screamed back and remained where he

119

was. After a few minutes, Maggie cam out, walked
down to us, talked rather loudly to the boys (my

Ojibwa was not good enough to understand it), and
led them off up to the house. (This was counted

as a punishment for each boy.) (Berens 52)

This is the same household as in the oreceding
incident, two weeks later. R is a boy of six and
Young Mukoman is his father. The children have
been unusually quarrelsome today. Young Mukoman
tells R to bring the ax. R just wanders around
"in a fog." Ygung Mukoman Speaks sharply to R
using word, "maja' ("get going"). R still wan-
dering in ”fog”. Young Mukoman cuffs him on
side of head, twice. R goes out door and comes
back with ax which Young Mukoman begins to
sharaen. (Berens 52)

This is the only incident in which I observed a

Child actually to receive a blow from an adult.

P is a girl of two. She is, as usual, staying at
her grandmother's while her mother works. Ben is

a boarder at the grandmother's house. P has been
running through the house. Ben has repeatedly told
her to stop, without success. Finally, he picks
her uo and outs her on a chair. She stays put.
(Flambeau 66)

Summarygof results for "fear of others" theory

These data offer no support for the "fear of
others" theory. The theory does not provide any basis
for the correct prediction of the behavior of even one
adult of this sample in the critical situation. Adult

Ojibwa rarely punish offending children even though the

children lack the capacity to retaliate.

120
The "fear of self" theony

Perhaos the alternative theory, which has been
called the ”fear of self" theory, will provide better
predictions of the behaviorof the sample adults in the
critical situation.

Briefly and informally stated, the "fear of self"
theory asserts that the typical Ojibwa refrains from
directing or attacking another Ojibwa who is annoying him
because he fears that to do so may do the other serious
injury (physical or psychological). If this theory is
correct, and if it is assumed that the Ojibwa believe
children to have the same vulnerability to direction and
attack as they believe adults to have, then adult Ojibwa
should treat children in very much the same manner in
which they treat other adults. From a formal decision
model of the typical Ojibwa as this simale theory assumes
him to be, it was calculated that the tywical Ojibwa,
when annoyed by another would choose Cl (neither direct
nor attack) with a relative frequency of .91; that he
would choose C2 (direct but not attack) with a relative
frequency of .Oh; that he would choose C3 (not direct
but attack) with a relative frequency of.O3; and that
he would choose Ch (both direct and attack) with a
relative frequency of .02.

The same data which were used to test the "fear

121
of others" theory will be used to test the "fear of self"

theory.*

Observed resoom:es: comoosite treatment

Of the 661 resoonses of these twenty-seven adults,
350 (.53) were c1, 288 (.hh) were C2, 13 (.02) were C3,
and lO-(.O2) were Ch. (Relative frequencies do not add
to 1.00 because of rounding.) Table 5.3 compared this
observed distribution with the theoretical distribution

deduced from the "fear of self" theory.

TABLE 5.3

DISTRIBUTION OF OBSERVED RESPONSES OF TWENTYQSEVEN OJIBUA
.ADULTS COMPARED WITH THEORETICAL DISTRIBUTION FROM "FEAR
OF SELF" THEORY

Course of Action
‘ 'cu C3 c2 c1
Theoretical frequency .02 .03 .0h .91
Observed frequency .02 .02 .hh. .V3

*The data for adults 990 and 992 are ommited here-
after. These adults were not members of the households of
any of the children involved in the study. Because of
the rule adopted earlier (Ch IV) that if more than one
adult was oresent, only the reSjonse of the adult pri-
marily resoonsible for the child would be scored, it was
impossible to score Cl (neither direct nor attack) for
either of these adults. Since Cl is predicted by the "fear
of self" theory to be the most frequent resoonse of
Ojibwa to offending children, it would be misleading to
include the resoonses of adults for whom this score
could not be included.

122

Figure 5.1 presents the composite observed dis-
tribution of choices of course of action compared with
the theoretical distributions derived from both the
"fear of others" theory and the "fear of self" theory.

As figure 5.1 shows, the correspondence between
the observed relative frequencies of choice ani the
theoretical relative frequencies deduced from the
"fear of self” theory is quite close for Ch and C3. The
observations, however, depart rather widely from the
predictions for C2 and C1. The sample adults chose C2
much more frequently than the ”fear of self” theory
predicts anl they chose C1 corresgondingly less frequently
than predicted. Nevertheless the general form of the
observed distributions is similar to the form of the

!

distribution derived from the "fear of self' theory.

I? f‘ 1‘ 1‘"
1ear o1 sel1

.80 theoretical:

>. '70

o

c

Q.)

g. .60

0.)

L

{H

c.) '50

>

OH _

p "fee r of others”

,2 .hO theoretical

C) a

e: /
.30

.20 observed 3

.IO

 

I 3 2 1

Course of Action

FIGIJ. x: 5.].

DIS TRIBUTIOII or o S-ER '13:) arseomes oz? ADULTS To TEEI-IDII 1e

CHILDREI.I 1.'.AE1,D JIITH TIIE THEORETICAL DISTR BUT IOIIS EROIJI
THE "FEAR OE OTHERS" THEOR AIID THE "IE OE SELE" THEORY

Observed resaonse : individual treatment

H

As in the case of the fear of others" theory,

I!

it is incividual behavior which the "fear of self theory

—.

12

K

1
purports to predict. Table S.h presents the data and a
K-S test of significance of the deviations from the

theoretical frequency for each adult.

TABLE S.h
OBSBAVED RESPONSES OE ADULTS T3 EEENDING CHILDREN AND
TEST CE IIYEOTI‘IES IS EROILI "EEAR OE SELE" TI‘IEORY E' R EACH
OE TWENTY—SEVEN ADULTS

f = observed number of choices of each course of action

 

 

 

 

 

Fo(X) = theoretical cumulative distribution of choices
under Ho from ”fear of self” theory

Sn(X) = cumulative distribution of observed choices

D = maximum /Eo(X) — Sn(X)/

p = probability associated with D under Ho

Adult No. 1 John Hunt L du F 1952

Course of action Ch C3 C2 C1

f O O 11 6 N =

E0(X) .02 .05 .09 1.00 D =

Sn(X) .oo .09, .65 1.00 p <

/Eo(X) —_Sn(x)/ .02 .05’ .56 .00

Adult No. 2 Ann Hunt L du F 1952

Course of action Ch C3 C2 C1

f I O 17 8 N =

Sn(X) .Oh .0h .69 1.00 p <

/Eo(X) - Sn(X)/ .02 .01 .60 .00

Adult No. 5 Old Mandémin L du E 1952

Course of action Ch C3 C2 C1

r o ' 2 6 h N =

E0(X) .02 .05 .09 1.00 D =

Sn(X) .oo .17, .6 1.00 p <

/FO ;<) "' Sn(X)/ 0.02 o 2 OS 00.5

17
.56
.01

26
.60
.Ol

12
.Ol

125

TABLE 5.h -— continued

Adult No. 6

 

 

 

 

 

 

 

 

Old Annie Mandemin L du r 1952

Course of action Ch C3 C2 C1
r 2 o 15 11 N = 28
ro(x ) .02 .05 .09 1.00 D = .52
Sn(X) .oz_ .91, .61 1.99 p < .01
/r 0(X) — Sn (x)/ .65' .02 .52 .00
Adult No 7 Rose Decker L du E 1952
Course of action (11 C3 C2 C1
r ' o ' 2 5 N = 7
I“0(X) .02 .05 .09 1.00 D = .20

n(3<) .00 .00 .29 1.00 p .20
/Eo(X ) - Sn(X)/ .02 .05’ .20 .00
Adult No. 15 Pauline Reed L du F 1966
Course of action Ch C3 C2 C1
f 0 0 37 8 N = IE5
?0(X) .02 .05 00C) 1000 D = .73
Snco v.00 .oo .82 1.00 p <. 01
/: 01x) - Sn(X)/ .02 .05’ .73 .00
Adult No. 20 Old Robinson L du F 1952,1966
Course of action Ch C3 C2 C1
f O l 25 1h N = hO
Eo(X) .02 .05 .09 1.00 D = .56
Sn(x) .oo .03 .65 1.00 p < .01
/r6(x) - Sn(X)/ .02 .02 .56 .00
Adult No. 21 Old Mrs. Robinson L du E 1952
Course of action Ch C3 C2 C1
E 0 l 8 23 N = 32
Sn(X) .00 .Qfig .28 1.00 p § .20
/Eo(X) - Sn(X)/ .02 .02 .19 .00
Adult No. 22 ' Young Robinson L du F 1952,1966
Course of action Ch C3 C2 C1
f 0 0 l 5 I = 6
Eo(X) .02 .05 .09 1.00 D = .08
Sn(X) .00 .00 3.17 1.00 p > .20
/To(X) — Sn(X) .02 .05 .08 .00
Adult No. 23 Eunice Robinson L du F 1952
Course of action Ch C3 C2 C1
f o o 2 7 N = 9
Eo(X) .02 .05 .09 1. 00 D = .13
Sn(X) .oo .oo .22 1.0a D > .20
/:-‘o(><) - Sn(:<)/ .02 .05 .13 .39

l

26

TABLE 5.h -- eontinued

Adult
Course of

p
.L

F0(3<)
/F0(\ ’) - Sn(k ’I/

Adult No. 25
Course of action
f

1‘0 (X)

Sn(X)

/Eo(X) - Sn(X)/

Course of action
f

FIG 0 211'-
action

Eo (X)

3n(x)

_/Eo(X) — Sn(X)/
Adult No . 110
Course of action
1% (x)

Sn(X)

/E0(X) - Sn(> ’)/
Adult No. h1
Course of action
f

Eo(X)

$3n(}<)

/Eo — Sn(X)/

Course of action
f
E0(> ’)

(\f

Sn(
/Eo(X) — SnIA ’I/

Adul t No 11-3
Course of action
f

Ro(x)

Sn(X)

/Eo(X) - Sn(X)/

 

 

 

 

 

 

 

 

Arlene Robinson L du F 1966
Ch C3 C2 C1
0 0 3 h? N
.02 .05 .09 1.00 D
000 __ .00 .06 1000 p
.02 .05 .03 .00
Jean Robinson L du E 1966
Ch C3 C2 C1
0 0 '7 22 N
.02 .05 .09 1.00 D
.00 .00 .2h 1.00 p
.02 .05“ .15“ .00
Annie Waswagen L du E 1952
Ch C3 C2 C1
1 o 3 ' 8 N
.02 .05 .09 1.00 D
L98 .08 .33 1.00 p
.06 .03 .2h .00
Jack Fitch L du F 1952
Ch C3 C2 C1
1 1h h2 N
.02 .05 .09 1.00 D
.02 .08 .31g, 1.00 p
.00 .03 .22 .00
Betty Fitch L du r 1952,
Ch C3 C2 C1
0 2 E3 25 N
.02 .05 .09 1.00 D
.00 .03 .6h 1.00 p
.02 .02 .55 .00
Old Mary Fitch L du r 1952
Ch C3 C2 C1
' o o o 3 N
.02 .05 .09 1.00 D
.og__ .oo .oo 1.00 p
.02 .05“ .09 1.00
Ben Ellis L du E 1966
Ch C3 C2 C1
1 1 8 h N
.02 .05 .09 1.00 D
.0? .11; .11 1.00 p
.05“ .09 .62 .Do

All 1| H V II II V II II VII ll V11 II

VI! ll

.Alln

SO
.20

.15
.20

12
.2h
.20

61
.22
.05

966

70
.01

.09
.20

1h
.62

127

TABLE 5.h -- continued

 

 

 

 

 

 

 

Adult No. 51 Marion Marshall L du F 1952
Course of action C2 C3 C2 Cl
r 1 1 DD- 19 N = 6?
:o(}< ) .02 .05 .09 1.00 D = .é2
X) .02 .03 ___ 311 1.00 o < .01
S/:I;l(3( (X) "' .SH(X )/ .00 .027 062 .00
Adult No. 101 Agnes Clauson L du F 1966
Course of action Cu C3 C2 C1
f 2 O ’10 8 N = 20
Fo(X) .02 .05 .09 1.00 D = .51
Sn(X) .10 .10 .60 1.00 p < .01
/ro(x) - Sn(X)/ .08 .055 .51’ .00
Adult No. 121 Mary St. Louis L du E 1966
Course of action C2 C3 C2 C1
f O ' 0 11 lb N = 25
Fo(X) .02 .05 .09 1. 00 D = .35
Sn(X) .oo .00 .11b 1. 00 p < .01
/FO (A) "" Sn(x )/ 002 005 033 000
Adult No. 63 Frank Mukkwa B.R. 1952
Course of action Cu C3 C2 C1
f O O 5 7 N = 12
Eo(X) .02 .05 .09 1.00 D = .33
Sn(X) .00 .00 .b2 1.00 p §=.lO
/?o(x) — Sn(X)/ .o2 .oS’ .33 1.00
Adult No 62 Maggie Mukkwa B.R. 1952
Course of action CD C3 C2 C1
f o o u 19 N = 23
Eo(X) .02 .05 .09 1.00 D = .08
Sn(X) .00 .00 .11 1.00 p > 020
“0(X) - Sn(X)/ .02 .DS* “:08 .00
Adult No. 70 Young Mukoman B.R. 1952
Course of action C2 C3 C2 C1
r ' o '1 "3 -1 N = 5
Eo(X) .02 .05 .09 1.00 D = .71
Sn(X) .oo .17 .80 1.00 p < .01
/%o(x) - Sn(X)/ .02 .12 .71 .00
Adult No. 71 Mary Mukoman B.R. 1952
Course of action C2 C3 C2 C1
r 1 o o 13 11 = 111
0(X) .02 .05 .09 1.00 D = .05
p > .23

 

sn(x) .0; .01_ .07 1.33
/"o(X ) - Sn(X)/ .09 .02 22 ‘

..J 1.00

Adult No. 76
Course of
f

ro(x)
Cn(/<)

/?0(X )-gSn(k ’)/

Adult No. 80
Course of action

X) - Sn()< )/

ifldult IR). 81
Course of action

Summary of results

128

action CL

 

 

 

 

TABLE 5.) _g -- continued
Abigail Eiéip B.R. 1952
C3 C2 Cl
0 O 6 111 N = 20
.02 .05 .09 1.00 D = .21
.oo .oo .30 1_go p > .20
002 .057 021- 000
Vbsse B.R. 1952
on C3 c2 Cl
0 ‘O l 2 N =
002 .0; .09 1.00 D = QZIL.
'oo .00 .33_ _~»1.oo p > .20
.02 .05? .2h . 00
Mrs. Wasse B.R. 1952
on C3 C2 Cl
0 o 2 11 N = 13
;02 .05 .09 1.00 D = .06
.oo .oo .15; 1.00 p > .20
.02 COSr— #:06 000 '

for "fear of self" theory

 

The individual data of Table 5.h tend to confirm

the conclusions

data with the two hypothetical distributions:

from a perfect fit,

from the comparison of

the composite

although far

the hypothetical distribution from

the "fear of self" theory is a better fit to the data

than the hynothetical distribution from the "fear 01

others" theory.

theory can be rejected at the

The hyoothesis from the

"fear of self"

.01 level for twelve of

twenty-seven individuals where the hyoothesis from the

"fear of others"

for twenty—seven of twenty-nine individuals.

teen of the twenty—seven individuals there is at

theory can be rejected at this

level

For thir-

(’i‘

lees

W

 

129
one chance in five that the observed sample of their
behavior could be a random sample from a population of
resoonses having the distribution deduced from the "fear
of self" theory. There were no cases of similar fit be-
tween the individual data and the "fear of others"

theory.

Separate treatment of the Berens River and Lac du
Flambeau samples

 

 

Some additional evidence in support of the "fear
of self" theory is provided by separating the data for
the Berens River peOple from the data for the Lac du
Flambeau peOple.

If the observations of the Berens River sample
are separated from those of the Lac du Flambeau sample,
we find that the observed frequencies of choice for the
Berens River sample corresoond much more closely to the
frequencies deduced from the "fear of self" theory than
do the observed frequencies for the Lac du Flambeau

sample.

Empirical modifications of the :fear of self" theory

 

The comparatively close correSpondence between
the data and the hypothetical distribution deduced from
file "fear of self" theory suggests that it might be

possible to modify this theory so that the observed

130
frequencies of choice can be deduced from it. The pro-
cess of modification begins with the observation that
the major differences between the observed relative fre—
quencies anl the hypothetical relative frequencies occur
in the greater observed relative frequency of choice of
C2 and the correSpondingly lower observed frequency of
choice of Cl.

2 Within the decision-theory framework an individ-
ual's probability of choice of a givenczourse of action
is taken to be a function of the expected value of that
course of action relative to the expected values of the
other courses of action. Thus, we account for the rel-
atively greater observed than predicted frequency of
choice of C2 by hypothesizing that C2 has higher expected
value for the typical Ojibwa than was originally assumed.
But, expected value depends on both the values assigned
to the compound outcomes and on the subjective estimates
of the probabilities of the various outcomes if C2 is
chosen. The values of the outcomes and the subjective
probability estimates were both assumed. The particular
figures used being chosen because they led to the
deduction of a frequency of choice of C2 which was
consistent with the reports of the behavior of typical
Ojibwa and because they did not seem unreasonable.

The expected value of C2 could be altered by

l3l
altering the assumed values of the outcomes or by altering

the assumed subjective probabilities (or both).

Rationale for modification of subjective probabilities

This section offers a Speculative rationale for
the hypothesis that Indian adults at higher levels of
acculturation regard directing as less dangerous to
children. The argument is basedcn the differences in the
frequency and intensity of Indian interaction with whites
at the two locations studied -- differences which roughly
parallel the differences in frequency with which adults
direct children.

At Lac du Flambeau there is a permanent population
of whites which is about equal in size to the Indian
population. Most of the permanent white residents are
owners of small, independent businesses and quite middle—
class in orientation. In the summer, the white pop—
ualtion is greatly increased by two categories of summer
people. Tge first category consists of people who own
summer homes on the reservation ani who are present enough
to get to know some of the Indian peOple. The second
category consists of people who are presentibr very
brief periods and who may never return again -- peOple
who stop at the numerous resorts and motels in the area.

Both of these roups of course have UldCIC~CIBSS
9 1 : .9

132
values.

The whites who are permanent residents exercise
considerabk> influence on the Indians. This influence
arises in part because of the Special status that the
northern Algonkians have apparently always given to the
white man. The attitude of dependence toward whites
with its corollary hOpe of receiving something from the
white man has resulted in their giving considerable
deference to whites. (See Barnouw 1963:152.) In
addition, many of the whites with whom they interact
most frequently actually do have power over the Indians.
The teachers and administrators of the school have
definite ideas on parental control of children which
they do not hesitate to eXpress to Indian parents,
they and the school board have been influential in getting

the tribal council to pass an ordinance making parents
reSponsible for the school attendance of their children.
This was, in 1952, the only tribal Ordinance which was
fully enforced.

The Indians are also in frequent and significant
interaction with the local merchants upon whom they are
realistically dependent for credit. These merchants also
give Indians unwanted advice on what they regard as
"moral” issues. The Indians have no trouble getting

the message that whites exnect Indians to control their

133
children in the same fashion that whites do.

The Indians have frequent opportunities to see
white parents constantly telling their children to do
this or that or not to do some other thing. These
occasions arise almost daily in the stores, the post
office, the laundromat and at school functions which
are well attended by both whites and Indians. my wife
and I were esoecially impressed with the frequency with
vhich white parents direct their own children in such
places in contrast to the Indians at Flambeau.

In any case, the Indians at Flambeau seem to
have gotten the idea that white people think that Indians
should do more directing of their children. They do
direct them more frequently than the Berens River people
do theirs, but less frequently than do whites. They may
have gotten the idea that directing is less dangerous
than they thought partly from observing the absence of
damage to white children. They must also have come to
believe it is less dangerous from having done quite a
lot of directing without disaster. It should be empha-
sized, however, that Indian parents seldom "sound as
if they mean it" when directing children.

Many of the same factors Operate at Berens
River, but to a much lesser extent. In particular, there

are fewer opportunities for the Indians to bbserve

13h
white parents and children in interaction and fewer

whites to tell the Indians to control their children.

Changes in the assumptions

following this reasoning, I have assumed that
file subjective probability of injuring a child by choice
of C2 is lower than originally assumed and that it is
lower for the Indians at Lac du Flambeau than it is for
the Berens River Indians. It was possible to find
subjective probabilities for each group which led to the
deduction of theoretical probabilities of choice of C2
very close to the composite observed probabilities for

the peOple of each location%.

Data and the modified theory: the post hoc nature of
these tests

Properly, each of these modifications of the
"fear of self" theory should be tested with data not
used in arriving at these modifications. Since this
was not practicable at this time, the original data
are being reused. The use of this data thus makes these

tests of the modified theory completely post hoc in

*These changes in the assumptions and the resulting
changes in the theoretical probabilities of choice are
presented in detail in the appendix.

_ 135
nature. Hopefully, at some time in the future, it will
be possible to make additional observations at Berens

River and Lac du Flambeau to test these modifications.

Data_and the modified theory: composite treatment
Table 5.5 compares the composite observed
relative frequencies for the Berens River sample with
the modified theoretical frequencies which result from
the changes in the subjective probabilities assumed for

the Berens River people.

TABLE 5.5

COMPARISON OE OBSERVED DISTRIBUTION OE CHOICES OE BERENS
RIVER ADULTS WITH MODIFIED THEORETICAL DISTRIBUTION

Courses of Action
Ch C3 C2 Cl

Modified theoretical .017 .023 .25 .71
distribution
Observed distribution .011 .011 .23 .7h
TABLE 5.6

COMPARISON OF OBSERVED DISTRIBUTION OE CHOICES OE LAC
DU ELAHBEAU ADULTS WITH MODIFIED {EORETICAL DISTRIBUTION

Courses of Action
Ch C3 C2 Cl

Modified theoretical .012 .016 .h? .50
distribution
Observed distribution .016 .021 .h? .50

136

Data and the modified theory: individual treatment

Table 5.? presents the individual data for the

c‘

seven Berens River adults and a K-o test of each of

the individual variations from the "fear of self" theory

as modified for Berens River.

TABLE 5.7
OBSE ED RESPONSES OE.ADULTS TO OFFENDING CHILDREN AND
TESTP OE HYPOTHESIS ERONIBERENS RIVER MODIFICATION OE
"FEAR OE SELE" THEORY FOR EACH OE SEVEN.ADULTS
f = observed number of choices of each course of action
Fo(X) = theoretical cumulative distribution of choices
under hypothesis from Berens River modification
of "fear of self" theory
Sn(X) = cumulative distribution of observed choices

- Sn(X )/

p = probability associated with D under Ho

D maximum /Eo(X)

 

 

Adult No. 63 Frank Mukkwa Berens River 1952
Course of action Ch C3 C2 C1

E O O S 7 N = 12
ro(x) .02 .ok .2~ 1.00 = .13
Sn(X) .oo .oo .12 1.00 p > .20
/Eo(X) - Sn(X)/ .02 .ou .13 .00

Adult No. 6h Maggie Mukkwa B.R 1952

Course of action Ch C3 C2 Cl

1 o u 19 N = 23
Fo(X) .02 .01 .29 1.00 D = .12
Sn(X) .oo .oo .17 1.00 p > .20
/?o(X) - Sn<>< )/ .02 .0b .12 .oo

Adult No. 70
Course of action

/F0(X) - 3n ><)/

Course of action

H)

Eo(X)
Sn(X)

/Eo(X) - Sn(X)/

Adult No. 76

Course of action
f

W(>)

«>1
/ro(x) - Sn(X)/

Adult No. 80
Course of action

1
F010

Sn(X)

/ro(x) — Sn(X)/

Adult No. 81

Course of action
f

E‘o (Ei)

811(‘()

/ro(x) — Sn(X)/

137

 

 

 

 

 

 

 

Young Xukoman B.R. 1952
Ch C3 C2 C1
0 3 1 N
.02 .ou .29 1.00 0
.00 .20 .80 1.00 p
.02 .16 .SI . 00
Mary Mukoman B.R. 1952
Ch C3 C2 C1
1 0 0 13 N
.02 .ou .29 1.00 p
.02 3,07 .91 1.00 p
.037 .03 .22 .OO
Abigail Sisip 3.3. 1952
Ch C3 C2 C1
0 6 1h N
.02 .0? .29 1.00 D
.00 .00 .30 1.00 p
.02 .01 .01 .00
Wasse B.R. 1952
Ch C3 C2 C1
O l 2 N
.02 .0b .2— 1.00 D
.00 .00 .33 1.00 p
.02 .Oh .0E .00
Mrs. Wasse B.R. 195
Ch C3 C2 C1
0 O 2 11 N
.02 .0b .29 1.00 D
.00 .00 .15;_ 1.00 p
.02 .0h .1h .00

V II H V 1! H V H H H H II

VI! ll

23
.51
.10

IS
.22
.20

3
.0h

.20

13
.IM
:20

The null hypothesis from the”fear of self" theory can be

rejected,

only one of these seven adults.

That is,

even at the .20 level of significance, for

there is a

probability greater than one in five for six adults that

the sample of his or her behavior came from a population

of reSponses having the relative frequencies predicted

by the "fear of self" theory as modified for Berens River.

138

Table 5.8 presents similar information for the
Lac du Flambeau sample. In five cases, the null hypo-
thesis derived from the "fear of self" theory as modified
for Lac du Flambeau can be rejected at the .01 level
of significance.* In two cases, the probability that
the null hypothesis is correct is approximately.lO.
In the remaining thirteen cases, the probability for
each case that the sample of behavior came from a
population having the distribution of responses deduced
from the Lac du Flambeau modification of the "fear of
self” theory is greater than one in five.

TABLE‘S.8

OBSERVED RESPONSES OF ADULTS TO OFZTET'IDING CHILDREN AND
TEST OF HYPOTHESIS FROM LAC DU FLAMBEAU MODIFICATION 0?
"FEAR 0? SELF" THEORY

f = observed number of choices of each course of action

 

Fo(X) = theoretical cumulative distribution of choices
under hyoothesis from Lac du Flambeau modi-
fication of ”fear of self" theory

Sn(X) = cumulative distribution of observed Choices

D = maximum /%o(x) — Sn(X)/

p = probability

Adult No. 1 John Hunt L du F 1952

Course of action Ch C3 C2 C1

f O O 11 6 N = 17

Fo(X) .10 .03 .50 1.00 D = .15

Sn(X) .00 .00 .65 1.00 p > .20

/ro(x) — Sn(X)/ .01 .03 .15 .00

 

__

%One of these cases conforms quite closely to the
predictions of the original "fear of self" theory and
another conforms quite closely to the Berens River mod-
ification of the theory.

139
TABLE 5.8 -- continued

 

 

 

 

 

 

 

 

Adult No. 2 Ann Hunt L du F 1952

Course of action Cu C3 C2 C1

f 1 0 17 8 N = 26
Fo(X) .01 .03 .50 -1.oo D = .19
Sn(X) .0h .02 .69 1.00 p > .20
/Fo(x) - Sn( )/ .03 .01 .19 .00

Adult No. 5 Old Mandémin L du F 1952
Course of action Cu C3 C2 C1

f o 2 6 L N = 12
F0(X) .01 .03 .50 1.00 D = .17
Sn(X .00 .127 .67 1.00 p > .20
/?o (X) - Sn(X)/ .01 .TL .17 .00

Adult No. 6 Old Annie Mandamin L du F 1952
Course of action (H1 C3 C2 C1

f 2 0 15 11 N = 28
Fo(X) . 01 .03 .50 1.00 D = .11
Sn(X) .07 .07 .61 1.00 p 2 .20
Adult No. 7 Rose Decker L du F 1952

COurse of action Ch C3 C2 C1

f 0 0 2 5 N = 7
F0(X) .01 .03 .50 1.00 D = .21
Sn(X .oo .oo___ .29 1.00 p > .20
/Fo(x) - Sn(X)/ .01 .03 .21 .00

Adult No. 15 Pauline Reed L du F 1966

Course of action Cu C3 C2 C1

r o o 37 8 N = AS
Fo(X) .01 .03 .50 1.00 D = .32
Sn(X) .oo .oo .82 1.00 p < .01
/Fo(X) - Sn(X)/ .01 .03 ".32 .00

Adult No. 20 Old Robinson L du F 1952, 1966
Course of action Cu C3 C2 C1

f o 125 1g N = no
F0(X ) .01 .03 .50 1.00 D = .15
Sn( (X) .00 ‘ .03 .6 1.00 p > .20
/%o(A)— Sn(A)/ .01 .oo .15 .00

Adult No. 21 Old Mrs. Robinson L du F 1952

Course of action Ch C3 C2 C1

f ' o 1 8 23 N = 32
70(X) .01 .03 .50 1.00 D = .22
Sn(X) .oo .93 .28 1.0+ p a .10

 

/Fo(>< ) — Sn( <)/ .01 .oo .22 .oo

Course of action
f

Fo(X)

Sn(X)

/F0(}<) - Sn(X)/
Adult No. 23
Course of action
f

T30 (if)

En(X )

/.‘—“O(}’) - Sn(3\)/
Adult No 2h
Course of action
F0(X)

Sn(X

/F0(X) — Sn(A)/

Adult N0. 25

Course of action
f

0(X)

n(X)
/20( D\) -SH(X)/

Adult No. 31
Course of action

Adult No. LOd

Course of action
f

F0(X)

Sn(X)

/F0(X) - Sn(X)/’
Adult N0. Ll

Course of action

4‘
J.

F0(X)
Sn(> ’)
/F0(X) -Sn1(k ')/

1210

 

 

 

 

 

 

 

 

TABLE 5.8 --. 9.922.112.1122.
Young Robinson L du F 1952
CL C3 C2 C1
0 O l 5 N
.01.03 .50 1.00 D
.00 _» .00 .17 1.00 p
.01.03 .33 .00
Eunice Robinson L du F 1952
cu C3 02 c1
0 0 2 7 N
.01 .03 .50 1.00 D
.00 .00 29 1.90 p
.01 .03 .237 .00
Arlene Robinson L du F 1966
on C3 C2 Cl
0 O 3 D? N
.01 .03 .50 1. 00 D
.00 .004_ .06 J_ 1. 00 p
.01 .03 .ldl .00
Jean Robinson L du F 1966
on C3 c2 c1
0 0 7 2 N
.01 .03 .50 1.00 D
.00 .00 .2h 1.00 p
.01 .03 .26 .00
Annie Waswagen L du F 1952
on C3 C2 C1
1 0 3 8 N
.01 .03 .50 1.00 D
.08 .08 .33 1.00 p
.07 .05' .17 .00
Jack Fitch L du F 1952
CL 03 02 c1
1 R. 1b L2 N
.01 .03 .50 .00 D
.02 .08 .31 1.00 p
.01 .051 .19 .00
Betty Fitch L du F 1952 1
cu C3 02 c1
0 2 LB 25 N
.01 .03 .50 1.00 D
000 003 06L 1.0an D
.01 .00 .1u .00”

.33
.20

ann

9
.28
.20

V” ll

50

Lil”!

.01

All ll

29
.26
.01

All ll

12

.17
.20

VII ll

12

.19
.01

A” H

\O

66

70
.111

§:.10

II

II

1&1

TABLE 5.8 -- Continued

 

 

 

 

 

 

 

Adult NO. L2 Old Mary Fitch L du F 1952
Course of action CL C3 C2 C1
f 0 0 0 3 N = 3
F0(X) .01 .03 .50 1.00 D = .50
/F0(X) — Sn(X)/ .01 .03 .50* .00
Adult N0. L3 Ben Ellis L du F 1966
Course of action CL C3 C2 C1
r 1 1 8 u N = 1L
F0(X) .01 .03 .50 1.00 D = .21
Sn(X) .07 .12 _3.71 1.00 p > .20
/Fo(X) - Sn(A)/ .06” .11 .21 .00
Adult No. 51 Marion Harshall L du F 1952
Course of action CL C3 C2 C1
f ' 1 1 LN 19 N = 65
F0(X) .01 .03 .50 1.00 D = .21
Sn(X) .02 .03 .71 1.00 p < .01
/F0(X) — Sn(A)/ .01 .00 .21 .00
Adult No. 101 Agnes Clauson L du F 1966
Course of action CL C3 C2 C1
f 2 O 10 8 N = 20
F0(X) .01 .03 .50 1.00 D = .10
Sn(X) .10 .10__! .60 1.00 p > .20
/F0(X) - Sn(A)/ .09 .07 .10 :00
Adult No. 121 Mary St. Louis L du F 1966
Course of action CL C3 C2 C1
f 0 0 11 12 N = 25
F0 (:4) .01 .03 .50 1.00 D : .06
anco .00 .00 .LN 1.00 p‘> .20
/F0(X) — SnCO/ '701 .03 .06 .00
Summary Of results for modified ”fear of self" theory

A ve'y slight modification of the "fear of self"
theory is sufficient to permit rather accurate predic-
tions 0f the relative frequencies of choice of the four

courses of action for the two samples as

slightness of the modification which r

rr-‘Ifi
LO

frauds. The

required to

1112
produce this improved accuracy of prediction is taken

to support the "fear of self" theory.
The Learning of Emotional Restraint

Group I theory restate,

In its most common form, this theory involves the
implicit assumption that restraint is motivated by
the values and subjective probabilities of the "fear
of others” theory. It is assumed, following Mowrer,
that the subjective probability of severe retaliation
following an offensive act is equal to the objective
probability of retaliation in the individual's previous
eXperience. Thus the actual probability of punishment
for an offensive act should be equal to the subjective
probability of retaliation which is required to eXplain
Ojibwa restraint. From the formal model of th: typical
Ojibwa as postulated by the "fear of others" flleory,
it was determined that a subjective probability of
retaliation of .75 would be sufficient to eXplain
Ojibwa restraint within that theoretical framework.
Additionally, it was deduced that, if the individual were
extremely cautious (if he were a nearly perfect maxi-
mizer of expected value), a relative frequency of
punishment of .h5 would be adequate to eXplain the

J 1

reported restraint. The null hypothesis to be trs;rr is

1&3
that the Ojibwa child receives punishment for offensive

acts with a relative frequency equal to or greater than

.h5.

; Group I theory: composite data

From the finding that Ojibwa adults do not punish
.offending children with the frequency predicted by the
"fear of others"theory, it seems unlikely that children
will experience punishment with a relative frequency of
.L5, but the theory calls for the experiences which the

children have as individuals.

TABLE 5.9

OUTCOMES OF 767 SITUATIONS IN WHICH CHILDREN BEHAVED IN
A MAI-HER OFFEI‘B ME TO ADULTS

Child's behavior becomes Child's behavior does no

acceptable following become acceptable fol-
adult reSponse lowing adult reSponse
Adult reSponse Adult reSponse
‘ ClCl C2C2 C3C3 CLCL C1 C2 C3 Ch Totals
No. ' 5 135 9 10 338’ 223 5 O 767

5 0.65 17.6 1.2 1.3 29.5 29.2 .65 0 100.1

Table 5.9 presents the composite data for all
recorded instances of "misbehavior" by children. The
adult reSponses which involve punishment are C3 and CL.
The total number of such resoonses is 2h. The relative

frequency of punitive reSponses if .031 which is sig—

1h1t
nificantly below the hypothetical relative frequency,
.115. (x2 = 513.1, p<.001).

There is a strong tendency for adults to adjust to
the behavior of infants and children less than two years
of age, which suggests the possibility of different
reactions to such young children. The composite data

ommitting_the six children less than two years of age

 

are presented in Table 5.10.‘

TABLE 5.10
ourcorms 0? SITUATIONS IN mucn A CHILD BEHAves IN A
A1IF‘“ 0er1~13 IVE TO ADULTS. CHILDREN UNDER TWO YEARS
or AGE Ol.l‘.fITTED

Child's behavior becomes Child's behavior does not

acceptable following Become acceptable fol-
adult reSponse lowing adult reSponse
Adult rGSponse Adult reSponse
C1 C2 C3 Ch C1 C2 C3 Ch Totals
No. 3 133 8 10 270 216 1 O 6L1
% .5 20.7 1.2 1.6 L2.1 33.7 .2 0 100.0

The total number of punishments received is now
nineteen rather than twenty—four% and the total number

of incidents is reduced from 767 to 7hl. The relative

*One rejected little girl between one and two
years of age received the five punishments which make
U115 difference. She was the youngest member of the
Fitch household.

1&5
of punoshment is actually reduced to .030 by this change.
It should be emphasized again that most of these punish—

ments are mild by our standards.

Relative frequency of punishment at Berens River and at
Lac du Flambeau

 

Table 5.11 presents the data for all observations

at Lac du Flambeau and at Berens River separately.

TABLE 5.11

OUTCOIJES OF SITUATIONS IN ‘GJ'HICH A CHILD BENAVES IN A
MANNER OFFENSIVE TO ADULTS. COMPARISON OF ALL OBSER-
VATIONS AT BERENS RIVER WITH ALL OBSERVATIONS AT LAC
DU FLAT-BEAU

Child's behavior becomes Child's behavior does not

acceptable following become acceptable fol-
adult rGSponse lowing adult reaponse
Adult resoonse Adult response
C1 C2 C3 CL C1 C2 C3 Ch Totals
BR No. 0 12 1 2 ‘75 11 O O 101
% O 11.9 1.0 2.0 7L.3 10.9 0 0 100.1
L“ No. 5 123 8 8 "305 212 ‘5 0 666'
5 .8 18.6 1.2 1.2 L5.8 31.8 .8 0 100.2

There were twenty-one instances in which a child
received punishment in 666 observations at Lac du
Flambeau. This is a relative frequency 0f.0315. At
Berens River there were three instances of punishment

in 101 observations for a relative frequency of .0297.

1&6
This slight difference is far from significant.
X? = .0L3h, p)’.80. This finding that the frequency
with which children experience punishment is low and
nearly equal at both locations is consistent with the
earlier finding that the sample adults punish children

with nearly equal low frequencies in both places.

Individual data and tests of the Group I learning theogy
The data for each child and a binomial test of the
hypothesis from Group I theory for each child are pre—

sented in table 5.12.

TABLE 5.12

GROUP I LEARN NG THEORY. NUNFER OF OFFENSIVE ACTS,

PRWBER OF PUNITIVE MESPONSBS RECEIVED, RELATIVE FRE-
OUENCY OF PUNISHMENT, AND BINOMIAL PROBABILITY OF OBSERVED
NUMBER OF PUNITIVE RESPONSES UNDER THE NULL HYPOTHESIS,
FOR EACH OF THIRTYFTWO CHILDREP

N = number of offensive acts by child

A = number of times child received a punitive reSponse
from adult

X/N 2 relative frequency of punitive reSponses from
adult
p(x) = binomial probability of x or fewer punitive re—

Sponses under the null hypothesis: (x/nz .115).

Lac du Flambeau children

 

Child HH Name N 3 x n 0(X)
1 Hunt 11 1 .091 .01393

2 Hunt 10 0 .000 .0025_

‘ TABLE 5.12 -- continued

 

Child HH Name
3 Hunt

10 manaémin
ll Handgmin
12* mandémin
18 Reed

19 Reed

20 Clauson
21 Clauson
22 Clauson
26 Robinson
28 Robinson
29 Robinson
30 Robinson
31% Robinson
3 6 Via swa gen
N5 Fitch

L6 Fitch
b7A Fitch

5 Marshall
56 Harshall
60 St. Louis
61 St. Louis

1&7

1~ 0 C) no <3 0 C) U) f0 0 Ix

GOOD

1‘.)

 

x n
.000
-OHS
.103
.000
.000
.000
.1511
.000

.000

.000
.000
.000
.000
.0H2
.0L5
.000
.139

p(x)
.00001

 

.00000
.00007
.95033
.OOOOL
.00000
.02691
.00253
.01522
.00000
.00001
.00001
.00001
.00000
.00001
.OOOOL
.00000
.00008
.00002
.00000
.00056

.00900

 

AChild less than two years of age

11,")
1 .k)

.LL

TABLE 5.12 -- continued

 

Berens River children

 

Child HH Name N i x n 2151
65 Uukhwa 13 0 .000 .000h2
66 Uuhhwa 13 0 .000 .000L2
67 Lhkkwa 13 0 .000 .00022
82 Mukoman 6 1 .167 .1635?
Sufi Mukoman 10 0 .000 .00253
87 Wasse - 12 0 .000 .00077
9! §§§is 7 0 .000 .01522
95 §I§Ip 9 0 .000 .OOh61

The null hypothesis (that the relative frequency
of punishment is not significantly different from .h5)
can be rejected at the .05 level for 30 of the 32 child-
ren (93.75%). If the .01 level were used, it could still
be rejected for 26 of the 32 children (81.25%) for whom

5 or more observations were recorded.

Summary of results of the Group I theory

 

It is clear that the observed relative fre-
quency with which children experience punishment in
both samples is significantly below the hypothetical

minimum relative frequency (.h5) deduced from this

 

AChild less than two years of age

lh9
version of Group I theory. Neither the composite data
nor the individual data offer any support for the theory

that the emotional restraint of typical Ojibwa is

produced by frequency and severe punishment.

group II learning theory

This application of the theory rests on the
assumption that restraint is motivated by the values
and subjective probabilities of th: "fear of self”
theory. Mowrer's notion that subjective probabilities
come to match objective probabilities is retained in a
modified form. Rather than holding that the subjective
probabilities of a given outcome match the objective
probabilities of the £333 outcome, it is held that the
subjective probabilities of a given outcome (severe
injury to another individual) following aggression,
retch the objective probabilities of a different
outcome (acquiescence by the other individual) following
very limited aggression.

In short,this theory holds that the learner

makes less direct inferences from his eXperience, while

 

the other theory holds that the learner makes only the
most direct inferences from his experience.
From the formal model of the typical Ojibwa as

postulated by the "fear of self" theory, it was deduced

150
that, if the adults acquiesce in the minor offensive
behavior of the child with a relative frequency of
.75, this would be sufficient for the child to infer
that the probability of serious injury to another fol-
lowing less minor aggression is equal to .75.

An adult was considered to acquiesce in the
offenSive behavior of a child whenever the adult chose
a course of action which was not adequate to produce a
change to acceptable behavior on the part of the child.
Thus all of the instances listed in the right-hand

1‘

half of table 5.13 are instances 01 adult acquiescence
in the offensive behavior of the child since these are
the instances in which the "Child's behavior does not

become acceptable following the adult's reSponse".

TABLE 5.13%
OUTCOMES Ls SITUATIONS IN WHICH A CHILD BEHAVES IN.A
LATHER.0723NSIVD To ADULTS. ALL OBSERVATIONS AT LAC
DU 21.211.213.210 AND 131331121153 RIVER '

Child's behavior becomes Child's behavior does not

acceptable following become acceptable fol-
adult reSponse lowing adult resoonse
Adult response Adult response
C1 C2 C3 Ch C1 C2 C3 Ch Totals
No. 5 135 9 10 380 22 5 0 767
% .65 17.6 1.2 1.3 h9.5 29.2 .65 0 130.1

AThis table duplicates table 5.9 above.

151

Coggosite data and the Grouo II theory

As can be seen from Table 5.13, 608 of the
total of 767 situations had outcomes involving adult
acquiescence. The overall relative frequency of adult
acquiescence is thus .793, which is greater than the
minimum relative frequency (.75) required by the Group
II theory. The composite data are therefore consiStent
with the Group II theory.

Because of the tendency of Ojibwa adults to
adjust to the demands of children less than two years
of age, table 5.1h nresents the data for all the

children two years old or older.

TABLE S.lh*
OUTcoMES or SITUATIONS IN WHICH A CHILD BEHAVES IN.A
TIME-ER OFFEl-IS IVE To ADULTS. HILDREN UNDER Two YEARS
OF AGE 02.17::ITTED

Child's behavior becomes Child's behavior does ngt

acceotable following become acceptable fol—
adult resoonse lowing adult reSponse
Adult resoonse Adult reSponse
Cl C2 C3 Ch C1 C2 C3 Ch Total

rm. ‘3 133 8 10 270 216 1 0 6U1
% .5 20.7 1.2 1.6 h2.1 33.7 .2 0 100.0

There are h87 outcomes in which the adult ac-

 

fiThis table duplicates table 5.10 above

152
quiesced in the child's offensive behavior in the total
of 6hl incidents. This gives a relative frequency of
acquiescence of .760 which is Still above the theor-
etical minimum relative frequency (.75).
Perhaos a description of some incidents involving
acquiescence on the part of adults will clarify the

meaning of acquiescence. The following incident occurred

 

in the Robinson household in the summer of 1966.

Old Robinson is sitting in the yard whittling.
Three of his grand-daughters, aged six to eight,
and some other girls are playing nearby with a
blanket. A couple of them are under the blanket
staggering and crawling around while the others
are oulling and pushing the ones underneath. Old
Robinson says, somewhat crossly, "You'll tear that
blanket." The six-year-old answers flatly, "No,"
and the game continues uproariously. No further
reaction from Old Robinson.

The following incident is from my notes on the
household of Wasse, the medicine man from Little Grand
Raoids, the most culturally conservative household in
the samole. A16 is Wasse's youngest child, a boy of
four. Albis mother is sitting on the ground inside the
tent, but near the door. In the back part of the tent,

a group of girls is slaying with the infant son of

Alo's eighteen~year-old sister.

'Alo showed considerable interest in the baby,
too, but after a few minutes, he climbed into his

153

mother's lap and reached inside her dress. In
a few moments he was "nursing" his mother's
undoubtedly dry breast. The baby's mother came

in and began nursing the baby. A16 continued to

”nurse” as long as the baby did. AIE'S mother

made no attempt to prevent his "nursing", nor did

she encourage him.

Comnarison of Berens River and Lac du Flambeau

Because of the interesting differences in

”directiveness ” between Flambeau and Berens River adults
3

revealed by the examination of adult behavior it may be

significant to contrast childhood exoeriences at the
two locations. Table 5.15 presents the data for all

observations at the two locations separately.

TABLE 5.15
0UTCoIIES 0E SITUATIOHS IN wHICH A CHILD BEI—IAVES IN A

lmflNER OFFENSIVE TO ADULTS. COMLARISON OF ALL.OBSER-

VATIONS AT BERENS RIVER WITH.ALL OBSERVATIONS.AT LAC
DUFUMEMU

Child's behavior becomes Child's behavior does not
acceotable following become acceptable fol-
adult reSponse lowing adult resoonse
Adult resoonse Adult response
C1 C2 C3 Ch C1 C2 C3 Ch Total
BR No. 0 12' 1" 2" 75' 11' O O 101
% O L11.9 1.0 2.0 7h,3 10.9 0 0 100.1
LF No. ‘5 123 8 8 305 212 5 O 666
5:; 08 1806 102 1.2 L508 3108 .8 010002

15h

There were 666 observations at Lac du Flambeau.
In 522 of these, the child received acquiescent resoonses
from the adult. Children at Lac du Flambeau thus
exoerience adult acquiescence with a relative frequency
of .785. At Berens River, the adults acquiesced 86
times in the 101 cases observed. Berens River children
exoerience adult acquiescence with a relative frequency
of .851, which is slightly higher than the relative
frequency at Lac du Flambeau. Both relative frequencies
are clearly above the minimum required by the Group II
theory.

In the case of the children who were two or

more years of age, the basic data are given in table

5.16

TABLE 5.16

OUTCOMES OE SITUATIONS IN WICH A CHILD BEHAMES IN A
.AA’IER OFFENSE IVE TO ADULTS. COLIPARISOI‘J OE CHILDREI‘I TWO
YEARS OLD AI'xlD OLDER AT LAC DU FLAT-BEAU ‘I'JITH SIMILAR
CHILDREN AT BERENS RIVER

Child's behavior becomes Child's rehavior does not

acceotable following become acceptable fol-
adult resoonse lowing adult reSponse
Adult resoonse Adult response
C1 C2 C3 Ch C1 C2 C3 Ch Total
BR Ho. 0 12 1' 2' 51 8 0 0 7h
3 O 16.2 l.h 2.7 6559 10.8 0 0 100.0
L? No 3 121 7 8 :19 208 l 0 567
5'3 .5 21.3 1.2 l.h 38.6 36.7 ’3 ’) £95)

155

For the Lac du Flambeau children, there were
h28 instances of adult acquiescence in 567 obser-
vations -- a relative frequency of .755. This relative
frequency is still above the theoretical minimum, but
just barely. For the Berens River children, there were
59 instances of adult acquiescence in 7h observations --
a relative frequency of .797.

The data for the six children who were less than

two years of age are oresented in table 5.17.

TABLE 5.17

OUTCOLES OF SITUATIONS IN WHICH.A CHILD BENAVES IN A
LANNER OEEENSIVE TO ADULTS. CHILDREN LESS THAN TWO YEARS
OLD: BERENS RIVER AND LAC DU ELAMBEAU COMPARED

Child's behavior becomes Child's behavior does not

acceotable following become acceotable fol-

adult resoonse loving adult resyonse
C1 C2 C3 Ch C1 C2 C3 Ch Total

BR No. 0 O O 0 2h 3 O O 27
Z 0 0 0 0 88.9 11.1 0 0 100.0

B No. 2 2 l O 86 h h 0 99
Z 2.0 2.0 1.0 O 86.9 h.0 h.O O 99.9

All 27 of the indicents involving children less
than two observed at Berens River ended with parental

acquiescence. Ninety-four of 99 incidents at Lac du

156

Flambeau ended with adult acquiescence*%

Individual Data and Group II theory

 

More important than the agreement between the
composite data and the requirements of the Group II
theory is the answer to the question, "For how many
children does the relative frequency with which they
receive acquiescent reSponses from adults depart so
far from the hypothetical relative frequency that it is
reasonable to conclude that the observed frequency is
unlikely if the hypothesis were correct?". Table 5.18
gives the basic individual data and statistical tests
of the data from which this question can be answered.
The table covers all children for whom five or more

observations were recorded.

____- _——¥ __

%*The difference between the relative frequencies
of adult acquiescence to these very young children in
the two places may be misleading. All of the cases in
which the child conformed to adult demands involved the
rejected little girl who also received more than one-fifth
of all the punitive adult behavior observed. When the
data from this child are omitted, there is no difference
in adult acquiescence between the two groups.

 

157
TABLE 5.18

GROUP II LEARNING-THEORY. NUMBER OE OFFENSIVE ACTS,
PRHBER OE ACQUIESCENT RESPONSES RECEIVED, RELATIVE
FREQUENCY OE ACQUIESCENCE AND BINOMIAL PROBABILITY OE
OBSERVED NUHBER OE ACQUIESCENT RESPONSES UNDER THE
HULL HYROTHESIS, FOR EACH OE THIRTY-TWO CHILDREN.

N

number of offensive acts by child

number of times child received an acquiescent re-
Sponse from adult

X

x/H = relative frequency of acquiescent FGSQOHSCS from
adult

p(x) = binomial probability of x or fewer acquiescent
responses under the null hypothesis (x/nz .75)

Lac du Flambeau children

 

 

 

Child Household E. x x N 2131.
1 Hunt 11 h .36h .0077
2 Hunt 10 7 .700 .h77h
3 Hunt 20 6 .300 .0000
10 ‘lbndémin hh 35 .795 .8101
11 Mandémin 29 22 .959 .6179
12% MandBmin 5 5 1.000 1.0000
18 Reed 17 12 .706 .h299
19 Reed 22 18 .818 .8h02
2O Clauson l3 9‘ .692 .h191
21 Clauson 10 8 .800 .7582
22 Clauson 7 6 .857 .0678

 

———

*Child less than two years of age.

i'l’l

158

TABLE 5.18 -- continued

 

 

 

 

 

 

Child Household E x. x N £d3fl_
26 Robinson 66 52 .788 .806h
28 Robinson 19 15 .789 .7h01
29 Robinson 22 17 .733 .680h
30 Robinson 19 13 .68h .3360
31% Robinson 58 58 1.000 1.0000
36 Waswagen 2h 22 .917 .9912
L5 Fitch 22 18 .818 .8h02
ho Fitch 51 38 .7h5 .5279
h7* Fitch 36 31 .861 .9668
55 Marshall 22 18 .818 .8h92
56 Marshall 50 35 .700 .2573
60 St. Louis 21 16 .762 .6366
61 St. Louis 56 50 .893 .9979
Berens River ghildren

Child Household fl 5 .513 2:51
65 Mukkwa 13 10 .769 .670h
66 Muhkwa 13 12 .923 .9767
67 Muhkwa l3 13 1.000 1.0000
82 Mukoman 6 h .667 .h68h
81:3:- islukoman 10 10 1.000 .0000
87 Wasse 12 10 .833 .3335
91: Si‘éIp 7 6 .857 .5167“
95 SIEIo .9 7 .773 .7021

 

xChild less

than two years of age.

159

The table shows that there are just two children
for whom it is Clearly'gfllikely that the observed
frequency of adult acquiescence is a sample from a
population of experiences in which the relative frequency
of adult acquiescence is .75 or more. For both of these
children, the probability that the Sample of experiences
came from the hypothetical population of exoeriences is
less than .01. For the other thirty children, the
probability that the samole came from the hypothetical
population is greater than .25. For twenty-three of the
children, this probability is greater than .50.

The two children for whom this hypothesis can
be rejected are both members of the Hunt family --
the family which has accepted most of the values of the
white middle—class. The fact that this hypothesis can
not be rejected for the middle child of this family
provides a problem. There is a possibkz explanation in
a statement by his mother. She volunteered the infor-
mation that this child had been so sickly when younger
that they had "Spoiled" him. If this is so, they have
not altered their behavior toward him to match their

behavior toward the others.

Summary for Grogp II learning theory

It seems clear that there is no basis in the data

160
for rejecting the Group II learning theory. As presently
formulated, its predictions are not precise enough in
the individual case to be entirely convincing as to its
correctness, but it is clearly a better euide than

V

is the Group I theory.

CHA RTER VI

THE WINDIGO BELIEF AND THE ":EAR OF SELF" THEOR

Introduction

Some slight additional support for the "fear
of self" theory may be secured from an examination and
interpretation of the windigo belief common to all the
Northern Algonhian tribes.

A number of studies of the "windigo psychosis
have been published (Folelson 1965; Hallowell l93h;
Landes 1938; Parker 1960; Teicher 1960). Some of these
have concentrated on the conclusions which may be
drawn from study of this pathological adjustment. I
shall not follow this practice, but I shall attempt to

draw some conclusions about U18 typical personality of

the Ojibwa from the widely shared beliefs about the

 

”windigo sickness” rather than concentrating on the
pathology of sick people involved.

I have drawn heavily on Teicher's monograph
(1960) in which he had emphasized the importance of
the beliefs about windigoes and windigo sickness.
Although my conclusions are somewhat different from
his, his compilation of cases of "windigo psychosis"
has been fundamental to my efforts.

161

162
The Windigo Beligf
The nature of the windigo belief has been well
presented by Teicher. A few quotations from his

monograph will serve to elucidate the belief here.

Throughout the kirge area inhabited by the
Algonhian—Speaking Indians of Northeastern Canada,
the belief in a superhuman, man-eating giant is
tenaciously held. This being is given a variety
of names, all referring to the same fabulous monster
who lives in the dark forest, preying on the Indians
to gratify his insatiable lust for human flesh.

An awesome and grotesque creature, some twenty to
thirty feet tall windigo is fearsome and blood
curdling (1960:2).

There are a variety of eXplanations as to how
the windigo came into being. Of these, the one
held most frequently is that the windigo was formerly
a human being who was transformed into his super-
human state of' by?] sorcery. In this transfor-
mation process, the individual developed a heart of
ice. His body swelled to the size of a pine tree
and became as hard as stone, impenetrable by a
bullet or arrow and insensitive to cold. (1960:3
citing Cooper 1933).

For present purposes file essential element of
this belief is the idea that any Indian can be trans—
formed into a supernatural giant with an insatiable
hunger for human flesh. In terms of this belief, "one
who develops this craving for human flesh or is con-
sidered to be in process of doing so is called a windigo"

(Teicher 1960:5).

163
Generality and Intensity of Belief
It seems clear that there were few Northern
Algonhians who did not share the windigo belief in the
past. Three paragraphs from Teicher's monograph (1960:h)
will indicate that many IRdians still accepted the

windigo belief as recently as 1950.

In 1950, L300 Saulteaux '[Ojibwa] living at
Island Lake, 326 miles northeast of winnipeg, were
completely disrupted and thrown into a state of total
panic and trepidation for an entire summer by the
rumor that a windigo was on the prowl. (Teicher
cites B. Paterson in The 1’InnIpegRreC Press, August
1952 for this statementwj

writing just a ew years earlier about the
“ontaonais— laskapi, a geographer reported that
numerous terrifying spirits still range the forests
...Windigo, the giant cannibal...who lives on human
flesh...no one among them would dare to doubt the
existenCe of the cannibal giant Windigo..." (Teicher
cites V.A. Tanner, Outlines of the Ggography, Life
and Customs of Newfoundland - Labrador, Cambridge,
Mass.: University Press, 19h? for this quotation.)

About 100 miles northwest of Sioux Lookout in
Northwestern Ontario is a small lake called Backwash
Lake. The contemporary Ojibwa believe that a windigo
resides there and claims at least one life each year.
Their fear is so intense that the area is studiously
avoided. (Teicher cites R.Alorenus, Crazx-White-
.Eag, New York; Rand—McHally, 1952. )

Internalization of the Belief
The quotations above suggest that the windigo
belief receives more than merely verbal assent. The

case reports quoted by Teicher empllasize how thoroughly

16h
this belief was incoroorated into the emotional reactions
of the Indians.

In fifteen cases, a verson considered to be in
process of becoming a windigo was killed before attacking
anyone. The intensity with which the windigo belief was
held is indicated by the relationships to the victim of
those involved in these oreventive executions. In
seven cases, the executions were carried out by band
members of unsoecified relationship to the deceased.

In two cases, youn men were executed by their own

fathers. A woman was executed by her two sons and

a man by his wife, later avoroved by his "closest rel—
atives." One of the oeople, a Mrs. Cochran was executed,
at her own request, by her brother—in-law, in the presence
of her assenting husband and children. Another, David
Meekis, was :xecuted by two of his brothers and two other
men. A young woman was killed by her father-in-law and

three other men with the consent of her husband. Yet

another, Me-new-as-cum, was killed by some aonarently

 

unrelated men at the urging of his wife.

The windigo belief was also shared by at least
some of the victims of the "windigo sickness". An
old woman who was believed to be becoming a windigo
told Hallowell's informant "...that if she became a

Q

wihtiqo she would be as tall as the trees and nothing

165
could step her...” (Teicher 1960:6h-65, from unpub-
1ished field notes of A.I. Hallowell). This woman was
cured through the efforts of a missionary.

In another case, "...a heloless, aged and ailing
..." old woman was renorted to have said that "...if
she isn't dead before the sun goes out tonight she
cannot be killed and will then begin to eat the
children" (Teicher 1960:88). This old woman was one of
those executed by her band before she could become in-
vulnerable and begin to eat the children.

Clearly, the windigo belief had been thoroughly
internalized by these old women, by the Island Lake
aeoole who Soent a summer in a ”state of trepidation
and aanic”, by those who avoided Packwash Lake, and
by all those who executed their relatives or friends
before the windigo transformation could become complete.

If it is assumed that such thoroughly inter-
nalized beliefs must be consistent with the personal
exoeriences of the individuals who held the beliefs,

on what exoeriences does the windigo belief rest?

Possible bases of the_windigg_belief in individual
exoerience

 

Teicher suggests that the belief has been
”...built up by inductions from actual exocriences of

starvation and cannibalism" (1960:110). But, given

 

166
the rarity of cases of actual cannibalism -~ Teicher
found forty-four cases for the entire area for a period
of 300 years -- it seems very unlikely that very many,
if any, of the total of seventy windigoes had had any
direct personal experience with cannibals prior to their
own illnesses. (fogelson 1965:88 accepts Teicher's

conclusion of rarity.) Even if they had had such ex—

(0

riences, such exacrienees would not account for belief

-

in an invulnerable giant —- particularly since most of
the Cannibals were clearly vulnerable when they were
killed to prevent their continuance of cannibalism.

Teicher, himself, notes this discrepancy.

Nevertheless, the unavoidable conclusion is that
windigo psychosis is a relatively rare phenomenon.
This is particularly striking in the face of the
wideSpread nature of the windigo belief and its
great importance in the lives of the people. It
clearly illustrates the way in which a belief may be
more important than the facts on which it depends
(1960:107).

It may well be the case, however, that actual instances
of cannibalism are_ngt "...the facts on which it[:the
belief] depends." t seems to be the case that such
thoroughly internalized beliefs can 39;, for most of
the believers, depend on having witnessed actual

cannibalism. It is certainly the case that belief in

a supernatural, cannibal giant does not depend on

l6?
having witnessed a real supernatural giant. There is,

however, an alternative experiential base for accepting

a belief in a cannibal giant.

There is a psychodynamic basis for the inter-
nalization of the windigo belief. The basis for inter-
ralization of the windigo belief which I shall propose
here is derived from Spiro's proposal of a basis for
acceptance by the Ifaluk of a belief in malevolent
ghosts. Since my proposal parallels Spiro's proposal
rather closely, I shall summarize Spiro's proposal,
quoting relevant passages from his article.

The beliefs of the Ifaluk concerning malevolent

ghosts called alus are paradoxical since:

...the aluS'are conceived of as the ghosts of
evil people, but according to both the Ifaluk and
ethnographic observation there are no evil people
in Ifaluh. If so, how can there be alus? Our
analysis, however, has revealed (by implication)
that there are evil people in Ifaluk. (1953:381).

Every Ifaluh has repeatedly eXperienced an evil person:

The'evgl_person that the_adult eXperiences is
himself, for everyone who exoeriences hostile
drives within himself is evil. Thus, psychologically
viewed, the tradition is true (1953:381 emphasis
in original).

 

 

This could also be interpreted to so

168
believe that anyone can be come a malevolent ghost
because at some level of consciousness, each of them
"knows" that it is possible for himself.

Similarly, the northern Algonkians never actually
see any cannibal giants, but they have internalized the
belief that any one of their own number can become a can-
nibal giant. If the parallel with Spiro's argument is
valid, it must be maintained that those northern Algon-

111 , I!
({nO‘v‘I ,

kians uho internalize the windigo belief must
at some level of consciousness that they could become
windigoes; that is, they must be, in some sense, aware
of a feeling of rage and of an "unconscious” (and
grossly Ixaggerated) sense of great destructive power.
This awareness of destructive power and hostility
is also a basic assumption of the ”fear of self"
theory which received support from the observations of
adult-child interaction.
The translation of "destructive power and
hostility" into a symbolic cannibal giant seems to be

quite consistent with what we know of unconscious

symbolism. It seems legitimate to apply to this case

l.‘

q

Freud's assertion that ....a colourless and aostract
expression of the dream-thought is exchanged for one that

s pictorial and concrete", (1938:361) since Freud also

H.

H

assures us ...that symbolism does not a pertain

eSpecially to dreams, but rather to the unconscious

169
imagination..." (1933:369).

Following a som what diff

('0

rent route, wallace

arrI ives at a conclusion consistent with the above:

Oral-aggressive impulses are projected in the
dread of cannibalistic monsters, the windigos; that

 

 

he windig is really a projection of the Ojibwa
personality under stress, is evident in the fact filat
tne Ojibv a are known to have a form of neurosis or

psychOSIs characteristically their own, in which
the victim imagines himself to be transformed into
a cannibalistic monster. (l952:l03 emo, hasis sup-
lied).

The problem 0; the apparent inconsistency between the

 

 

e as powerful and Ojibwa dependenCe
Psychological dependency is a generally rec—
ognized characteristic of the Ojibwa (Parl {er 1960:
606-7; Wallace 1952:10h; Barnouw 1963:151). How can
this dependency with a self-image of "helplessness" be
consistent with a self-image of peter and hostility?
There are (at least) two approaches to the ori IgI ins
of dependency in the literature. Horney (1937) can be

taken to represent one of these approaches and Kordiner

(1939) the other. Both writers emphasize the objective,

I

 

or factual, dependence of the child on its parents
(liorney 1937: 3 Kardiner 1939 W3l-3 t) From this point

tIey diverge markedly.

170

Horney tacitly assumes that the infant not only
;§_deoendent, but feels deoendent. She then builds
her theory of neurosis on the perpetuation of this
dependency. She takes this psychological dependency to
be the ultimate motivation for the repression of hos-
tility which she regards as the most frequent basis for
neurosis (1937:63).

hardiner, on the othethand, carefully dis-
tinguishes between the infant's factual deaendence and

the infant's feelings:

...infants do not oerceive their helplessness,
but on the contrary feel as if they control the

world (l939:3h—3S).
He goes on to exolain how this occurs:

The helplessness of the child is conceded by the
parent, aid is given unsolicited...the mother can
interoret his every discomfort and can do for him
what he can not yet do for himself...the control
over the environment exerted by the infant through
the mother's agency is very like the one we sub-
sequently observe in magic nractices...A few
vocalizations of the child set in motion a series
of complicated changes in the outer world which the
child does not understand, but which terminate in
easing his tension...During this period of magic
control through the mother, we cannot sneak of the
child's attitude of dependency, For such an
attitude to exist, it is necessary to have a wish
tension, a knowledge of the inadequacy of resources
to satisfy it by oneself, and an attitude of de—
manding this satisfaction from someone else (1939:

35-36) 0

171

throbservations of the relations of Ojibwa
mothers and their infants are, with one exception,
examples of Kardiner's description. (The one exception
was the little girl of eighteen months at Flambeau who
received five of the total of twenty-four punitive
reSponses received by all observed children). For all
of the others, a cry was followed by nearly instant
attention and satisfaction of the infant's needs. If

Kardiner's theorizing is correct, this infantile eXper-

A

ience should lay the foundations for a leClln: of power.
The assumption that Ojibwa have this feeling of power
(at least where their aggressive impulses are concerned)
is a fundamental element of the ”fear of self" theory.
The same assumption is also basic to the explanation
"presented above of the windigo belief.

Kardiner continues:

The period of magic omnipotence in the infant
does not go on very long: for the child is propelled
by the process of growth, which simultaneously changes
both the needs and the resources of the child.
However, the resources never catch up with the
needs...Whenever a situation develops which the child
cannot meet, and which is accompanied by a recog-
nition of the limitation of his resources and an
exoectation of outside help, we have an attitude of
dependency...The attitude of dependency can be said
in every case to be due to a failure of resources;
but this failure may be due to the social pro-
hibition of the exercise of certain impulses, to
constitutionally determined inadequaeirs or to
accidental interferences (1939237).

172

The Ojibwa child manages to retain his belief in
his own omnipotence in at least one area -- the area of
his power to injure others. When he annoys an adult,
the adult acquiesces in the annoyance three-fourths
of the time. According to the cognitive learning theory
employed here, this should result in the child's coming
to believe that he is too powerful for the adult to
control. This belief then leads him to inhibit his
aggressive impulses for fear of injuring those whom
he also loves. This then has the consequence that, when
someone makes him angry, he cannot take action to end
the other's annoying behavior. In short, his resources
fail every time somebody angers him. f this inter-
pretation is correct, his belief in his own power results
in his helplessness before the aggression of others.

Horney, I find, has long ago enunciated this
idea; although she does not trace it to feelings of

power :

Repressing a hostility means "pretending" that
everything is all right and thus refraining from
fighting when we ought to fight, or at least when
we wish to fight. Hence the first unavoidable con
sequence of such a repression is that it generates
a feeling of defenselessness, or to be more exact,
it reinforces an already given feeling of defense—
lessness. If hostility is repressed when a person's
interests are factually attached it becomes possible
for others to take advantage of him (1939: h).

173
The main reasons why awareness of hostility may
be unbearable are that one may love or need a
person at the same time one is hostile toward him

... (1939:66).

This inability to cope with the annoying behavior
of others is a large and important aSpect of inter-
personal relations and it seems unlikely that very many
people could maintain a genuine indifference to their
inability to deal effectively with such situations.
Horney, seems to hold that inhibition of aggressive
impulses may be sufficient to account for general de-
pendency. Writing of sources of "a weak and helpless

H

attitude toward life she asserts:

Or, most frequently, he may have a general
notion of having inferiority feelings. These feelings
are the result rather than the cause of his tendency
to recoil from self-assertion (1939:256).

Since the feelings of dependency are relatively
conscious while the earlier self-image as powerful and
destructive is closer to completely unconscious, the
inconsistency of the two beliefs need not become
evident to the individual who simultaneously accepts
both of them. It is possible that the belief in their
own helplessness helps them to control their aggressive

impulses.

CIL’ 71:51“: VI I

CCLCT UBICXEIIMID 137LICATIONS

The Theories of Motivation

 

Two hypotheses, one from each of two theories
of the motivation of Ojibwa restraint, have been
tested with data derived from observation of 767
inCidents of adult-child interaction.x These incidents
involved twenty-two adults and twenty-four children
from Lac du Flambeau, Wisconsin and seven adults and
eight children from Berens River, Manitoba.

The tests of these hypotheses, and therefore
of the theories from.whieh the hypotheses were derived,
were independent of each other. That is, the rejection
of one hypothesis did not entail the acceptance of the
other hyiothesis and the theory from which it was
deduced. The statistical tests could have resulted in
the rejection of both hypotheses and both theories.

These two theories of the motivat ioI of Ojibwa
restraint (one from Hallowell and one from Freud,

u *

%For the test of the "fear of self" theory, two
adults and ten incidents in which they figur:d are omitted
from the statistical analysis.

171!—

, 175
Bettelheim, and Erikson) were partially formalized
(see appendix) in the framework employed by some decision
theorists (Ackoff 1953, 1962; Churchman 1961: Churchman
and Ackoff l9h7a, 19L7b; Churchman, Ackoff, and Arnoff
1957). This formalization has two related purposes.
First, to make the relationship of the hypotheses to
the theories so rigorously deductive that tests of the
hypotheses are clearly tests of the theories from which
the hypotheses were deduced. Second, to provide
quantitative estimates of the subjective probabilities
of certain outcomes which would be necessary to explain
(within each of the theories) the behavior of the typical
Ojibwa. These estimates of the subjective probabilities

are also used in testing file theories of the manner in

which Ojibwa restraint is learned.

The "fear of others" theory
Statement of theory and hypothesis

I have called the theory taken from Hallowell
the "fear of others" theory. The basic notion of this
flieory is that the restraint characteristic of the typical
Ojibwa depends on the value he attaches to avoiding
serious injury to himself and on his belief that, if

he offends another Ojibwa, there is a significant

probability that the other will injure him through sorcery.

176
The test of this theory rests on the assumotion
that the Ojibwa do not believe that children are casable
of sorcery (Landes 1937:117), from.which (in combination
with the rest of the theory) the hynothesis is deduced
that they should be much less restrained in dealing wifl1

an offending child than they are with adults.

Conclusions

The data are not consistent with this hypo—
thesis. It can be rejected at the .01 level for 27 out
of 29 adults. The data for the other two adults, although
not statistically significant, is inconcsistent With
the null hysothesis also. Of course, if the assumption
that children are believed to be incaoable of sorcery
is false, then the hyoothesis which was tested can not
be deduced from the theory. The data oresented could
then be used to argue that the Ojibwa "really” believe
that children have supernatural powers. The assumption
that children are believed to have magical powers,
however, seems inconsistent with the readiness of the
Ojibwa to attribute such powers to adults and their
insistence that the child "...fast for knowledge and
abilities or he will be helpless in the struggle for
life" (Landes 1937:117). Since the Ojibwa do not attack
even the children, who cannot retaliate, I conclude that

the restraint characteristic of the typical Ojibwa is

177

not due to fear of retaliation by others.

Implications
Mast generally, and most obviously, this con-
clusion implies that fear of retaliation is not the

only motive people have for not attacking others.
The "fear of self" theory

Statement and Hypothesis

I have called the theory taken from Freud
(1963:62-63)_and other nsychoanalysts (Bettelheim
1950:207-208; Erikson 1963:195-208) the "fear of self"
theory. The underlying idea of this theory is that
restraint of the degree characteristic of the tynical
Ojibwa depends on the value he olaces on not doing
serious physical or osychological injury to other
neoole and on his belief that, if he directs or attacks
another Ojibwa, there is a high probability that such
injury to the other will occur. The original hynothesis
tested for this theory asserted that, when offended by
a child, adults would choose neither to direct nor
attack with the same very high frequency'with which they
choose that course of action when offended by another

BdUIto

178

This hyeothesis was accented for fifteen of the
twenty-seven adults. Under this hynothesis, the observed
frequencies of choice of the ounitivg courses of action
are very close to the eredicted frequencies. The ob-
served frequencies of choosing to direct and not to attack
are higher than nredicted and the observed frequencies
of choosing neither to direct nor attack are lover than
oredicted. When it was also noted that the frequency
of choice of directing was higher at Lac du Flambeau than

Barens River, it was hypothesized that the higher fre-

O

f‘
q.

quency of this choice was a lunction of level of ac-
culturation. Following this acculturational hypothesis,
the assumed subjective nrobabilities of moderately or
seriously injuring a child through directing him were
reduced —- slightly for the Berens River group and
somewhat more for the Flambeau people.

Ideally, these new hynotheses should have been
tested with new data. Since this was out of the question
with the available resources, they were tested against
the previously used data. Following these modifications,
there were no rejections for the seven Berens River cases
and only five rejections (all at .01) for the twenty
Lac du Flambeau cases. Six of the seven Berens River
cases and thirteen of the twenty Flambeau cases must be

accepted even at the .20 level. If the observed average

179
frequencies of choice for each location
are used, the deviations from the predicted frequencies

are very slight.

Conclusions

The "fear of self” theory is roughly consistent
with the data even under the first hypothesis tried.
Under the hypothesis that adults at higher levels of
acculturation regard directing as less dangerous to child-
ren, the observed behavior is in surprisingly close
accord with the predictions. The supports the theory that
Ojibwa restraint is motivated by the desire ggt to injure
others seriously and by the belief that any attack on

another will result in serious injury to the other.

i

In so far as, the 'fear of self" theory

implies that the Ojibwa conceive of themselves as

powerful and dangerous to others, it is supported by

.1"

the analysis of their acceptance of the windigo belief.

Implications

If the "fear of self" theory is, indeed, even

approximately correct, it implies that positive feelings%

... A_~

*By "positive feelings” I mean ”friendliness" or
"liking". I think that such feelings may exist in all
people who have had satisfying relationships with others.
These feelings may Spring from dependency, but it does not

seem to be necessary to assume that this is glwevs the case.

 

180
toward others play a more important role in human behavior
than most of our explicit theories concede. This
suggests that predictions of another's behavior on the
basis of"a rational assessment of his self-interest" ~-
83 that phrase is usually construed -- are likely to be
in error. This should not be taken to imply that there
are no people who do not act exclusively on the basis

of narrow self-interest as least some of the time. I

do mean to imply that altruism should not be excluded

3 pr ori.

It seems likely that a theory which took into
account both the fear of retaliation and the fear of
one's own anger in differing proportions would give
Abetter predictions for people who are less rigidly re—
strained than the Ojibwa. There is no reason why such

a theory can not be formulated.
Theories of the Development of Restraint

General

Two hypotheses, one from a stimulus-reSponse
type of theory and one from a more cognitive type of
theory have been tested. The same data used in testing
the theories of motivation are used in testing these

1

learning hypotheses, but these data are new analyzec

181
in terms of the learning GXperiences of each child.

I"
i
.5

As in the case 0* the motivational theories,

f‘

the tests of these hypotheses are independent 0* each
other. Rejection of one hypothesis does not imply

acceptance of the other.

Group I, or Sjfi, theory

 

Statement of theory and hypothesis
This theory asserts that an individual will

exaect any given outcome to follow a particular kind of
act only if he has chosen that particular kind of act
and the given outcome has followed it. Mowrer (1960b:
3h?) states that the subjective probability of a given
outcome for an individual tends to equal the objective
probability of its occurrence in his previous exaerience.

From the formal ’

'fear of others" theory, it was deduced
that the minimum subjective probability of retaliation
necessary to account for the restraint of the Ojibwa
is .h5 (see appendix). Hence, the hypothesis tested
is: Ojibwa parents punish offending children with a

1."
...

relative .requency of at least .h5.
Conclusions
The average observed relative frequency of

punishment of offending children is .031. 1 en trsted

182
individually for each child, this hypothesis was rejected
at the .05 level for thirty of the thirty-two children.
It seems clear that these Ojibwa do not learn restraint
as a consequence of punishm nt for offensive behavior

in childhood.

Implications

Many students of child development have con-
cluded that a simple stimulus-rCSponse theory is inad-
equate to account for all of human learning (of Howrer
1960b; Bandura and walters 1963; Sears, Rau, and Alpert
1965). The conclusion of this part of this study supports
that view.

I

Grpup II, or cognitive, theory

Stated in terms of subjective probabilities, this
theory holds that the child infers that the probability
that another would be injured if he were to attack the
other is equal to the probability that adults reSpond
acquiescently to his offensive behavior. From the formal—
ization of the "fear of self" theory it was deduced that
the minimum subjective probability of injuring another by
an attack must be .75 to account for the high degree of
restraint of the typical Ojibwa.

.1

Less precisely, but perhaps more clearly stated,

183

the high frequency of adult acquiescence leads the child
to believe that there is a high brobability that, if he
were to attach an adult, the nrobability that the adult
would be seriously injured would be high. That is, the
high frequency of acquiescence leads the child to be—
lieve that he is oowerful and dangerous. Therefore, the
hyoothesis tested is: Ojibwa adults resoond acquies-

J.

LIL”: O

Cently to ffensive behavior of children with a

relative frequency of at least .75.

Conclusions

The average observed frequency of adult acquies-
cent resoonses to the offensive behavior of children is
.76. Vhen tested for each child individually, this
hynothesis must be accented for thirty of the thirty-two

7"

Children. ror all thirty of these children the proba-
bility that the observed samale of their behavior came
from a distribution in which the relative frequency of
acquiescent resaonses by adults is greater than one in
four. For twenty-three of the children, this probability
is greater than one in two.

I therefore conclude that the restraint character-

istic of the tyaical Ojibwa is a consequence oi adult

acquiescence in the offensive behavior of children.

Implications

This conclusion suggests that at least some of the
aeople in other societies who are emotionally restrained
reached that state by the same route as the Ojibwa. These
findixgs suoport flue asychoanalytic hynothesis that adult
failure to control the child may lead him to overcontrol
himself.

When an adult is observed to be anxious or inhib—
ited about certain kinds of activities, there is a tend-
ency to attribute this anxiety or inhibition to "severe"
socialization concerning the activities about which he is
anxious or inhibited. If my conclusions about the develop-
ment of restraint by the Ojibwa are correct, then the
attribution of anxiety or inhibition to severity of
socialization is not always correct. At least in some
cases, the anxiety or inhibition may be due to social-

ization practices that would not ordinarily be described

as "severe."

Implications for the future of the Ojibwa

It seems clear that the Ojibwa (in the U.S. at
least) have no real choice but to become full partici-
pants in the dominant society. They annarently can not
maintain a really seoarate existence, and they can not

continue as deoendent and second-class citizens without

185
complete loss of self—respect. But, their typical
personality creates a serious barrier to their becoming
full particioa ts in our society.

For the Ojibwa (or other northern Algonhians) to
become full participants in our soc’ety, they will have
to become able to compete openly. If I am correct that
this inability to compete is a consequence of an un-
conscious conviction that to compete with another is to \
threaten serious harm to him, then this change can not
be brought about by merely reducing the external barriers
(such as racial discrimination) to their competing in
our economic and social systems. Since the greatest
barrier to their fuller participation in our system is
internal, some way of changing their belief in their own
destructive powers must be found.

If my contention that this belief in their own ”/
powers is a consequence of the extreme acquiescence of
Ojibwa parents to the offensive behavior of children,
then the development of this belief could be limited by
giving them more experience with adults who find it easy
to control children.

At present, nearly all of these Indians are ac-
quiring some experience in childhood with people who do
not hesitate to control them. These controllers are their

teachers. Since they do not go to school unti: thev are

v

six or seven, this experience of people who are capable

of withstanding the "powers” of the Indian child probably
comes too late to change the underlying self-image. The
Indian children are already quite restrained before the
age of six. At Lac du Flambeau there has been a "head-
start” program which has been very popular with the Indian
children. This early eXperience with people who are not
so acquiescent as Indian adults, may do something to
counteract the belief of the children that they are too
powerful for Others to withstand. The children might,

of course, come to conclude that, although whites are
strong enough to controlthen, Indians are not. This would,
however, still facilitate their assimilation into the
social system of the dominant group. It might be in-
teresting to investigate the behavior of those children
who participated in ”Headstart" for varying lengths of
time.

Limitations of the Study and uggestions

Q
for Future RCsearch

Theoretical limitations
One merit of attempting to develop a partially
formalized theory is that it forces the scientist to

make many of his assumptions eXplicit. This may be hard

on the scientist, since exolicit assumptions are ready

187
targets for criticism and investigation, but, by the
same token, scientific progress is accelerated.
Naturally, all of the assumptions, eXplicit and

implicit, of this study are potential targets for further

examination. I wish to call attention to just two of

The f

Ho

statements about restraint into the assumption that the
typical Ojibwa chooses to neither direct nor attach with
a relative frequency of at least .90. This quantitative
assumption conditioned all of the other quantitative
assumptions, eSpecially those of the subjective probabil-
ities and of the decision rule. In the absence of

O

quantitative data about interpersonal relations, this
seemed a reasonable assumption. If this assumption is
erroneous it has had numerous effects on my conclusions.
To illustrate, let us suppose that .90 is too
low for this probability. If that is the case, then,
according to my data, Ojibwa adults treat children very
differently from adults in directing the children as
often as they do. On the other hand, if .90 is too
high, tien the difference in treatment of children and
adults is smaller than I have concluded. Only a quan-

titative study of the reactions of adults to other adults

rho have offended them can answer this question. I?

rst of these is my translation of Hallowell's

188

such a quantitative study of adult interaction should
become available, it would permit adjustment of my present
data and the drawing of more accurate conclusions from
them.

Another of my assumptions which seems reasonable
but which may prove erroneous is the assumption that
the learner equates the probability of adult acquiescence
following offensive behavior with the probability of
serious injury to the other following an attack on the
other. The whole area of indirect learning is in need
of a great deal of investigation. The fact that these
assumptions did not result in unreasonable conclusions
suggests that these assumptions may be aporox’mately

correct.

Statistical limitations

 

Since these conclusions are not based on a random
sample of the Ojibwa population they can be generalized
only to the extent that the sample used in this study is
representative. The sample was not designed to represent
the whole pOpulation, even of the two groups studied,
but only those which were taken to approximate the typical
personality and even here, no claims to genuine random-
ness can be supported. The most that can be said, is

that the Flambeau sample seems to be quite heavily

189

weighted toward the culturally conservative families.

Observational limitations

 

The nroblem of recording the observations is
significant. In most cases, notes had to be recorded
from memory after leaving the observed oeople. Sound
movies would be ideal from the point of view of recording
all the interaction, but would be an even greater handicap
to undistorted interaction than a notebook and pencil.
Perhaos, a small, finger-operated device which could
record general tyne of offensive behavior and the resgonse
of the adult could be developed. This would provide
basic data and would facilitate the filling in of details

from memory.

interview data

 

The "fear of others" theory exolains Ojibwa

restraint as arising from a distorted oerceotion of the

oowers of others to injure or kill through magical means.
The ”fear of self" theory exolains restraint as a con-
sequence of the individual's distorted perception of

his own powers to injure others. Data gathered from
"depth" interviews with a sample of Ojibwa adults and
children might have added to the evidence for or against

H

either interoretation. Data from relatively ”suoer;

: :r“
lCiC..L

190
interviews, reaching only the most conscious perceptions
of interoersonal relationships would have provided ad-
ditional material for interpretation. The effort to
understand both their actual adult-child behaVior and
their verbalized views of the adult-child relationship
mi oht have resulted in additions to our kn owledg cc of
the connections between verbalized conscious percentions
and overt behavior.

Unfortunately, very little interview data was
collected so the investigation of these interesting
topics must await future field work.

Suggestions for future research

One oossibility which grows out of the tests of
the "fear of self" theory is the investigation of less
acculturated Ojib groups to see if the frequency of
directing is in fact a function of level of accultur-
ation.

Another investigation of some significance would

be a study of the ef 1fects on restraint, for the children

q

involved, of the ”headstart" program at Lac mu Tlambcau.
Following the Group II learning theory, we should CX‘CCt
that Ojibwa children who have the exnerience of being

in charge of oeoole who are not hesitant to control

children should show some lessening of the cha ra ctc.r -

istic restraint.

191

Finally, theories of interpersonal behavior more
general 'han the two presented above can be developed
and tested by somewhat similar techniques of observation.
$ueh theories could combine the vahies and subjective
probabilities of the two theories presented here and they
could contain additional values and beliefs. Such theories
could also be applied to the study of intergroup behavior,
with reference to competing or conflicting groups within

a society or to intersocietal conflict or competition.

APVEHDIX

IZATION OF THE TWO THBOiIES

(a

O? HOTIVATION

Introduction

 

The purpose of this appendix is rigorously to
derive some consequences of theories of Ojibwa person—
ality and its development to permit better tests of these
theories. Since the candidate explanatory theories of
Ojibwa typical personality are stated informally and
imprecisely, the first step is to put them into a more
rigorous form from which the reported bheavior of the
typical Ojibwa can be deduced. This will permit tests
of some additional corSequences of these theories and
also provide better estimates of the values of certain
variables needed in testing the theories of development
of Ojibwa typical personality. These theories are very
simple, perhaps overly simple, but the formalization
of these simple versions should facilitate testing and
enrichment of these simple theories.

This presentation can be greatly facilitated
(though at the cost of learning some new symbolism) by
the introduction of the conceptual system employed in
contemporary decision theory. The presentation of
decision theory which follows is based princirailf

192

193
on the works of C.W. Churchman and R.L..Ackoff (Churchman
and Ackoff 19h7a, 19h7b; Ackoff 1953, 1962; Churchman
1961; Churchman, Achoff, and Arnoff 1957).

This system will be introduced by presenting its

H, H

elementary, or vrimitive concepts and then using these,

first, to restate Hallowell‘s description of typical
Ojibwa behavior, and, second, to explicate Hallowell's

"Common-sense" theoretical eXplanation of this behavior.

Elementary concepts

 

The elementary concepts of the system are:

(1) an individual, IR, whose behavior is to be pre—

 

dicted or explained.

(2) a set of courses of action, Ci, which the individ—

 

ual can choose.

(3) a set of possible outcomes or consequences, Oj,
of the courses of action.

(h) the probability, :1, that the individual will
select Ci in a Specified environment.

(q) the probability, §;j, that Oj will occur if Ci is
selected by Ik in a Specified environment.

(6) the value, 2;, of Oj to IR.

(7) the environment, N, which consists of all the

 

conditions which determine Eij. The corer ts

19h
logic and mathematics are also presupposed in the

construction of the system.

Ojibwa behavior

 

Using these conceots we can begin to exalicate
the Ojibwa tyaical aersonality as described by Hallowell.
We begin with some of the behavior which ibllowell seems
to regard as central to the Ojibwa typical personality.
He refers to ”...suppression of any impulse to tell

I

someone elee what to do..." (1955:135) and to ”...absence

\C)

of overt aggression in face—to-face situations..." (l~SS:
278). "Telling someone else what to do” apaarently
describes a "course of action." We will abbreviate this
description to "directing" and denote it be "cl." Sim-
ilarly, "overt aggression” describes a different course
of action which will be shortened to ”attacking" and will
be denoted by ”c2."

For the nresent, we will restrict attention to the
compound courses of action, Ci, which can be foreme
from these two "primitive" courses of action.* These

compound courses of action are:
C1 = cl'c2' -- neither direct :3? attack

*"Direct" and "attach" can be defined within the
systeg of Churchman and Achoff, but for now they will
be treated as "primitives" of the "fear of others"
theory.

195

C2 ‘ cl c2' —— direct and not attack
C3 cl'c2 -- not direct but attack
Ch = cl c2 -— both direct and attack

”cl'” designates all courses of actionxfihichdo not
involve cl. It is read ”not-cl," or "cl-prime," or in
any other locution which retains the meaning as in C1
above.

Hallowell's description indicates that for la,
the tynical Ojibwa, Pl (the probability that he will
choose Cl) is very high and P2, P3, and PL are all low.
The examples he gives indicate that this is the case
even when some other aerson, Ib, has annoyed him. it
is assumed from Hallowell'sirenorts thatthe frequenqy of
chaice_of C1 is aaproximately .90. This assumation
conditions all the later assumptions concerning the
subjective arobabilities of the various outcomes given
choice of each course of action. Therefore, to the
extent that this figure is in error, the conclusions
based on the tests of both the following theories will

be in error. Research is oarticularly needed to correct

or validate this assumation.

The'Tear of others" theory of Ojibwa aersonality
The aroblem is to exalain the very high Pl

(nrobability of choice of C1) by Ia even when Ia is

annoyed. "Common sense" suggests that people rtg;ri“

196
from directing or attacking others when they fear that
the others may retaliate with suoerior force. Hallowell
anaears to adopt this view when he attributes Ojibwa
restraint to the fear that the other person will retal-
iate by means of sorcery (1955:1L8). It appears that
this theory is intended to exalain the behavior of adults
in interaction with other adults, since the Ojibwa

b

(D

lieve that children lack magical powers as well as
nhysical canability to injure adults.

If the assumptions in this account are made
somewhat more explicit the theory seems to assert that the
typical Ojibwa would "like to" choose a course of action
which would make the other oerson stop annoying him and
would also ”like to" injure his annoyer, but he refrains
from choosing courses of action (directing or attacking)
which are likely to have such om;comes because he fears

that doing so would lead the other to retaliate, pro—

ducing an outcome he does not want, severe injury to him-

 

self. This will be recognized as the ”balance of power"
or "mutual deterrence" theory, the assumptions of which
are so familiar as to seem almost necessary.

This theory can be cast in the framework of
contemporary decision theory where the individual's
probability of choice of a course of action (pi) is

u ,.,

' -‘ asp.“ ‘ «a
’exaected valet ,n.

(.1
r—e
...;
D

exalained as a function of the

197

individual of the course of action.

The "exnected value" of a course of action (ci)
is calculated by multiplying the value (Vj) of each
outcome (Oj) by the arobability (Bij) that Oj will
occur if Ci is chosen. The theory being explicated
emaloys subjective arobabilities, that is, the arob~
abilities which the tynical Ojibwa is assumed to believe.
This theory assumes that the individual will tend to choose
that course of action which has the highest expected

value in a given situation.

Requirements of the decision theory_framework

To exnress this theory in the decision theory
framework will require that we make exalicit not only
the courses of action and outcomes with which the theory
is concerned, but also the assumed relative ralues (Vj)
to la of these outcomes and the subjective probabilities
(Eij) that each of the outcomes will occur if each course

of action is chosen, which are also assumed in this

theory.

The outcomes

 

The courses of action with which this theory
is concerned have already been SjQCified in above. The

outcomes which are mentioned in the informal presentation

198

 

above are, first, Ib stpps_pnngying Ia, which will be

H H

designated

 

01; second, Ib is injured, which will be

designated ”02;" and, third, Ia is severely injured, which

 

will be designated "03."

Compound outcomes

 

The outcomes are clearly not exclusive; they may
occur in any combination. By means of a Boolean ex-
pansion, an exclusive and exhaustive classification of
these possibilities can be constructed (as was done in
the case of the courses of action). Since there are
fliree "primitive" outcomes (oj), there will be eight

compound outcomes (OJ) as follows:

 

01 = 01 02 03 -- Ib stops annoying Ia, Ib injured,

‘ Ia injured. '

02 = 01'02 o3 -- ib does 332 stop annoying Ia, Ib
injured, Ia injured.’

03 = 01 02'03 -- Ib stops annoying Ia, Ib ngt injured,
Ia injured " '

Oh = 01'02'03 —- Ib does not stop annoying Ia, Ib‘ngt
injured, Ia injured ' '

05 = 01 02 03' -- Ib stops annoying Ia, Ib injured,
Ia ggt_injured '

06 = ol'02 03' —— lb does 222 stop annoying Ia, lb

injured, la ggt_injured

07 = 01 02'03' —- lb stops annoying Ia, lb 3g; injured,
13.29: injured '

08 = 01'02'03' -— Ib doeS'ppt stop annoying Ia, Ib ngt
injured, Ia apt injured

The value assumqtions of the "fear of othersfiwjiferv

l. ... b-

Deduction of Ia's high probability of e};~:ir

(l)

tuation demands quite detailed assumptions

Ho

C1 in this 3
about the relative values of the outcomes in this sit-
uation and such assumptions do appear to be implicit in
Hallowell's statements. I shall try to estimate the
relative values of these outcomes to the typical Ojibwa

(as described by Hallowell) using a technique suggested

*J

C.

by a procedure develOped for a somewh t

O.)

i ferent pur-

I

pose by Churchman and Ackoff (Acmoff 1953:2h-27, 375-

376: 1962:87-91). Many researchers attempt to avoid
making detailed estimates of the relative value of
outcomes by concerning themselves only with the Single
outcome of greatest value. This does not avoid the
problem since it amounts to assigning maximum value to
one outcome and zero value to all other outcomes. In
many cases, this amounts to a greater distortion than
any other possible error of estimation. Therefore,
assignment of relative values by judgmentsfin“ which the

bases are

explicit is a permissible research pro~

ade
cedure (See Achoff 1962:82 Churchman, 5333 al 1957:115-116).

3 __
The first step in the procedure is to rank these
eight outcomes in accordance with the assumed preferences

of the typical Ojibwa. This theory seems to assert

that the typical Ciibwa strongly prefers any compound

outcome involving 03' (la is not severely injured) to any

r‘\/

of the outcomes involving 03. That is, each of 0:, g,

200

07, or 08 is strongly preferred to 01, or 02, or 03 or
Oh. It seems clear that Ia is assumed to value both
01 and 02, given 03', hence his first choice is 05.

His second choice denends on his are erence for ol'o2

or 0102'. The assumption is made here that stopaing

the annoyance is preferred to injuring the offender,
although, since the Ojibwa are "hostile" they might pre-

fer injuring the offender, the assumption is also made
that he arefers any of the above to ol'oZ'. Therefore
we arrive at the following ranking of these outcomes on
the basis of the assumntions of the "common-sense"

theory:

 

OS (Ib stops annoying Ia, Ib is injured, la is ngt injured)

07 (lb stops annoying Ia, Ib is Egg injured, la is not
injured) '

06 (lb does not stop, Ib is iniured, Ia' not ured)

W
q

is in
08 (Ib does not SUJa, lo is not injured, la is no
injured)

On the assumption that the ranking within the
group of outcomes which include injury to la is the same
as within the group which does not include injury to la,

this grouo ranks as follows:

01 (Ib stops, Ib is injured, la is injured)
03 (lb stops, Ib is not injured, la is injured)

02 (Ib does not stop, lb is injured, la is injured)
Oh (Ib does not stop, Ib is not injured, I: is 'n’ re?)

ng (or relative valuation) of outcomes

.4
.‘3‘
H 0
U)
"3
C.)
Q
L.

"fear of others” or

is a distingu'shing feature of the
”balance of power” theory.

The next step in the procedure is the assignment
to the outcomes of numbers which represent the relative

values to la of the outcomes (preserving the ranking of

F“ "4"

values above, of course). lhe lear of others” the

O

ry is
not so clear regarding the fine adjustments within this
ranking. I have therefore assigned these numbers on the

\J

basis of two considerations. First, the previously men—
tioned great preference for any Oj containing 03' to

one containing 03. Second, to choose figures which will
lead to predictions in accord with Hallowell's reports

(a great range of figures wiieh preserve the ranking will

meet this consideration). I have therefore assigned the

following numbers as value estimates:

 

 

OJ VJ 0i vi
o; 100 01 5
97 90 03 L
06 Go :3 3

It is conventional to standardize these f'gures by
dividing each by the sum.of‘all of them. When this is

et:

J _

done, we

‘3

 

 

Qj, vj Cj_g vj
OS .29 01 .015
O7 .76 03 .012
06 .23 02 . 339
08 .13 Oh 0.0

inese figures will be entered in the matrix to be des-

4

The subjective nro‘ability assumptions of the ”fear of
a

L ... II J. ‘.
'tUGLS tneory

 

It is convenient to present the probability
figures (Eij) in~a matrix in which the outcomes are
listed horizontally across the tap and the courses of
action are listed vertically on the left. The values
to la of the outcomes are listed below the outcomes.
Later, the exnected values of the courses of action and
the probabilities of choice of the Ci will be listed

vertically on the right.

 

OJ
e1 01 ea 93 oh g; 36 oz 08
VJ .015 .uolz .012 o .29 .23 .26 .18
C1 311 312
C2
C3 :35

203

In this matrix, Ell reorese.ts the subjective prob-
ability that Ol will occur if Ia chooses Cl, 312 rep-

resents the subjective orobability that 02 will occur
if Cl is chosen, E35 represents the subjective
arobability that 05 will occur if C3 is chosen, etc.
These probabilities are determined by the prop-
erties of the environment (N), or, rather, since we are

dealing with subjective probabilities, by what Ia be-

 

q

lieves the nroperties of tne environment to be. As
this theory has been exoounded by Hallowell, these en—
es and intentions

l t

(7'
Ho
l-‘I
l4 0

vironmental properties are the caoa

directly through fos-

Ho
e-O
,4

al factors operate

H.

of Ib. Soc

 

tering the belief in sorcery whicl is regarded as the
effective factor in producing restraint. The nrob-

C

C)

(1‘
H.
H
H.
(....
H.
U)

which are inserted in the matrix thus reo—

sent in numerical form the capabilities and intentions

"3
('0

which the *3pical Cjibwa imputes to other Ojibwa adults,

(

0

according t this theory.

H

n general "common sense" assumes that directing
(C2) someone is more effective in producing a desired)
change in the person's behavior (01) than not directing
him (Cl); that attaching him (C3) in some fashion is

much more effective than merely telling him, and that a

combination of both directing and att cking (Ch) is

slightly mor effective than attacking alone.

r101,
L.‘ (.1-

The Oj

Ho

bwa seem to believe that Cl is very

unlikely to injure the other person (02), that directing

*‘3
[-do
:5

((2
DJ
,4
0‘)
(1
0
k4
f“)

I

him (C2) has considerable likelihood of inju
esteem, and that attaching him (C ) is very likely to

\J

injure him and hence that both directing and attacking

:3‘
f—u

m (Ch) has slightly greater nrobability of injuring him

than attacking alone. It is Significant in determining

4 o

the subjective drobaoilities held by the typical Ojibwa

V.
,4
f.)
(9°
(f-
:3.
(D
(
(....
Ho

bwa are osychologically injured by what the
average white American would regard as a mild verbal
attack. hey do not feel that it takes much "to really
tell someone off."

The only course of action (of the set under
consideration) which does not have a high orobability of
oreducing retaliation causing severe injury (as the Ojibwa
are renorted to see it) is Cl (neither directing nor
attaching). As Hallowell reverts the Ojibwas' beliefs,

q

attacning someone is more likely to produce retaliation
by sorcery than directing him, but the nrobability is
high in both cases. Tne probability that the retaliator
will be successful in seriously inj

attacker seems also to be very high in the belief of all
but a few individuals. As Hallowell pictures it, each
oerson seems to feel that nearly every adult can and

will injure him severely by sorcery on (whet we rr 1:;

205

‘

as) provocation totally inadequate to the intended

H0

In“

njury (the ntended injury 18 oiten eeath).

In calculating the probabilities of the compound
outcomes (oj) for each of the compound courses of action
(Ci), a numerical probability consistent with the above
assumations will be assigned to each of the primitive

outcomes (oj) for each course of action. If the

nrobabilitv of o‘ is n then the arobabilit of o" 18
- u . J -

q

l-p. Each of these primitive outcomes is assume' to be
an independent event (each of which is contingent upon
Choice of some Ci), hence the probability of a compound
outcome (Oj) is the nroduct of the probabilities of the

three orimitive outcomes (or their negations) which compose

the compound outcome.

Calculation of the numerical nrobabilities (Eij)

 

For the purpose of assigning numerical values to
the orobabilities of the primitive outcomes, the total
range of probabilities will be divided into twelve
classes, each of which (except for the two end classes)
will be represented by its midpoint. These classes are

as follows:

 

 

flange of subjective Range
nrobabilities: renresented by:
virtually zero (<.Ol) O
.01 - .lO .0
.ll — .20 .l
02 ‘- 030 CL—
031‘- QE'O 03

.‘d
a

051 "' .60

.61 - .70

O
IJ-x'] G\\f1-'

O \FLU‘LU1\T1\J'L\I‘L\JL\})\I1 \Ji

.7]- "' 0:0 0
0

DUI — .90 '1‘.)
.01 _ an O
/ .z'l 0/
virtually certain (p>.99) 1.0

The following table reorese nts the tentative as-

signments of nrobabilities to each of the primitive out—

 

comes for each of the courses of action based on the

"fear of others" theory (where both Ia and lb are adults).

1H8 UUED SUBJECTIVE j"’33P“?ILITY O“ OCCUTTTUCE OF EACH
a? TH: TIIXITIVE OUTC 1E8 GIVEN CHOICE 0? ECU COURSE
0? ACTION

01 02 03 01' 02' 03'
Cl .15 O O .8 1.00 1.00
C? .39 075 07:7 06) .25 0.2:)
J fir’ fl ’
C3 0 75 DOS 0 (3:!) 0 C/ 0 05 015
Ch .85 .95 .85 .15 .05 .15

Then since the nrobabilit of a compound outcome eiven
, r - ’ v
choice of a course of action, is the product of the prob-
abilities of the three primitive outcomes, £21, for er-

ample, is the product of p(ol), p(oZ), and 1(93), ““ in

(4

207
T .75 x .75 = .2 . The remaining thirty-one Eij

vere calculated by a similar nrocedure. When these are

entererl in the matrix, it looks like this:

 

 

[I I IL

was noz<nzmo5n do%owo500v are moonwwwomwmozm 0% are ooaeocsg

noflooesa ooeammm ow moflwo: mam Jammomfloa coHoéu

as 9H am am mg. (m om owl so
5%

me o o o o o o .Hm .mm
mm .mo .wo .oau .Hm .oos .Hm .omm .oru

om .oe .mo .oum .oeu .HH .omm .oomo .ooeo
2. .am .Hm .owo .oommu .Hmll.ll .gm: .oomm .Ireboum

[vii .on ‘1 .0000 .oem .0 .mo .mw .mo .1 .Hm

ocaooamm mza

OH H o-om. H Somwroa mmwoow Sow mwwmox
on H OH on. u ameOOd mun Sow mafimox
Ow n o~.om H Sea mmaoow oma mwwmox
or H OH on H anS QHaQOd m3; warmer
H OH on ow H HU maoom m330%mzm Hm. Ho Maucaoo. Ho wagesmo

II
II

oH‘om om
0H ow.ow
o~.on_om
OH on ow.
o~.om om.
ow om.omg
H oH_om.ow.

 

H
H

r

Ho meow Bowmwoe osso%wbe Hm. HU

H
H

 

(“J -~

0 O '3 O C) C) O C)
J C\\I‘L‘J.“\.~) 2‘0 *“‘
ll

 

HU mwovm m330%~3m Ho. Ho new MSQCama. Ho 3
HU comm 30d wdoe m::o%wbm Hm. He Dow ”Susana. Hm bow ”Bunsen

Ha goom sow whoa mSSQKHSQ Hm. Ho MSDCwna.
HU mwoom mSDoKHSQ Hm. Hv.www.HDECwao. Ho Hzgcaoa
Ha mono WWW mace m::o%wdo Hm. Hc.mww ngmwoo. Hm ”Sucsom
H6 mwovm mSSOKMSQ Hm. Ho HSECon. Hm mow “Sheena

 

Hm MDQCqma

Msuewom. Hm Sow ”madame

 

3d HSchma

 

 

209

This matrix oresents information to the effect

H
,_1
(D

that this theory asserts that Ia believes that, if

c ooses C1, the orobability of any outcome other than

(D

h

7 or OS is virtually zero. It also informs us that la
is held to believe that C2 has a relatively high prob-
ability of producing Oh, but, since Oh does involve
severe injury to Ia and does ngt involve the end of the
annoyance nor injury to Ib, he nuts no value on this

outcome.

Justification 0: these calculations

It may be asked, "Why make such calculations as
these when the theory has been stated so imprecisely?"
The justification for this effort lies in the belie f
that such an attemot to state this theory precisely may

more ouickly lead to research specifically designed to

;
I

test these assumptions and also, as will be shown later,
this set of assumations and calculations (or ded1ctions)
does orovide something aéproximating a rigorously
deductive exnlanation of Hallowell's reports of the very
low frequency of directing or attacking behavior among

the Ojibwa. These figt res could be rounded to the nearest
tenth or modified in other ways and they would still
result in somewhat similar deductions about Ojim

behavior. These figures, however, seem to ha?” thr i5vary

wnrk particularly'well.

10

P)

T“ J" :
oxoectrd value of a course 01 action

 

The exoected value of iven course of action

(1)

is the 31m of the oroducts 01 the value of each outcome
andthé nro'ability that the outcome will occur if the
given course of action is chosen. In the 83? Mel of

the statistician, the exoected value of C1 is exnressed:

jVj. For E “(C ) in the theory under
1 i

j
sv(c1) = E E1
CO:

5 eeration we write:
J=1
O
U
:v(ro) == E :2 3vg = .“o(.013) + .36(.0089) +
i=1
.oa7(.o1r) + .2(o) + .oa7(.29) + .12(.23)+ .oaa(. 26) +

a v
I

their values entered in hatrix Al:

‘3
H.

as a function of engcted value

 

If I interoret the "fear of others” theory

4

that the prob-

L)

correctly, it can be construed as asse ertin

OJ

ability, Pi, that n individual will choose any course
of action, Ci, is a function of the exoected value (for
him) Of all th? Ci. It is 8 frequent gracticc in studies

of decision-making to assert that the individual should

always choose the course of action which has the highest

211

Eran X DH

 

om OH on ow imp: lpm: lam om, ow. m<AoHv m<onHv mm
<u .oem .ooao .oem .0 .mo .mw .mo .Hm .ooomo

oh 0 o o o o o .Hm .mm .Ho .0000 .0“
m .mo .wo .004 .am .oma .Hm .orm .oee ,.ooq .ooouo .owa

nu .oe .mo .owm .ooe .He .owm .ooMo .ooeo .ooo .ooomm .omo

or .om .Hm .owa .ooom .Hm .onm .ooom .ooem .om: .oooem .ome

OH
0m
ow
of
om
00
3a

(\3
CC

[I

H

I!

ll

hos nos<ozwezw woweamzoa. are momomeOmwMozm 0% #30 ooSoocsm oedOOBem
mam ooaooeso nocamem ow modwo: one oaommsamo omHoEu

ow om ow
o~.om om
OH om.ow
o~.om.ow
0H om ow.
o~.om ow~
QH on‘om.
o~.om.ow_

ll

HU
Ho
HU
Ho
HU
Ho
Ho
Ho

OH H 0H.om~ H Somwfimd gmaeow 30w mwwmox
Om H ow om. n mwaooa use Sow mwwmo:
Ow u OH‘nm H Sow wamow vow mflwmo:
or H OH om H vow: mmamow mag mfiwmox

mwomm m3:o%wsa Ho. Ho Hagcwma. Hm Magcaoa

doom WWW mdoo mdnofiwam Hm. Ho Magmamo. Hm “Sheena

mfioUm mSSOKHZQ Ho. Ho mow Hbgcdegu Hm Magmama

momm.WWm mnoo msso%wzo Hm. Hv.www Rwandan. Hm HDgCamQ
mwoem manofiwsm Ho. Ho ”Sheena. Hm nod Magcnma

doom new mwov m3:o%mbm Ho“ Ho MSQCaeQ. Hm Sow Hmucwma
macaw mzbo%wam Hm. Ho mbwmmnbcdam. Hm nod.H:gcdom

doom mow mwov m3:o%mbm Ho» Ho now Hzgcamp. Hm Sow Mageamo

 

 

 

 

 

 

 

 

exaected value, that is, Pi = 1.0 for the course of action
which has the highest exaected value for him in the
given environment. This is often regarded as the best

1

decision rule for the individual, but it seems unlikely
that many individuals will be observed to follow this

rule with oerfect consistency. It seems, however, that
some aaar Ximation to this rule is asserted by the theory
under consideration.

One way to aaprox'mate thls decision rule is to
exorcss the individual's probability of choosing any
course of action as proportional to some power (such
as the square or cube) of the exaected value of the course
of action. As higher oowers are used, this rule approx-

"naximization" rule

imates more and more closely to the
(i.e. Pi aaaroaches 1.0 for the course of action which
has the highest EV and aperoaches O for all others).

(See Howrer 1960b:330-3b5 for summary of experimental

evidence on aaoroximate maxim1zing.)
A decision rule tentatively to be enoloyed in this theory
In the Ojibwa case with which we are concerned

here, we will somewhat arbitrarily adont the rule that

4

O

the typical Ojibwa's probability f choosing any course

of action is orooortional to the cube of its exaected
' ' ISV3 (C.i‘)~___

value. In mathematical symbols, Pi : h .

o ‘ I

1 =1

213

:J‘
(”0

Assuming that t tyoical Ojibwa does have the values and

[-30

X

subjective orobabilities attributed to him in Matr

A and follo 8 this decision rule, his probabil'tie of

U)

choosing the four courses of action are as follows:

Ia's probability of choice of C1 2 Pl = .91

Ia's probability of choice of 2 = P2 = .039
Ia's orobability of choice of C3 = P3 = .029
Ia's orobability of choice of Ch = PM = .021

many questions can be raised about these prob-

Pg

abi ities of choice and about the value and subjective
probability assumptions from which they were derived,
but since these figures arejguite consistent with Hallowell's

descrigtion of the tyoipal Ojibwa, these questions will

 

be postponed until later.

Personality

 

Any such rule, combined with the information
about values and subjective arobabilities necessary to
calculate the exvected values of the courses of action,

is the oersonality of an individual. Such information

 

about an individual is his personality since it permits
us to predict (arobabilistically) his behavior in any
situation. This is a nominal definition of personality
in the sense in which Spiro uses the term nominal in

"Culture and Personality: the Natural History or r

211i
False Dichotomy" (lQEl). The personality of an indivd-
ual is thus the aroduct of the scientist's study of the

ndividual just as the "law of falling bodies," S =% eta:

Ho

Ho

3 the oroduct of the scientist's study of falling bodies.
Both the aersonality of an individual and the law of
falling bodies serve the same function: each permits

the arediction of the behavior under varying (but

soecified) conditions of the entities to which the h
3

8’7p1y.

sonality to oredict behavior in a different

 

C
m
o
o
"3
J
a)
'1

If we assume that our typical Ojibwa places the
same relative values on these outcomes in situations

which differ only in the subjective probabil ties (

a

ij)
of occurrence of these outcomes, and that his choices
follow the same decision rule, we will predict different
orobabilities of choice of the same four courses of
action in this new situation. This amounts to saying
that his oersonality remains constant but his behavior
changes to meet the (at least subjectively) different
situation. Let us examine a somewhat different situation
which is central to this study -— the situation in which

‘

Ia is dealing, not with another adult, but with a child.

Ho

tion of the changed 3

C”)
J
'3
O
H.
H)
H.
O
(\

tuation

For this change» situation, we assume that the
relative values attached to the outcomes by the typical
Ojibwa adult remain the same. That is that he still
wishes to have the annoyance stopped, that he wishes
strongly to retaliate and injure the annoyer, and that
he most strongly wants to avoid serious injury to himself.
It seems unlikely from Hallowell's descriptions that the
subjective probabilities of the various outcomes will

f"
T

be changed except or the believed probability that an
Ib who is a child can produce serious injury to an

adult. The Ojibwa believe that no child has ny

C.)

supernat"ral powers. They are eXplicit that children
are extremely vu nerable as a result of this lack
(Landes 1937:1l7-118). Hence, for all courses of
action, the probability of 03 must be virtually zero and

the probability of 03' must be virtually 1.0.

 

Deductions ”ran this changed situation
Calculating the subjective Eij from this one change
in the subjective nrobabil'ty assumptions, and cal-

culating the exnected values of the courses of action

and the resulting Pi, we get the results diSplayed 1n

Katrix Bl, (next page) where Ia is an adult and lb is

I40

a child. Thus, this theory from which t is possible

 

 

nu 0H 00 mar. of ,m ow ow. om m<nnnyrxlm<wfiouv am

<e .om .oome .oem .0 .mo .mw .mo .Hm .oomo

as o o o o o o .Hm .mm .Ho .ooom .HH

mm o o o 0 .mo .:0 .omm .Ho .m: .oH: .mm

ow o o o 0 .4H .mT, .oww .on .mm .omm .wm
MW e: o o o o .ou .H: .orw .ooqm .mu .omo .wm

n/M . . _

217

to deduce the reported restraint of Ojib adults in
interaction with othc adults has, as another of its
logical consequences, that adults, when annoyed by

I

children, can be exoected to choose courses of action

involving attack (C3 and Ch) with a frequency of .67.

 

 

H H

(D

An alternative version o? the 7ear 0? other theogy
Hallowell seems to feel that the Ojibwa are not

inclined to take any chances on retaliation even though

the subjective nrobability of such retaliation may not

be hi h. I” the subjective arobability ofo3 is reduced

(3

for C2, C3, or CD, there are two oossible ways of
retainiig high Pi. The value to the individual of ‘o3'
might be increased or the in di vi c-ual's tendency to max-
imize exnected value might be increased. Since the value
assigned to 03' could be increased only slightly, the
alternative of greater tendency to maxiriize must be
adopted. If the individual's probability of choosing

a course of action is assumed to be proportional to the

tgnth_nower of ts exnected value (rather than to the
third nower as in Matri' A1), a P1 of .91 can still be
attained if the subjcc i eorobabilities displayed below
are adopted. The subjective orobabilities of all the

oriznit ive outcomes excedt o3 and 03' are retained un—

changed as indicated in the following table;

218

CD

TA L3 8.2
SUB JBCTIVE RR33ABILITIES OF -RIRILI‘ OUT O"Zo GIVEN
CHOICE OR C1,C2, C3, OR CL. ”RRKR OF OT rlERS"R HEORY,
NEAR RERRECT LAXI HI ATIVN, ILINIQUM PROBABILITIES OR 03.

01 02 03 01' 02' 03'

Cl .15 O O .85 1.00 1.00
P .33 .75 .h5 .65 .25 .55
c3 .75 ,9: .55 .25 .05 .b5
Ch .85 .95 .55 .15 .05 .HS

”hen the sub oj.ct1ve nrobabilities of the com-

pound outcomes are inserted in the matrix, we get
(see Uatrix A2)
These arobabilities of choice are consistent with

q -

Hallowell's resorts as are those in Latrix Al. If we
now make the same assumntion about the lack of danger

in directing or attaching a child that was made above
(9. 215) while retaining Pi as arooortional to the tenth
no er of the exoectcd value of Ci, we get: (See

LIatrix 32 p. 221) If this version of the " fear of others”
theory is more nearly correct than that presented in
Katrices Al and El, then the orobabi lity that an adult
will punish an oftenuino child is even higher than in
the version of Katrices Al and 01. Pi in Matrix B

are .11, .22, .35, and .32, while in Xatrix BB they are

011’ .11, .52, and .360

19

R)

Rogmalization of the "fear of self” theoqy

f‘

Since this theory 01'1ers to e: <jlain the same

1“

observed behavior as the fear of thers neory, it is
identical in many resnects and similar in others. In
1"

the ”fear of others” theor orimitive o; is ”lb is in'uree
’ _

" In the "fear of

(D
H
D

L...
O
(3..
O

—-_

a.e orimitlve 03 is "Ia i

self" theory, 02 is ”lb is injured mod mately, " and 03

 

is "lb is injured seve rely.” The some assumptions about
the relative values of the compound outcomes to Ia

are made in both theories (although these outcomes have
different meanings in the two theories), and the same

assumntions are made about the _:robabili ties of occur—

rence,g1ven choice of Cl, C2, C3, or Ch, of these

nrimitives. Consequently, Matrix Cl is identical with

There may be some question about the meaning of

compound outcom es con tai ing combinations of ”moderate

H

injury to Ib" (and its negations) with serious injury

to Ib" ( and its negations). These combinations are

H

aossible so long as "moderate injury" and serious

in Ju ry" are de: ined qualitatively. For example, an

individual can receive both bruises (moderate) and

fl

1ractures(serious) or he could be illed (serious)

1 ._,1,~rg
.. a '\.

1
LA-.

without receiving bruises. (It should not be come

L9“?

R)

1'11 '0

IZO'O

{\J

€13

\A)
\O

EI'

C)
R)

C)
C)

C)

LIO' 01°

3500'

€I'O

‘i‘n'o

130°C

C")

P)
\l)

IO

0

I'Q

{6'0

 

BRIBA

10'

C)

68’

PO

‘1’.

('3
U

(c-01 X)

IO

(1

£0

fig

’7‘?

\fIfifbe'y
J\ um ..‘

221

0w.

)3

.NLIt‘

 

O

. ,

C)
C)

b)

0
<3
(D

\1

K

I

P)
O\

p...-

\fl

(,3

C)
CD

CO

Q

‘~..‘

0
U)

 

O”)?
(..L__L'_

from these examoles that injury is restricted to physical
injuries. Any action which reduces the probability that
an individual will produce an outcome which he values,

produces injury and is classed as an o :ensive action.

 

 

Imelicatinns o“ the ""ear of self" throty ’0? “dU1t
interaction with offendine childrci

 

 

ective nrobaoilities assumed in
the ”fear of others” theory were modified in accordance
with assunntions about tie subjective probability of a
child's izjurinl someone through sorcery, this theory
aredicted that adults would punish an annoying child
with a nrobability of .67.
"fear of others" theory is not supported by the analysis

‘

of the data (Ch 5). lor does it seem possible to Maze

H.
O
H.
t’)
{—1.
Q
,J
(‘9‘

with the data by minor modilications.

In testing the "fear of others” theory, 1t was
hyéotlie81zed that adults believed that children were
inca able of retaliating for adult direction or

nunishment. This hynothesis via 3 justi fied'on the bas1s

n

o; Lande s' s statemerts about the believed vulnerability

f‘

0: children due to their lack of supernatural powers.

In estin he ”fear of self theory, the

(Q
(4-

critical issue is the beliefs of adult Ojibwa concerning

the effects on children of directions or nunishment.

 

: a < l
ad on an em lest. mm srmo: OFONI om m<Aodv m<mA0uV ed
<e .oem ..ooao .oem .0 .mo wmd .mo .cm .ooqmm

on o . o o o o o .Hm .om .Ho .0000 .ae
om .mo .mo .oaq .Hm .ooa .Hm .omm .ore .ooa .ooomo .owo
ow .0H .m. .owm .oee .HH .owm .oomo .ooca .ooo .ooomm .omo
0: .mm .Hm .omo .ooom .Hm .omw .ooom .ooem .om: .oooeo .ome

H

H

OH om om
o~.om ow
OH om¢ow
OH.om.ow

OH on ow.
o~.om ow.

OH om.ow_

o~.om.ow.

[I

II

II

was cob<osmosd sowmaesoo.
convocba oocamom ow mowwo: use

t I

are meeonMOmdwobm ow
daemonaod voHoéu

are ooBeocSQ ochOBam and

O~ n 0H.om_ H momwvma awqond Doe mwwmox

om n opom. n dwdeow and bow madmofi

Ow H 0H.om H Sea dmaoow dew mflfimox

a? n o~ om H dedv dmdonw and mdwmox
manofimboe mwoem. HU mahcemd Bomewmme%. HU “Saddam mm<oamH%
mano%m:nm doom vow meow. He Magdaom SomewoonK. HG ”Susana mm<owm~K

mzso%mboe
mSSOszom
mo<mdeHK

mDSQKmsoo
mano%m30®
mo<oao~%

m::o%m:oo
mo<awaH%.

o::o%msom deem :ow mace.

mwoom. He Dow.mdueqod Boaoamde%. Hy HDQCnom mo<owm~%
doom Sow mace. He Sow ngcana Bodmaowew%. Hy HDQCdem
mwoem. He HSQCnod Badmamwo~%e Hv Sow WDQCdom mm<owm~%
doom don mmoe Hv MSQCeeQ Bocmdmde%. HU Dow MSgCeeQ

mwoam. Hv Dow HDQCwod BommnmdeK. HU Dow Hagcned

HU now HSchmn Bodaamwowk. He Dow

Magdaod mo<edeK

22h

Kore than one hyoothesis about adult beliefs is pos—
sible. The first hynothesis to be tested was that adults
believe that children are affected by direction or
attack to exactly the same extent that adults are af-
fected. This leads to the deduction that the adult's
nrobability of choice of each of the Ci is identical
whether the offender is another adult or a child.

Tested under this hynothesis, the "fear of
self" theory eredicted adult behavior better than the
"fear of others" theory did under the hypothesis of no
fear of retaliation from the child. The results were,
however, far from ideal. The null hypothesis could be
rejected for twelve of the twenty-seven adults.

The observed behavior diffe ed from the theoret-

"5

ical exoectations under this hynothesis chiefly in the

.p
.L

higher relative requency of choice of C2 and lower
relative frequency of choice of Cl. It was also observed
that the deviations from exdected behavior were consid-
erably greater at Lac du Flambeau than at Berens River.
These observations suggested that, (l) the Ojibwa regard
children as less vulnerable to direction than adults,

and equally vulnerable to attach, and, (2) that since

P2 is much greater at Lac du Flambeau that belief in

the ability of children to take direction is a function

(id)
of level of acculturation.

Q

On the hypothesis that adults believe that direc-
tion is less likely to injure a child than an adult

and on the'hyoothesis that the extent of this bel iel is
a function of level of acculturation, new estimates of
the subjective orobabilities of the various outcomes,
given choice of C2, were nrenared. It was nossible to

find subject eorobability estimates which led to

deduction of Pi very close to the average observed Pi
for both locations. All sub ojective probabilities for
C3, and Ch remained unchanged. Table 8.3 disolays

t‘

the subjective probabilities 01 various outcomes, given

choice of C2, as ori iei nally estimated and as modified

\J

for both communities

TABLE 8.3

sur JEC rive rroaAaILI'ries or men or 111:: euro. ores, GIVEN
chair or c2 as onIeIm .LLY ESTILhTED AID AS LCBIFIBD
tr? :alaen LOCAerr. '“fiLP or seLr”'n:xrng ADULTS In
Iur,nar11en WITH CHIL en

01 02 03 01' 02' 03'
Original C2 .35 .75 .75 .65 .25 .25
Berens C2 .35 .55 .L5 .65 .h5 .55
Flambeau C2 .35 .25 .15 .65 .75 on;

Using these subjective arobabilitieo, the “”w

26

\

I"

deduced probabilities of choice for each locality are as

disnlayed in Table 8.L.

PROBABILITIBS O? CH3 C3 O? EACH CF THE<307£SES OF.ACTION,
AS ORIGINALLY CALCULATE? AND AS HODIFIBD FOR EACH
LOC1LITY. "TEAR OF SELF" TTeORY.ADULTS IN INTERACTION
31TH CHIUNUII

Original Berens Flambeau
P1 .91 .71 ‘ .50
P2 .03; .25 .h?
P3 ~039 - 23 .016

PL .021 .017 .012

rarer?" a“ eirso
A\-—J‘ a. \ul‘n .JU 1.21

953 The design of social research. Chicago,
University of Chicago Press.
196a Scientific methodz' optimizing aoolied

. research decisions. New York, John Wiley.
Bandura, Albert and Richard H. Walters
1963 Social learning and oersonality development.
how York, Holt, Rinehart and Winston.

marnouw, Victor

1950 Acculturation and oersonality among the
Wisconsin Chippewa. American Anthropol-
ogical Association memoir 72. ‘

V?

1963 Culture and personality. homewood, Illinois

Dorsey.
earrett, €.A.
1711 The dream dance of the Chipoewa and Ken-
ominee Indians of northern Wisconsin.
Billetin of the Public Huseum of the City
f Nilwaukee.

Berkowitz, Leonard
1962 Aggression: a social psychological analysis.
hew York, EbOraw—Hill
196M The development of motives and values in
s

the child. New York, Basic Boon

I

Bettelheim, Bruno
1950 Love is not enough. Glencoe, Illinois,
The Free Press

\

Do

(1’)

teohen T.

h Ojibwa socialization: some asoects of
narent-Child interaction in a changing
culture. microfilm, Ann Arbor, University
chrofilms, publication 11,55h.

1956 An interactional study of Ojibwa socialization.

'American Sociological Review, 21:191-198.

gs

kit?

9
If7

\
I

1950 Culture change and the personality of ’
Ojibwa children. American Anthrooologist,
’n
60.

Buss, Arnold H.
1961 The osychology of aggression. flew York,
JOhI’l Kiri 16y.

227

m
{\J
C)

d

lorman A.
19f? The osychology of behav or fisorders.

i
Boston, Houehton-mllflin.

Cameron

117

C urchman, v.w.
1961 Prediction and ootimal decision. Bnglewood
Cliffs, N.J., Prentice—Hail. ‘
l9h7a Psychologistics. (Mimeo), Philadelphia,
University of Pennsylvania Research Fund.

1

Churchman, C.W. and R. L. Ackofl
l9h7b An ey oerimental definition of me Mrs nality.
Philosophy of Science, 111:301 ~332.
Churchman, C.‘1-‘1’., Achoi‘f, P1.L., and B.L. Arnolf
1957 Introduction to operations research. New
York, John Wiley.

Densmore, Frances
1913 Chippewa music II. Bureau of American
Ethnology Bulletin 53.
1929 Chippewa customs. Bureau of American
Ethnology Bulletin 86.

C d r p l

‘0.

Bollard, thn, Boob, Leonard W., 1111er, Neal 3., and
Robert R. Sears.
1939 Prust ra ti n and aggression. Yale University
Press, 1*Iew Haven.

I

.4

Mlson, Erik h.

1963 Childhood and society, 2nd Ed. le York,
“’.Th lbrton.

Poeelson, Raymond D.

1965 Psychological theories of lwindigo "psychosis
and a orelim inary afirlica tion of a models
aonroaCh. In Sjiro, Z'Ielford E. (Ed),
Context andmmeani1ng in cultural anthro—
polog‘. New York, Free Press.

H

Proud, Sigmund
1933 The basic writings of Sig1u nd Freud,

qr

translated by A.1. Brill. 1ew York,
Modern Library.

1963 Notes uoon a case of obsessional neurosis.
reorinted in Rieff, Philip, Ed., Three

case histories. ew York, Collier Books.

 

f)“

Hagen, Everett E.
1T6? On the theory of social change. Homewood,
Illinois, Dorsey.
Biol 1 owel 1 , A. I .
1933. Culture and mental disorder. Journal of
Abnormal and Social -rychology 29:1—?.
1036 "The passing of the Hidewiwin in the Lake‘
Winnineg Region. American Anthronologist,
38:32-P1. '
1955 Culture and exwerience. -llilade
University of Pennsylvania Press.

 

l, C.G.
lQLZ The T"unction of general laws in history.
The Journal of Philosophy 39. Renrinted in
Heigl, El crbert and Se11a rs, u ilfrid
Readings in Philosyshical analysis, New
York, App nleton-Cent ury-Crofts.
liicl , liarwjld

on
y62 The southwestern chipnewa. AAA Lbnoir 92.

, Walter J.
1885 Hidewiwin or Grand medicine society of the
'3

Ojibwa. burear of American Ethnology
Report No. 7.

Horney, Karen
1937 The neurotic personality of our time.
New YOrk, Norton. ‘
o r‘ ' " 1
1939 New ways 1n psychoanaly31s. mew'Yorx,
w
110 rton.

Kardiner, Abram
1939 The individual and his society. New York,
Columbia University Press.

Kelly, George A.
1955 The osychology of oersonal constructs.
W.W. Norton and Co., New York

Kroeber, Alfred Louis
19b? Cultural and natural areas of native Korth
America. Berkley and Los Angeles, Universit y
California Press.

230

Landes, Ruth
1937 "The Ojibwa of Canada" in Head, mar
Cooperation and competition among a
oeonles. New York, leraw-Hill.
1938 The abnormal among the Ojibwa Indians.
Journal of Abnormal and Social Psychology

33tlh-33-

Linton, Ralph
1939 Foreword in Hardiner, Abram, The individual
and his society. New York, Columbia
University Press.

HeClelland, David C.
1961 The achieving society. Princeton, N.J.
D. Van Nostrand Co.

warer, C. Hobart '
1950 Learning theory and personality dynamics.
New York, Ronald Press Co.
1960b Learning theory and the symbolic processes.
New York, John Wiley}
NewCOT.: b T.H.
19b? Autistic hostility and social reality.
Human relations, 19h7, l, 69- 86

Orlansk cy, Harold
1 )‘9 Infant care and personality. Psychological
Bulletin h6:1-h8.

Parker, Seymour
1960 The wiitiho psychosis in the context of
Ojibwa personality and culture. American
Anthropologist 62:603-623.

Parsons, Talcott
1961 Social structure and the development of
oersonality. in Kavlan, Bert, Ed., Studying
oersonality cross-c culturally. New York,
haraer and Ro‘.r

Ritzenthaler, Robert E.
1953 The impact of a small industry on an
India n community. American Anthropologist,

55,], 113—1. 1113.

231

Sears P.R. fiaccoby, Eleanor 3., and H. Levin

9 ‘ 3
1957 Identification and child rearing. Stanford,
California, Stanford University Press.
Siegel, Sidney ~
1956 Nonparametric statistics for the behavioral
sciences. flew York, McGraw—Hill.

1er, rge D.

1955 Sociocultural and Psychological processes
in Lenomini acculturation. Publications
in Culture and Society'Vol. 5. Berkley,
University of Claifornia Press.

le Old E.

1531 Culture and oersonality: the natural
history of a false dichotomy. Psychiatry
1h.

195° Ghosts: an anthropological inquiry into

learning and oercention. Journal of

Abnormal and Social Psychology, '

\J

48:376~33L.
1961 Social systems, nersonality,a and functional
analysis. in Kawlan, Bert, Ed., Studying
personality cross— ~cu1tura11y. Mlaw York,
Harfier and Row.
Teicher, Morton I. _
960 Windigo psychosis. Seattle, American
Ethnological Society.

[-0
{‘1

Tet ing, David R. and Jerry R. Hutchins
1966 Comparison study 01 percentage 01 pure
form (P) resoonse .3 among two Ojibwa Indian
groups. Unpublished te—r m paper, 1H chigan
State University.

It, Jam". a nd COUICS, J.

952 The role of communication in the reduction
of interpersonal hostility. Journal of
Abnormal and Social Psychology, t7: 770-777.

, Anthony 1.C.

'51 Some osychological determinants 01 culture
change in an Iroquoian community in W.N.
Penton (ed.), Symposium on local diversity in
Iroquois Culture. Bureau of American

. T‘thnology Bulletin 1M9

2 The modal oersonality of the Tuscarora
Indians, as revealed by the Rorschach test.

T

Bureau of American Dthnology. Bulletin 150.

x)

23;

D...

John W.X., an Irving Child
Child training and personality.
Yale University Press.

New Haven