SEX DIFFERENCES

The

procedure of analysis so farhas not considered the possibility that personality differences between thesexes

may

be reflected in the

Rorschach

data.

As

amatteroffact,thedata dosuggestsucha dif- ference, asaglanceat table5, below, shows.

Within

the

modal

class of 26records, 61.5 percent are male,although in thesample of70 asa whole, males comprise only 51.4percent. Testing the significanceof this difference

by

the critical ratio method, it is

found

not to be significant at the .05 level of confidence; that is to say,

we might

expect, if this experiment were repeated,

and

^if the true difference

were

nil,

more

often than once in 20 repeats, to find a difference as large as this between the percentages of males

and

of females in the sample

and

inthe

modal

^class.

Table ^5.

—

Sexdifferences

Personalitytypes Male

78 BUREAU

OF

AMERICAN ETHNOLOGY

[Bull. 150 alone a description of the personalities of all of the

members

of the

modal

or

any

other class.

He

conceives the

modal

type to be an

"ideal,"

an

axis

around which

revolve,

more

or less closely, according toindividual variation, the "real" personalities.

A

^large proportion of the sample (26 out of 70, or 37.2percent) produce Korschach rec- ords

which

in their quantitative aspectscannot bedistinguished

from

one another, at conventional levels of statistical significance, insofar as their relationship to the

modal

values of the total

sample

is con- cerned. Sixteen out of 70 (22.7 percent) produce Rorschachs

which

differ in various respects but retain the

major

outlines of the

modal

type;

and

28 out of70 (40.0percent) differsignificantly in

major

out- line,being subdividedforour purposesinto4 deviantsubtypes.

The

writervisualizes thisdistributionasasortof three-dimensional

normal

curve, described

by

a figureshaped

somewhat

likean inverted saucer (fig. 5).

The modal

typeisrepresented

by

^{thevertical axis}of thefigure,

and

islocatedatthe point of

maximum

height (frequency)

.

The

several variables (determinants, locations,

and

ratios) are meas- ured alongthe variousdiametersof thesaucer, allpassingthroughthe axis,

which

is the point defined

by

the average values for the

modal

class.

Any

single personality

would

be described

by an

irregulartwo- dimensional projection, the perimeter of

which was

defined

by

the scores obtainedineachof the 21 variables.

While

this visualanalogy perhaps ought not to be carried

through

to its logical mathematical consequences,

which may

involve relationships not inherent in the

phenomena,

it

may

serve asa

rough

descriptionof

how

the actual per- sonalitiesare distributedaboutthe

modal

oraxial type.

Inasmuch

as thefrequency distributionof

most

of the categories is

anything but symmetrical (usually they are J-curves, or

skewed and

truncated),itisobviousthattheempiricaldata

do

notacuaily fitthis hypothetical saucer.

The

failure of Rorschach scores to distribute themselves symmetricallyis,ofcourse,irrelevant

from

thestandpoint of individual scoring

and

interpretation,

which

does not have to take idealcurves into consideration.

On

theotherhand,itseemsarbitrary to

assume

that all individuals

who

give

OM

are identical in theirin- capacitytoproduce

M.

Failuretoproduce

IM

inresponseto thispar- ticular series of ink-blots does not necessarily

mean an

absolute ab- sence of the capacityto see

humanlike

actioninink-blotfigures.^^

The

people

who

give

OM

constitute a large class

whose M

^capacityis not

measured

at all

by

the Rorschach test but is subliminal to it.

For

^That professional Rorschach workers recognize that a zero score is not an adequate measure of the variable is evidenced by the use of the testing-the-limits period (see Klopfer, 1946) and of supplementary devices (like the Levy Movement Blots) for the elucidation of human movement concepts (see Schachtel, 1950). Schachtel says on this point (p. 97)^: "Rorschach mentions thatgivingno

M

responsesisonlya relativemeasure of the capacity for kinesthetic perception, since the seeing of movement purposely has

Wallace]

MODAL PERSONALITY

OF

TUSCARORA

INDIANS

79

f^,_pdPdS

B^GtJEE5.

—

The"saucer" distribution of personality variables.

practicalclinicalpurposesin

Western

society,it

may

be thataninabil- ity to produce even

\M on

the Rorschach is sufficient for

most

diag- nostic purposes, but theclinician is still left with the problem of the differential interpretation of cases of Oil/.

By

adding another 10 cards (andthewriterisnotsuggesting thatthisshould be done!)

some

of

which

resembled active

human

figureseven

more

closely than, say,

Card

3,

many

individuals

who

have given OJ/ on the

Standard

Ror- schach

would

be

found

togiveoneor

more M

^{on the}

^expanded

^series.

In

other words, the Rorschach test is so constructed that it does not

measure

the

M

^{capacities of people below a certain point; it} simply

lumps

all

M

capacitiesbelowthatpointina subliminal group. Itisas ifonecould onlymeasurepeoples'heightsinmultiples of5feet.

Midg-

ets

would

simply be "invisible" in such a system of measurement.

The

J-curve itself,

which many

of the distributions describe,

and which

hasachieved considerable reputationinthe"J-curve hypothesis of

conforming

behavior,"

might

be

adduced

here in defense of the adequacy of the obtained measures.

But

the J-curve can be a trickystatisticaldevice.

Any phenomenon which

is"really"normally distributed can be arranged in the

form

of a J-curveif one

lumps

to- gether a sufficiently large

number

of cases

toward

one extreme as havingin

common

the property of failing to pass or ofexceeding a

been made somewhat difficultin his inlt blots. Mostpeoplewho do not give

M

responses inRorschach'stest seehumanmovementinthe Levy MovementBlotsInwhich liinesthetic perception has been facilitated by the design of the blots and by the task given to the subject in which he is aslted to say what the figures on the blots are doing. Thus the absence of

M

does not mean that the person without Jf lacks the capacity for empathlc projectionand forcreativeexperience."

80 BUREAU OF AMERICAN ETHNOLOGY

(Bdli^ 150 certain criterion value.

Where

the

phenomenon

is such (as, for in- stance, in overtime parking time) that

an

absolute lower limit not onlycan be imagined butisactuallyreachedin

measurements

of equal class interval, J-curve

may

be

an

accurate description of the distri- bution.

But where

varying kinds

and

degrees of conformity are

lumped

together as"conforming" torule,

and

^{only deviations (in one}

linear direction) are

measured and

counted, one indeed often does findthat

more

people

conform

thannot,but one doesnot

have

a

com-

plete frequency distribution.

As

far as Rorschach scores are con- cerned,

where

one does not

have

a rule to

which

to conform,

where

there is

no

"telic continuum," interpretation of

an

apparent J-curve asbeingsignificantinthelightof the "J-curve hypothesis ofconform- ing behavior" is not justified.

AUport

specifies three conditions of applicability of the J-curve hypothesis, one of

which

is that "there

must

be

some

kindof law, rule,regulation,orcode,inthelanguageof

which

one

may

find stated the 'proper' or required

mode

ofbehavior

by which

this purposeis to be achieved."

As

Allport says, only the satisfaction of this, along with the other

two

conditions, will

answer

the objection "that

we have

not extended ourrecords far

enough on

the left" (Allport, 1934, p. 62).

Rorschach

data are not the kind for

which

the J-curve hypothesis

was

designed.

Thus

itishereargued thatthecapacities represented

by

the various Rorschach scoringsymbols are in

any

population distributed

on

two- sided,bell-shaped, essentially symmetrical (possibly normal) curves,

and

thatif a "better"

Rorschach

test

were

available, this fact

would

beapparent.

Hence

the saucer distribution actually depicts the prob- able distribution of capacities, rather than the actual distribution of

raw

scores (submodalscoresbeingunrecognizable, for instance,

when

the

mode

^itself is zero).

In

orderto further conceptualize thissaucer distribution, it should also be pointed out that the distribution of scores represented

by

it

arenot

raw

scoresbut standardscores,expressed intermsof standard deviation units

from

the average scores of the

modal

class.

The

use of standardscores will

make

the variousdiameters nearly equivalent.

But

even withallthe conceptualpushing

and

pulling

we have

already done, perfectly symmetrical bell curves are still not likely really to exist.

We must

expect that our saucer will always have

lumps and

dents, here

and

there, caused

by skewed

distributions. It is a hand-

made

^saucer, not amachine-perfect productI

It is notable that fully 40 percent of the adult population possess personalities

which

aredistinctly not "typical" Tuscarora. It

would

at thispointbepertinentto ask, althoughthe questionis

momentarily

unanswerable, whether other communities

on

investigation

would

Wallaces]

MODAL PERSONALITY

OF

TUSCARORA

INDIANS

81 show

a similarproportionof statistically deviant individuals.

On

a priori grounds, in

some

societies,one

might

even expect a larger pro- portion ofstatistical deviants.

In many

ways, the Tuscarora can be regarded as a very small,

homogeneous

society in

which

change is

notprocedingat different rates or on different levels for several geo- graphical or socioeconomicclasses.

More complex

societies,orpopula- tionsclassifiable as unitsof study

by some

other criterionthan social integration (e. g., linguistic dialect),

may

veiy well be

found

to be

more

variable.

The

apparently high incidence of statistical deviancy

among

the Tuscarora is not to be taken at face value.

The

"deviant" class, as

was

explained above, actually includes four

major

groups;

and

these deviant groups are very loosely defined.

The

largest single deviant

group

(the13 recordswith

F

percent greaterthan60.5)

make up

only 18.6percent of thetotal sample

—

^{justhalf the}

^number

^{of those within}

the

modal

type.

The modal

typestillretainsitstitleas the

most

fre-

quentsinglepersonality typeidentifiable withinthetotal sample.

In

evaluating thesignificanceofdeviancy,it

must

again be

empha-

sized that deviation in personality structure is consonant with full participation in social activities

and

with efficient utilization of the culture for the satisfactionof needs.

The

use of the

word

"deviant"

to describe

nonmodal

personality types does not in

any way

imply

"neurotic," "asocial," "maladjusted," or

any

other adjective connot- ing the square-peg-in-round-hole

phenomenon. The

one psychotic, indeed (hospital diagnosis:paranoid schizophrenia), included in the adult sample,falls into oneof the deviantgroups

{F

percent

>

60.5).

But

sinceadventitioussituationalfactors inadultlife,however,

would

play

some

^part in producing severe personal discomfort or even a clinicallyrecognizable disorder, it

might

be expected thatevena

few modal

⁽ⁱⁿ

Korschah

behavior) individuals

would show

traces of psychopathology in their social behavior, in spite of their essential modality in the

major

structure of personality. Deviant individuals would,

on

apriori grounds, seem

more

likely to include both the best ad^justed

and

worst

ad

justedindividuals. This studyhas not included, however,

any

systematic investigation of the distribution of levels of adjustment.

The

one extremeextrovert in thewholeadultsample

{M:^G=l:'l)

is a middle-aged

woman. She

is a clan mother, is widely respected

and

^liked,

and

is even something of a ^{social leader.}

Her

house is a popular meeting place for various social occasions: club meetings, parties, receptions.

Whether

she is completely happy, is to be doubted; both her Rorschach

and

her interpersonal behavior

show

signs ofstrain.

The

pointis, however, that she has functionedsue-

82 BUREAU

OF

AMERICAN ETHNOLOGY

[Bull. 150 cessfullj'^during

most

ofherlifeinasociety

which would

superficially

seem

to be

most

uncongenial.^®

Another

deviant case

{F

percent

>

^{60.5) is}anold

man who

iswidely respected as an herbalist

and

is personally well liked

by most

of the

community,

although he is feared

by some who

think he is a witch.

His Rorschach shows

in

him

the presence of

enormous

hostility; he hasdefiniteobsessive-compulsivecharacteristics.

But

hispersonality difficulties

do

not cripplehim.

He

has been able to sublimate

much

of his aggression in his herbal practice,

which

enables

him

to heal those

whom

in his fantasieshe has killed; his life is a never-ending search for

an

herb, a sure cure fortuberculosis,

which was

revealed to

him

in adream.

Thus

Tuscaroracultureprovides

ways

ofbehavior

which

areusable notonly

by modal

typesbutalso

by

a

wide

rangeof deviants; certain specialized rolesaremade-to-order safety valves for persons

who

are

"dangerously" deviant. These deviants can behave in a

manner

con- genial to their personalities,

and

at the

same

time in a fashion

which

is well adjusted to the functioning of the whole social system.

A

society produces not only deviant personality types, but provides socially tolerablepatterns of behaviorfor them.

CORRELATIONS BETWEEN MODAL PERSONALITY, NATIONAL