Tactics of Manipulation

Abstract

Manipulation is one means by which environments are altered to correspond to characteristics of individuals. We conducted two studies to identify the manipulation tactics that people use to elicit and terminate the actions of others. Factor analyses of four instruments revealed six types of tactics: charm, silent treatment, coercion, reason, regression, and debasement. Tactics of manipulation showed strong individual difference consistency across contexts. The charm tactic, however, was used more frequently for behavioral elicitation, whereas the coercion and silent treatment tactics were used more frequently for behavioral termination. Manipulation tactics covaried significantly across self-based and observer-based data sources with personality scales of Neuroticism, Extraversion, Ambitious-Lazy, Arrogant-Unassuming, Quarrelsome-Agreeable, and Calculating and with characteristics of subjects' social environments. We draw implications for an interactionist framework of person-environment correspondence, for an expansion of the taxonomic task that faces personality psychology, and for identifying links between personality and other scientific disciplines.

Full text

Journal
ol
Personality
and
Social
Psychology
1987.
Vol.
52,
No.
6.1219-1229
Tactics
of
Manipulation
David
M.
Buss
University
of
Michigan
Mary
Gomes
Stanford
University
Dolly
S.
Higgins
and
Karen
Lauterbach
Harvard University
Manipulation
is one
means
by
which environments
are
altered
to
correspond
to
characteristics
of
individuals.
We
conducted
two
studies
to
identify
the
manipulation tactics that people
use
lo
elicit
and
terminate
the
actions
of
others.
Factor
analyses
of
four
instruments revealed
six
types
of
tactics:
charm,
silent treatment, coercion, reason, regression,
and
debasement.
Tactics
of
manipulation
showed strong individual
difference
consistency
across
contexts.
The
charm tactic, however,
was
used
more
frequently
for
behavioral elicitation, whereas
the
coercion
and
silent treatment tactics were
used
more
frequently
for
behavioral
termination. Manipulation
tactics
covaried
significantly
across
self-based
and
observer-based
data
sources
with
personality scales
of
Neuroticism,
Extraversion.
Ambitious-Lazy, Arrogant-Unassuming,
Quarrelsome-Agreeable,
and
Calculating
and
with
char-
acteristics
of
subjects'
social
environments.
We
draw implications
for an
interactionist
framework
of
person-environment
correspondence,
for an
expansion
of
the
taxonomic
task
thai
faces
personality
psychology,
and for
identifying
links
between personality
and
other scientific disciplines.
Natural selection favors
people
who
successfully manipulate
objects
in
their environment. Some
manipulable
objects
are in-
animate, such
as the raw
materials
used
to
build
shelters,
tools,
clothing,
or
weapons. Other manipulable objects
are
alive.
These
include
predators
and
prey
of
different
species
as
well
as
mates, parents, children,
rivals,
and
allies
of the
same species.
Manipulation
of
living
objects
may be
defined
as the
various
means
by
which organisms
influence
and
exploit
the
sense
or-
gans
and
behavioral machinery
of
other organisms (Dawkins
&
Krebs,
1978;
Krebs&
Dawkins, 1984).
Among
group-living
species
such
as
ours,
manipulation
of
conspeeifics
is
especially
important. People
who
lack
the
ability
to
manipulate others
may
fail
to
elicit
parental
care, acquire
resources,
establish
reciprocal
alliances, elevate
in
hierarchies,
or
attract
mates.
Existing
people
had
ancestors
who
successfully
manipulated members
of
their
own
species
in
these ways.
Peo-
ple
who
lacked
such
skills
are no
one's
ancestors.
By
what
tactics
do
humans
manipulate
one
another?
Under
what
conditions
are
different
tactics
displayed? What
are the
relative frequencies
of
alternative manipulation
tactics?
And
how
do
others
respond
to
specific manipulative attempts? These
and
related
questions address
a
central issue
that
has
galvanized
the
field of
personality psychology
for the
past
2
decades:
What
is
interactionism
and how can we
best
conceptualize
and
empir-
ically examine links between features
of
people
and
features
of
their environments?
Person-Environment
Correspondence
In
the
decade
since Magnusson
and
Endler
(1977)
published
their
volume
on
interactionism
in
personality psychology, there
This study
was
supported
in
part
by
National Institute
of
Mental
Health Grant
MH41593-02
to
David
M.
Buss.
We
thank Lewis
R.
Goldberg
for
exceptionally
helpful
suggestions.
Correspondence
concerning this article should
be
addressed
to
David
M.
Buss, Department
of
Psychology, University
of
Michigan,
580
Union
Drive.
Ann
Arbor, Michigan
48109-1346.
has
been
strong
consensus about
the
importance
of
both person
and
environment variables.
No
compelling conceptual
frame-
work, however,
has
emerged
to
incorporate
their
interaction.
The
dominant response
to
calls
for
interactionism
has
been
the
analysis
of
variance
(ANOVA)
framework,
in
which
interaction
is
conceived
as
nonadditive statistical interaction
of
person
and
situation variables crossed
in
experimental design. Despite
its
frequent
use,
enthusiasm
for the
ANOVA
approach
has
waned
as
telling limitations have been
noted
and
documented empiri-
cally
(e.g.,
Ekehammar,
1974;Golding,
1975).
One
limitation
is
that studies
can be
constructed
at
will
to
manipulate variance attributable
to the
person component,
the
situation
component,
or the
interaction component. Selection
of a
weak situational manipulation,
for
example,
results
in
less
variance attributable
to the
situation,
whereas selection
of an
inappropriate, ill-conceived,
or
poorly measured person
vari-
able attenuates variance attributable
to
persons.
A
second limi-
tation
is
that crossing
levels
of
persons
with
levels
of
environ-
ments,
an
essential
part
of the
ANOVA
paradigm,
does
not
often
occur
in
nature. Outside
the
psychological
laboratory,
people
are
rarely randomly assigned
to
conditions.
But
perhaps
the
most telling limitation
is
that interactions
in the
A\OVA
sense
do not
capture
the
dynamic interchange
and
mutual
influence
between people
and
environments
that
most psychologists
view
as
central features
of the
concept
of
interaction.
An
alternative
approach
is to
identify
the
links
between
fea-
tures
of
people
and
features
of
their
environments
that
occur
in
everyday
life
(Buss. 1984b, 1985a). This
person-environment
correspondence framework
has
Ihree
essential
components.
The first is
descriptive
and
involves
documenting
empirically
the
nature
and
domains
of
person-environment correspon-
dence that occur
in
people's
lives.
The
second component
is
causal
and
involves
identifying
the
mechanisms
and
specific
ac-
tions that
are
responsible
for
producing obtained
person-envi-
ronment
correspondences.
The
third
component
entails
exami-
nation
of the
consequences
that
follow
from
obtained
person-
environment
links.
1219

1220BUSS. GOMES,
H1GGINS,
AND
LAUTERBACH
Three
Mechanisms
of
Person-Environment
Correspondence
Three essential mechanisms
are
posited
to
causally produce
person-environment correspondence
(Buss,
1985a): selection,
evocation,
and
manipulation. Selection involves nonrandom
choices
of
interpersonal
and
physical milieus. Mate selection
is
a
dramatic example
of the
importance
of
this mechanism
in
producing
person-environment correlations (Buss.
I984b).
Nonrandom
selection
of a
mate
results
in
subsequent exposure
to a
prolonged
act
environment that shows stability over time
(Buss.
1985b;
see
also
Snyder,
1981;
Snyder
&
Gangestad,
1982).
The
second person-environment mechanism, evocation,
may
be
denned
as
nonrandom
and
unintentional elicitation
of
reac-
tions
from
the
environment. Researchers have conceptualized
evocation
in
the
context
of
behavioral genetics
and
developmen-
tal
psychology (Plomin,
DeFries,
&
Loehlin,
1977;
Scarr
& Mc-
Cartney,
1983)
and
have studied
it
empirically
in the
context
of
parent-child interactions
(Buss,
1981). Highly active children,
for
example, appear
to
evoke
"upper
limit
control"
behavior
from
parents
that
is
designed
to
reduce
the
noise
and
intensity
that
such
children
typically generate. Less
active
children
do
not
elicit
such responses
and so
inhabit
a
more quiescent
and
peaceful
interpersonal milieu.
Manipulation,
the
third person-environment mechanism,
is
denned
by the
tactics used intentionally
to
coerce,
influence,
change,
invoke,
and
exploit
the
environment.
No
insidious
or
malevolent
intent
need
be
implied
by the
mechanism
of
manip-
ulation.
Conceptually, manipulation
is the
broadest mechanism
of
person-environment correspondence
because,
in
principle,
there
are
counlless actions that
a
person could
use to
influence
the
nature
of the
environment subsequently inhabited. Manip-
ulation
differs
from
selection
in
that selection involves choosing
to
enter existing
habitats,
whereas manipulation entails altering
those
environments
already
inhabited.
Several
areas
of
personality research appear
to
deal
with
the
ways
in
which
people shape
or
manipulate their environments.
The
work
on
Machiavellianism (Christie
&
Geis,
1970),
for ex-
ample,
has
explored
a
personality style that
is
characterized
by
manipulativeness,
cynicism
about
human nature,
and
shrewd-
ness
in
interpersonal behavior.
In the
context
of
laboratory
ex-
periment
games, those high
in
Machiavellianism
appear
to
dis-
play
"an
acute
and
opportunistic sense
of
timing"
(p.
159)
and
appear able
to
capitalize especially
on
situations containing
am-
biguity
regarding rules.
In a
different
context, Patterson (e.g.,
Patterson
&
Bechtel,
1977)
has
examined coercion
in the
con-
text
of
families
and
identified discriminative stimluli that mod-
erate
the use of
coercive behaviors.
Both
the
evolutionary perspective
and the
person-environ-
ment
correspondence framework point
to the
importance
of
manipulation
as a
central
process
in
social interaction.
These
perspectives
are
valuable
in
guiding
research
toward this impor-
tant
line
of
inquiry.
However,
in
spite
of
offering
a
valuable heu-
ristic,
neither perspective
offers
specific
predictions
about
which
tactics
of
manipulation
will
be
used,
in
what contexts
they
will
be
used,
how
effective
they
will
be, or
precisely
how
each tactic
will
be
linked with specific features
of the
social
en-
vironment.
Manipulation
implies
influence
for a
reason, purpose,
or
goal.
In
principle, manipulation tactics
should
varv
with
the
goals toward which
they
are
directed.
Tactics
used
with
the
boss
to
obtain
a
higher salary would
be
expected
to
differ
from
those
used
with
the
spouse
to
obtain
a
backrub
or
with
a
friend
to
obtain
the use of a
car.
For
this
first
empirical
probe,
we
chose
two
broad
conditions
in
which
to
study
tactics
of
manipulation:
(a)
behavioral instigation,
or
tactics used
to get
another
to do
something,
and (b)
behavioral
termination,
or
tactics
used
to
get
another
to
stop doing something. Close
relationships
were
selected
as the
context
in
which
to
study
tactics
of
behavioral
instigation
and
termination.
The
basic
purposes
of the
study
were
as
follows:
(a) to
provide
a
first
empirical examination
of
tactics
of
manipulation
that
are
used
for
behavioral instigation
and
termination
in the
context
of
close relationships;
(b) to
identify
the
relative
frequencies
with
which each
tactic
is
performed;
(c) to
identify
performance
frequency
differences
in
manipulation tactics between behav-
ioral
instigation
and
termination;
(d) to
examine
the
cross-situ-
ational consistency
of
individual
differences
in
the
use
of ma-
nipulation
tactics
between
the
instigation
and
termination
con-
texts;
(e)
to
identify
the
links between more
traditionally
assessed
personality
variables
and
tactics
of
manipulation;
(f)
to
examine
person-environment
correlations
for
manipulation
tactics
in the
form
of
couple correspondence;
and (g) to
identify
the
connections
between
use of
specific
manipulation tactics
and the act
environment that
people
inhabit.
Preliminary
Study:
Nominations
of
Acts
of
Influence
Subjects
Sixty-seven college
undergraduates
participated
as
subjects
for the
preliminary study
of
nominating acts
of
influence
in the
context
of
close
relationships.
Procedure
Each subject received
one of two
nomination
forms.
I he first
con-
tained
this instructional set: "When
you
want
to get
your
romantic
part-
ner
to do
something, what
do you do
(please
list
specific
actions)?"
The
second nomination
form
contained these instructions: "When
you
want
to get
your partner
to
slop doing something, what
do you do
(please
list
specific
actions)?"
The
goal
of
these procedures
was to
obtain
a
reason-
ably
diverse
set of
influence
tactics
in
the
context
of
close
relationships.
We
examined
the
large number
of
nominations
for
redundancy.
meaning,
coherence,
and
thematic
content.
Three
judges
nominated
the
major categories within which
the
tactics
of
influence
belonged.
Allow-
ing for
slightly
different
labeling
of
similar content
categories,
the
three
judges nominated
five
categories
in
common
and two
distinctive
catego-
ries.
The five
common
categories were retained
for the
main
study.
The
three judges then independently nominated
the
seven best exem-
plars
for
each
category.
Nominations
that
received
at
least
two
endorse-
ments
were
retained.
This
produced
a
35-act
instrument
{see
Buss
&
Craik,
1983,
1984)
consisting
of
seven
act
exemplars
for
each
of
the
five
categories:
(a)
reason
lactic
(e.g.,
"I ask her to do
it"),
(b)
rc#/T.U7<w
lactic (e.g.,
"I
whine
until
she
does
it"),
(c)
coercion
taclic
(e.g..
"I de-
mand
that
she
does
it"),
(d)
charm
lactic
(e.g..
"I
compliment
her so
that
she
will
do
it"),
and (e)
debasement
lactic
(e.g..
"I
debased
myself
so
that
she
would
do
it").
Two
structurally analogous forms were developed.
The first
contained
the 35
actions
in the
form
of
behavioral
instigation,
or
getting
the
part-
ner to do
something.
The
second
form
contained
the
same
set of
tactic
stems
but
ended
with
the
goal
of
behavioral
termination.
For
example.

TACTICS
OF
MANIPULATION
1221
the
act "I
curse
at her
until
she
does
it"
became
"I
curse
at her
until
she
stops doing it."
The
complete instrument
for the
behavioral instigation
condition
is
provided
in the
Appendix.
Main
Study:
Assessment
of
Manipulation
Tactics
Method
Subjects
Subjects
for the
main study were
118
undergraduates composing
59
dating
couples. Couples were recruited through ongoing classes,
fliers
placed
in
dormitories,
and
announcements
in the
student newspaper.
Because
our
goal
was to
examine tactics
of
manipulation
in the
context
of
close relationships,
we
asked subjects
how
long
they
had
dated
(a
requirement
of at
least
6
months
of
prior dating
was
imposed).
Materials
Among
a
larger
battery
of
tests
and
assessment measures,
we
used
the
following
instruments
for
this study.
Tactics
of
manipulation. Four
different
forms
of the
35-act instru-
ment
described
in
the
preliminary study
were
generated
from
a 2 X 2
matrix.
The first
axis consisted
of the
goal
or
context
of the
influence,
behavioral instigation
or
behavioral termination.
We
developed
a
self-
report version
of
each form
as
well
as a
structurally analogous observer-
report version. Four instructional sets were
developed,
one for
each ver-
sion.
For
example,
the
self-report version
of the
behavioral instigation
form
contained these instructions:
When
you
want
to get
your partner
to do
something
for
you,
what
are you
likely
to do?
Look
at
each
of the
items listed below
and
rate
how
likely
you are
to
do
each when
you are
trying
to
gel
your
partner
to do
something
None
of
them
will
apply
to all
situations
in
which
you
want
your partner
to do
something,
so
rate
how
likely
you
are,
in
general,
to do
what
is
described.
If you are
extremely likely
to
do it,
circle
a 7. If you are
not
at
all
likely
to do it,
circle
a
I.
If you
are
somewhat
likely
to do it,
circle
4.
Give intermediate ratings
for
intermediate likelihoods
of
performing
the
behaviors.
Eysenck
Personality
Questionnaire
fEPQ).
The
EPQ
(Eysenck
&
Eysenck,
1975)
is a
self-report instrument that contains
90
true-false
items.
Four scales
are
scored
from this instrument: Extraversion, Neu-
roticism, Psychoticism,
and a Lie
scale
designed
to
detect
dissembling.
Interpersonal
Adjective
Scales
(IAS).
The
Interpersonal
Adjective
Scales
(128-item
version; Wiggins,
1979)
were developed
to
represent
a
reasonably comprehensive taxonomy
of the
interpersonal domain
in the
form
of a
circumplex
structure.
The
16
scales, each containing eight
adjectives,
are
Dominant, Ambitious, Extraverted.
Gregarious,
Agree-
able, Warm, Ingenuous, Unassuming, Submissive, Lazy, Introverted,
Aloof,
Quarrelsome, Cold, Calculating,
and
Arrogant.
In
self-report
form,
subjects
rate
how
characteristic
or
uncharacteristic each adjective
is
on a
9-point scale. Adjectives
are
intermingled
and are not
identified
by
the
scale
to
which they
belong.
Interviewer
judgments
about couple
relationship.
Each couple
was
interviewed
by a
pair
of
interviewers drawn from
an
eight-member
team. Each interview
lasted
about
30
min.
A
dozen
standard
questions
were
posed
to
each
couple.
Questions
posed
by the
interviewers
in-
cluded
the
following:
How did you
meet? What
are the
similarities
and
differences
between you?
Do
youthinkyouwillbe
together
5 or
10
years
from
now?
In
addition
to the
standard
questions, interviewers were
trained
to
probe
further
into issues raised during
the
course
of the
inter-
view.
Directly
following
each
interview,
the two
interviewers indepen-
dently
rated each couple
on a set of
relationship variables:
How
well
matched
do you
think
the
couple
is? How
similar
is the
couple?
Who
has
more
power (who takes
more
control)?
How
long lasting
will
the
relationship
be?
Ratings were
made
on
7-point
scales.
To
obtain
a
more
reliable
assessment
of
each
couple
relationship
variable,
we
composited
the
scores
for the two
interviewers.
The
correlations between
the
inde-
pendent
interviewers
were
.65
(how
well
matched).
.44
(hov*
similar),
.57
(who
has
more power),
and .53
(probability
of
termination).
Procedure
Subjects completed
the
self-report versions
of the
tactics
of
manipula-
tion, IAS,
and FPQ in
their spare time. Couples were
then
tested
in
groups ranging
from
4
(two couples)
to 12
(six
couples)
individuals.
Couples were
separated
for the
duration
of the
testing session
to
prevent
discussion
of the
instruments.
In
this testing
session,
subjects
completed
the
observer
forms
of the
tactics
of
manipulation
and
were
interviewed.
Total
confidentiality
was
assured
for all
responses.
Not
even
the
subject's
partner could
see the
responses without expressed
written
permission.
Results
Sex
Differences
in
Manipulation
Tactics
We
conducted
the first set of
analyses
to
examine
whether
significant
sex
differences
existed
in the use of
tactics
of
manip-
ulation.
T
tests
for sex
differences
were
conducted
for
each
of
the 35
acts
of
manipulation
for
each
of the
four
instruments.
Although
significant
sex
differences
occasionally
emerged
for a
given
instrument,
only
one
act
showed
significant
sex
differ-
ences
across
more
than
one
data
source.
For
example,
the
item
"He or she
whines
until
I do it"
showed
greater
female
than
male
performance
frequencies
for the
observer
data
source
in
the
instigation
condition,
r(90)
=
2.82,
p <
.006. Similarly,
the
act
"I
allowed
myself
to be
debased
so
that
he or she
would
do
it"
showed
significantly
greater
male
than
female
performance
for
the
self-report
data
source
in the
termination
condition,
i(108)
=
2.30,
p<
.025.
However,
neither
of
these
sex
differences
was
replicated
in any of the
other
conditions
or for the
other
data
source.
Only
the act "I
repeated
the
request
from different
angles"
showed
significant
sex
differences
across
more
than
one
data
source
(observer
data
sources
for the
instigation
and
term
i-
nation
conditions),
but
even
this
sex
difference
was not
repli-
cated
in the
remaining
conditions.
We
concluded
that
in
this
sample,
at
least,
strong
sex
differences
that
replicate
across
con-
ditions
and
data
sources
do not
exist.
Factor
Analyses
of
Manipulation
Tactics
To
identify
the
major
dimensions
along
which
tactics
of ma-
nipulation
vary
and to
confirm
the
initial
rational
item
group-
ing,
we
conducted
four
separate
factor
analyses
using
varimax
rotation,
one for
each
of the
four
manipulation
instruments.
Inspection
of the
factor
loadings
revealed
highly
similar
factors
and
loadings
across
the
four
instruments.
Six
major
factors
were
identified
across
the
four
instruments
and
analyses:
Charm,
Si-
lent
Treatment,
Coercion,
Reason,
Regression,
and
Debase-
ment.
All but
Silent
Treatment
were
also
rationally
generated
a
priori.
The
four
highest
loading
acts
on the
Silent
Treatment
factor
had
been
earlier
classified
intuitively
as
Regression.
The
factor
loadings
for the
four
separate
factor
analyses
are
shown
in
Table
1.
A
seventh
factor
emerged
on one of the
factor
analy-
ses.
It
appeared
to be a
"reciprocity"
tactic
involving
two
items:
"I
give
up
something
so
that
he or she
will
do it" and "I
give
him or her a
small
gift
so
that
he or she
will
do
it."
Because
of

1222
BUSS.
GOMES,
HIGG1NS,
AND
LAUTERBACH
Table
1
Factor
Loadings
of
Tactics
of
Manipulation
Specific
act
I
compliment
her so
she'll
do it
(stop
it)
I
act
charming
so
she'll
do it
(stop
it)
I
try to be
loving
and
romantic
when
1 ask her
I
give
her
a
small
gift
or
card
before
I ask
I
tell
her
I'll
do her a
favor
if
she'll
do it
I
don't
respond
to her
until
she
does
it
(stops
it)
I
ignore
her
until
she
does
it
(stops
it)
I
am
silent
until
she
agrees
to do it
(stop
it)
1
refuse
to do
something
she
likes
until
she
does
it
I
demand
that
she do it
(stop
it)
I
yell
at her
until
she
does
it
(stops
it)
I
criticize
her for not
doing
it
(stopping
it)
I
curse
at her
until
she
does
it
(stops
it)
I
threaten
her
with
something
if
she
doesn't
do it
I
give
her
reasons
for why she
should
do it
(stop)
I
ask her why she
doesn't
do it
(stop
it)
I
point
out all the
good
things
that
will
come
from
doing
it
(stopping
it)
I
explain
why I
want
her to do it
(stop
it)
I
show
her
that
I
would
be
willing
to do it for her
I
pout
until
she
does
it
(stops
it)
I
sulk
until
she
does
it
(stops
it)
I
allow
myself
to be
debased
so
she'll
do it
1
lower
myself
so
she'll
do it
(stop
it)
I
act
humble
so
she'll
do it
(stop
it)
Instigation
Self
Observer
Charm
tactic
.70
.85
.86
.47
.35
Silent
treatment
tactic
.79
.82
.79
.43
Coercion
tactic
.61
.80
.55
.78
.60
Reason
tactic
.79
.58
.78
.76
.45
Regression
tactic
.76
.75
Debasement
tactic
.71
.86
.30
.73
.75
.72
(.26)
(.25)
.90
.87
.70
.57
.69
.52
(.18)
(.17)
.49
.43
.69
.69
(.23)
.65
.55
.65
.77
.50
(.23)
Self
.64
.74
.70
.56
.46
.82
.76
.79
.65
.62
.66
.62
.57
.42
.77
.61
.70
.48
.55
.75
.82
.82
.87
.42
Termination
Observer
,7S
.80
.76
.56
.42
.84
.80
.88
.47
.73
.84
.61
.58
.55
.86
.64
.67
72
(.34)
.70
71
.79
80
.45
Note.
For
expositional
clarity,
only
the
male
version
of the
acts
are
presented.
Factor
loadings
in
parentheses
reflect
those
for
which
the
highest
loading
occurred
on
another
factor.
its
lack
of
robustness across conditions
and
data sources, this
seventh
factor
was not
carried forward
in
subsequent analyses.
Because each
of the 35
acts
of
manipulation
are
keyed
in the
same direction,
it is
possible that individual
differences
in the
use
of the
response scale
may be
confounded
with
scores
on
various
tactics.
An
alternative
data
analytic strategy,
therefore,
would
be to
examine subjects' relative
use of a
tactic compared
with
his or her use of
others. Note that
this
alternative approach
addresses
a
different
set of
questions
and
ignores
a
subject's
overall
elevation
or
depression
on
tactic
use.
For
example,
a
man
who
reported performing,
and
whose partner
reported
him
performing,
many
acts
of
manipulation,
but
whose
use of
coercion
was
slightly
higher
than
his use of
regression, would
receive
similar scores upon
transformation
as
would
a
male sub-
ject
who
reported (and whose partner reported)
that
he
per-
formed
few
acts
of
manipulation,
but
whose coercion
use was
nonetheless
slightly
higher than regression use.
Nonetheless,
this alternative data
analytic
strategy
was ex-
plored.
We
standard scored responses
for
each subject across
the 35
acts
of
manipulation
for
each data source
and for
each
of
the two
conditions.
These
transformed scores
were
then
factor
analyzed
in the
same
manner
as
were
the
untransformed
scores.
The
results proved
difficult
to
interpret.
No
clear
factors
emerged across
data
sources.
Items that were assigned
a
priori
to the
same scale, such
as
coercion (e.g.,
"I
yell
at him so he
will
do
it";
"I
curse
at her so
she
will
do
it"),
and
which
loaded
on
the
same
factor
in the
untransformed factor analysis, loaded
on
different
factors
in
some
of the
factor
analyses
using
the
trans-
formed
scores
(these analyses
are
available
from
David
M.
Buss).
For
subsequent analyses,
we
elected
to use
composites con-
structed
on the
basis
of the
results
of the
factor
analyses
of the
untransformed
scores. Four composites were computed
hy
summing
the
acts shown
in
Table
1 for
each
tactic—one
com-

TACTICS
OF
MANIPULATION
1223
Table
2
Frequencies
of
Tactic
Performance
Across
Conditions
Manipulation
tactic
Reason
Reason
Reason
Reason
Charm
Charm
Charm
Charm
Regression
Regression
Regression
Regression
Coercion
Coercion
Coercion
Coercion
Silent
treatment
Silent treatment
Silent
treatment
Silent
treatment
Debasement
Debasement
Debasement
Debasement
Condition
Instigation
Instigation
Termination
Termination
Instigation
Instigation
Termination
Termination
Instigation
Instigation
Termination
Termination
Instigation
Instigation
Termination
Termination
Instigation
Instigation
Termination
Termination
Instigation
Instigation
Termination
Termination
Data
source
Self
Observer
Self
Observer
Self
Observer
Self
Observer
Self
Observer
Self-
Observer
Self
Observer
Self
Observer
Self
Observer
Self
Observer
Self
Observer
Self
Observer
M
5.31
5.38
5.28
5.05
3.13
3.11
2.51
2.56
2.17
2.27
2.39
2.28
1.81
1.71
2.40
2.35
1.79
1.89
2.04
2.47
1.44
1.35
1.42
1.27
HD
1.12
0.86
1.05
1.26
.28
.26
.09
.19
.23
.43
1.41
.39
0.98
0.85
.23
.24
.02
.11
.50
.53
0.82
0.82
0.82
0.63
Nole.
Means
and
standard
deviations
shown
are
divided
by the
number
of
acts
composing
each
composite
tactic
so
that
relative
frequency
can
be
evaluated.
posite
each
for the two
data
sources crossed
by the two
condi-
tions.
Note that
the
present
use of two
data
sources minimizes
the
importance
of
response-style problems. Subsequent analy-
ses
can
examine
the
robustness
of
relationships
that
emerge
across
data
sources that
do not
share
the
same response
style.
Relative
Performance
Frequencies
Across
Conditions
and
Data
Sources
To
compare
the
relative frequencies with which each major
manipulation
tactic
was
reported
to be
performed,
we
divided
the
means
and
standard
deviations
for
each
of
the
24
compos-
ites
by the
number
of
items forming each
one.
These
results
appear
in
Table
2. As
shown
in
Table
2. the
reason tactic showed
the
highest performance
frequency
across
all
conditions
and
data
sources.
This
was
especially apparent
in the
following
acts:
"I
asked
him or her to do
it";
"I
explained
why I
wanted
her or
him to do
it";
"1
gave reasons
why he or she
should
do
it";
and
"I
point
out all the
good
things that
will
come
from
doing it."
The
charm tactic
was the
second highest
in
performance
fre-
quency (overall
M =
2.81).
This
was
followed
by the
regression,
coercion,
and
silent treatment tactics,
which
had
overall
means
of
2.28, 2.07,
and
2.05,
respectively.
The
debasement tactic
composites showed
the
lowest performance
frequencies
across
all
conditions
and
data sources.
Especially
rarely
performed
manipulation acts
were
"I
allow
myself
to be
debased
so
that
he
or she
will
do it" and "I
lower
myself
so
that
he
or
she
will
do
it."
Agreement
Between
Self
and
Partner
Data
Sources
Agreement
or
lack
of
agreement between users
and
recipients
of
manipulation tactics
poses
interesting conceptual issues
that
preclude interpretation
as
simple
reliability
coefficients.
Recipi-
ents
may
most
accurately perceive tactics
when
they
are
least
effective.
Recipients
may be
impervious
to
tactics that work.
Similarly,
acts
of
influence
may
become
habitual
so
that
not
even
the
actor
has
accurate self-knowledge. Some tactics
may
be
more readily observable than others
and
thus
lead
to
greater
agreement between self
and
observers. Table
3
shows
corre-
lations between
self-
and
partner ratings
of the six
tactics under
conditions
of
instigation
and
termination.
Also
shown
are the
conditions
for a
total manipulation score generated
by
sum-
ming
across
the 35
acts
of
manipulation.
The
mean agreement across tactics
for the
instigation
condi-
tion
was
.42,
whereas
the
mean agreement
for the
termination
condition
was
only .25. Thus there appeared
to be
greater
agree-
ment between self
and
partner
on the
frequency
of
instigation
tactics when compared with termination tactics. This
finding
was
especially apparent
with
the
reason,
charm,
and
debase-
ment tactics;
the
silent treatment,
coercion,
and
regression
tac-
tics showed approximately
equivalent
levels
of
agreement.
Across conditions,
the
coercion tactic
received
the
highest
self-
Table
3
Agreement, Consistency,
and
Context
Differences
for
Tactics
Agreement3
Tactk
Charm
Silent
treatment
Coercion
Reason
Regression
Debasement
Total
scored
Instigation
.36"*
.28"
.53"*
.77"*
.33"
.27*
.44***
Termination
.11
29**
.53*"
.17
.28"
.12
.42***
Consistency*
Self
.72***
.71***
.78***
.80***
.79***
.89***
.91*"
Observer
.71*"
.68*"
.80"*
.74*"
.74*"
71"*
.90*"
Context0
Self
5.58"*
4.99***
7.09*"
0.4ft
1.96
0.51
2.28*
Observer
6.99*"
4.89*"
7.67***
2.59*
0.51
1.10
0.51
a
Agreement
signifies
correlation
between
self
and
partner
observer,
b
Consistency
signifies
correlations
across
instigation
and
termination contexts.
1
Context
signifies
(tests
between
the
instigation
and
termination
conditions.
d
Total
score
consists
of the sum of
all
35
acts
of
manipulation.
*p<-05.**p<-01.***p<.001.

1224
BUSS.
GOMES,
HIGGINS,
AND
LAUTERBACH
Table
4
Correlations
Among Composited Tactics:
Same
Data Source
Tacuc
1.
Charm
2.
Silent
treatment
3.
Coercion
4.
Reason
5.
Regression
6.
Debasement
1
2
—
.41***
.25**
—
.28"
.58***
.29**
.15
.26**
.68***
.31**
.23*
3
.29**
.56***
—
.26*
.51***
.30**
4
.40*"
.21*
.18
—
.14
.01
5
.26"
.52***
.53*"
.15
—
.31**
ft
.41"*
48***
.49***
"O*
.14***
—
A'o/t1.
Correlations
among
tactics
assessed
through
self-reports
appear
above
the
diagonal;
those
assessed
through
the
observer
data
source
appear
below
the
diagonal.
*p<.05.**/)<
01.
***;;<.001.
other
agreement,
whereas
the
debasement tactic received
the
lowest
self-other agreement.
The
total manipulation
score
showed
modest agreement
across
data sources.
In
sum, cross
data-source
agreement appeared modest overall,
slightly
higher
for
instigation
than
for
termination,
and
highest
for the
coer-
cion
tactic across conditions.
Cross-Context
Consistency
of
Individual
Differences
Also
shown
in
Table
3 are the
cross-context consistency
co-
efficients
for the six
tactics
and
total score
and the two
data
sources
separately.
These
reflect
the
degree
to
which rank order
of
individual
differences
in
frequency
was
maintained
across
conditions.
The
individual tactic consistency
coefficients
were
reasonably
high, ranging
from
.89 to
.71.
The
total scores
showed
consistencies
of
.91
and
.90. Overall,
it is
clear that indi-
viduals
do
maintain
well
their
relative
positions
in
performance
frequencies
of
manipulation
tactics across conditions
of
instiga-
tion
and
termination.
Cross-Context
Differences
in
Performance
Frequencies
One
intriguing
issue
is
whether
there were main
effects
owing
to
condition.
Did the
sample
as a
whole
use
different
tactics
of
manipulation
in the
context
of
behavioral instigation when
compared
with
the
context
of
behavioral termination?
The
(
tests
shown
in
Table
3
suggest that
the
answer
is
affirmative.
The
charm
tactic
was
more
often
used
for
behavioral instigation.
Table
5
Correlations
Among Composited
Tactics:
Separate
Data
Sources
Observer-repon
data
source
Self-report
data
source
1
1.
Charm
— .02
2.
Silent
treatment
.19 —
3.
Coercion
.11
.26
4.
Reason
.00 .15
5.
Regression
.22*
.34**
.20*
.10
.24* .19*
.37"*
.15
.32"
.11
— .04
.51***
.20*
.21*
—
.23* .21*
.38*"
.07 — .05
ft. Debasement
.16
.46*
.44*
-.02
.46*1
•p<.05.
";x.OI.
"/J<.001.
The
silent treatment
and
coercion tactics
were
more
often
used
for
behavioral termination. Thus,
it
appears that
subjects
used
aversive
controls
to
terminate unwanted
behavior
of
others,
whereas
they used reward
differentially
to
instigate desired
ac-
tions
of
others.
The
fact
that these results occur
with
approxi-
mately equal magnitude
for
each
of the two
data sources sepa-
rately
lends
credence
to
their robustness.
Correlations
Among
Composites
Table
4
shows
the
correlations among
the
composite manipu-
lation
tactics
for the
self-reported data source (above
the
diago-
nal)
and the
partner-reported data source (below
the
diagonal)
for
the
instigation condition (correlations
for the
termination
condition were similar
and may be
obtained
from
David
M.
Buss).
In
general, there
was a
positive
manifold
in the
matrix
such
that most tactics
of
manipulation
were
positively
corre-
lated
with
one
another.
One
possibility
is
that
a
response
style
may
have
inflated
the
magnitude
of
these correlations.
To
examine
this
hypothesis,
we
correlated
the
self-reported tactics
with
the
ob-
server-based
tactics.
These correlations appear
in
Table
5.
As
shown
in
Table
5,
there
was
again
a
positive
manifold
be-
tween
the
various tactics
when
assessed
through
different
data
sources.
The
magnitudes
of
these positive correlations
were
slightly
lower
than analogous correlations between tactics measured
through
the
same
data
source, suggesting
the
possible operation
of
a
response
style.
However,
the
overwhelming positive
manifold
between tactics measured through
separate
data
sources also sug-
gested
a
substantive
interpretation—namely
that people
who are
above
the
mean
on
performing
one
tactic tend
to be
above
the
mean
on
performing
a
variety
of
other tactics.
This
appeared espe-
cially
true
of the
relations between
the
regression tactic
and the
tactics
of
coercion,
debasement,
and
silent treatment.
In
contrast,
the
charm
and
reason
tactics were less strongly correlated
with
the
other
tactics
of
manipulation.
Correlations
With
Standard
Personality
Variable's
Tables
6 and 7
show
the
correlations
between
the six
manipu-
lation
tactics
and
total score
with
the EPQ and the
IAS.
respec-
tively.
Because
the
instigation
and
termination tactics
were
highly
correlated
(Table
3), and to
conserve
space,
only
the
cor-
relations
for the
instigation tactics
are
shown
in
these
and
subse-
quent
tables (the entire
set of
correlations
may be
obtained
from

TACTICS
OF
MANIPULATION
1225
Table
6
Eysenck
Personality
Questionnaire
X
Tactics
of
Manipulation (Instigation)
Tactic
Charm
Silent
treatment
Coercion
Reason
Regression
Debasement
Total
score
*/><.05.
***/;<
.Odl.
Self
data
.24*
.05
.11
.20*
.05
.09
.18
Extraversion
Observer
data
.16
-.03
-.09
-.06
.13
.22*
.05
Self
data
.08
.33***
.20*
.01
.28**
.25*
.24*
Neuroticism
Observer
data
-.01
.24*
.11
.00
.19*
.09
.12
Ps\
Self
data
.1
1
.(14
.08
.10
.05
.08
.10
'choticism
Observer
data
.!«*
.(14
.07
.19*
.14
.16
.12
David
M.
Buss).
Correlations
are
shown separately
for the
self-
reported
and
other-reported manipulation tactics.
As
shown
in
Table
6, the
Neuroticism scale
of the
EPQ
corre-
lated
significantly
across
data
sources
with
tactics
of
manipula-
tion.
Specifically,
those
who
scored
high
on
Neuroticism tended
to
perform high frequencies
of
silent treatment
and
regression
manipulation tactics.
Table
7
shows
correlations
between
the
self-
and
other-re-
ported manipulation tactics
and the
IAS. Noteworthy
are
those
relations that were
significant
for
both
data
sources.
Use of the
charm
tactic
was
correlated
across
data
sources with
the IAS
Calculating scale.
Use of the
silent treatment tactic
was
corre-
lated
positively
across
data
sources with
the IAS
Calculating,
Cold,
and
Quarrelsome scales
and
negatively
with
the
Agree-
able
and
Gregarious scales.
Both
the
coercion
and
regression
tactics
were correlated
across
data
sources
with
the IAS
Arrogant (positive)
and
Unas-
suming
(negative) scales.
The
coercion
tactic
showed additional
correlations with
the
Quarrelsome
(positive)
and
Agreeable
(negative)
scales. Interestingly,
the
debasement tactic covaried
significantly
with
the
Lazy (positive)
and
Ambitious (negative)
scales.
These
findings
suggest that those
who are
relatively
sloth-
ful
and
lacking ambition
influence
others
by
subjugating them-
selves;
the
more ambitious appear
to
avoid self-subjugation
and, instead,
use
reason
as a
tactic
of
influence.
In
sum, there were clear links between
standard
personality
dimensions
and the use of
specific
sorts
of
manipulation
tactics.
Use of the
silent treatment
was
linked with neuroticism
and
with
the
quarrelsome quadrant
of the
Wiggins circumplex.
Use
of
the
debasement
tactic
was
linked with
the
lazy
or
nonascen-
dant
quadrant
of the
circumplex.
And use of
reasoning
to in-
fluence
others
was
linked with
the
ambitious
and
ascendant
quadrant
of the
circumplex. Interestingly,
the
total manipula-
tion
score
was
positively correlated, with
lASs
calculating
across
both
data
sources.
Correlations
With
Couple
Relationship
Variables
As
shown
in
Table
8,
manipulation
tactics
also correlated
across
data
sources
with
couple relationship variables,
as
judged
by
male
and
female interviewers
who had no
knowledge
of
the
manipulation scores
of the
subjects. Perhaps
the
most
striking
findings of
Table
8 are the
pervasive
negative
corre-
lations
between
all
manipulation tactics
and the
variables
i«V/
matched
and
couple similarity.
The
less
well-matched
and
less
similar
couples
tended
to use
more
manipulation
tactics.
This
was
especially apparent
for use of the
silent
treatment,
but it
also appeared
with
charm, reason,
and
debasement.
Whether
the man or the
woman
was
judged
to
have
more
power
within
the
relationship
was
correlated
significantly
across
data
sources
by
using regression
tactics.
Specifically,
if
the
woman
had
more
power,
more tactics
of
regression
tended
to be
used than
if the man had
relatively
more power.
Across both
data
sources,
high
probabilities
of
judged
rela-
tionship
termination correlated
positively
with
high
use of
coer-
cion
and
debasement
manipulation
tactics.
In
sum.
tactics
of
manipulation
showed
coherent
links
with
independently
as-
sessed
characteristics
of the
couple
relationship.
Person-Environment
Links
The
last
set of
analyses centered around
links
between
the use
of
manipulation tactics
and the
tactics
to
which
one is
exposed.
Within
the
current
framework,
these
are
conceptualized
as
per-
son-environment
links.
Table
9
shows these
links
in
two
forms.
It
gives
both
the
correlations between
the
observer-reported
tac-
tics
and the
self-reported
environment
and the
correlations
be-
tween
the
self-reported tactics
and the
observer-reported
envi-
ronment.
Thus
both sets
of
correlations
involve
links
between
independent data sources. Interpreted here
are
those
analogous
correlations
significant
for
both analyses.
The first
striking
feature
of the
matrix
is
that
all
correlations
are
positive
in
sign,
which
suggests that elevated
use of any ma-
nipulation
tactic tends
to be
associated
with
being
the
recipient
of
higher frequencies
of
other tactics
of
influence.
Because
all
correlations
are
based
on
data
derived
from
separate
sources,
this
overall positive manifold must
be
interpreted
as a
substan-
tive
finding.
A
second
finding
concerns reciprocity. Does
the use of a
given
tactic tend
to be
associated
with
receiving
that tactic
in
return?
The
relevant correlations
are
shown
in the
matrix.
Of the
12
relevant
correlations,
all
were positive,
and 8
reached
statistical
significance.
For the
total manipulation
scores,
these
person-
environment correlations were
.33 (p <
.05)
for
the
self-re-

1226
BUSS,
GOMES,
HIGGINS,
AND
LAUTERBACH
-,
oo
O
r-
5i
m
C
- r
—
—
r-i
(N
r-i
r-l
c
—
(
I
1
n
«
a
C
r*-,
r-4
—'
r*'!
!••":
O
i~^l
—'
so
'I
O
l~~-
—
n
O
<--i
r-l
—'
<N
— O
f>
O
fi
— —
<
V)
— so
f*"-
ONU-,
m^cr-i—'
—
-«T
O
s
—
—
r>i
—
—
r-i
-~
O — —
r-i
—'
rsi
<
<NOr)r-i
—
mo
—
r^C
* * *
# #
—
—
—
m
— o —
oo
r-ino
Q-oo'-r^Q-ri
—
r-ioo
O
—
rNOOOOrN<5
ported
tactics
with
the
observer-reported
environment
and
.41
(p<
.01)
for the
observer-reported tactics
with
the
self-reported
environment. Thus there appears
to be
considerable
reciprocity
in
these person-environment
links.
Couple correspondence
seems particularly strong
for the
charm
and
coercion
tactics.
Several additional
findings in
Table
9 are
noteworthy.
Use of
the
silent
treatment
was
correlated
with
receiving
regression
and
debasement tactics. Those
who
used
regression
tactics
tended
to be
recipients
of
coercion
tactics.
Finally,
use of de-
basement tended
to be
associated
with
receiving
both
silent
treatment
and
coercion tactics.
In
sum,
these results suggest that
the
currently
assessed
tac-
tics
of
manipulation
are
linked
not
only
to
standard
personality
dimensions such
as
neuroticism,
ambitiousness,
and
quarrel-
someness,
but
they
also
show coherent
links
to the
interpersonal
environment
to
which
one is
exposed.
Discussion
We
have
identified
six
distinct
tactics
of
manipulation,
and
these
tactics
emerge
from
the
factor
analyses
of
four
instru-
ments.
The six
tactics,
ordered
by
their
frequency
of use
from
most
to
least
as
assessed
with
the
four
instruments,
are
reason,
charm, regression, coercion,
silent
treatment,
and
debasement.
Agreement between
self
and
partner
on
relative
performance
varied
across
the
tactics.
The
highest self-other agreement
oc-
curred
for the
coercion lactic,
perhaps
because
it is the
most
overt
and
open
to
observation.
The
debasement
tactic
showed
the
least
self-other
agreement, perhaps because
its use is
more
covert
and
less readily observable. Subjects
showed
moderately
high
consistency
in
using manipulation across
the
contexts
of
behavioral instigation
and
behavioral
termination.
This
indi-
vidual
difference
consistency ranged
from
a low of
.71
to a
high
of
.89.
Despite
strong consistency
in the
individual
difference
sense.
the
sample
as a
whole deployed
different
tactics
for
behavioral
instigation
and
termination.
Specifically,
the
charm tactic
was
used
more
for
instigating
the
behavior
of
others.
In
contrast,
the
coercion
and
silent treatment tactics were used more
frequently
for
terminating unwanted behavior
of
others.
Differential
de-
ployment
of
manipulation tactics depending
on the
goal
oc-
curred
with
similar magnitude across
self-
and
observer-re-
ported
data
sources
and is
therefore
a
robust
substantive
finding.
Manipulation
tactics showed correlations across data sources
with
standard
personality variables.
In
particular, those
high
on
EPQ
Neuroticism tended
to use
regression
and
silent
treatment
tactics more than
those
who
scored low. Strong
validity
was
pro-
vided
for the IAS
Calculating scale.
High
scorers tended
to use
all
manipulation tactics
relatively
frequently.
This
finding was
strongest
for
charm, silent treatment, reason,
and
debasement,
which
showed
significant
correlations
for
both data sources.
Those using
the
reason tactic
relatively
often
tended
to
score
high
on IAS
Ambitious.
In
contrast, those
who
used debase-
ment
relatively
frequently
tended
to
score
high
on IAS
L.a7.y.
Coercive
and
silent treatment
tactics
covaried
across data
sources with
IAS
Quarrelsome
(positive)
and
Agreeable
(nega-
tive).
These
results provide validational
information
for
several
IAS
scales
and
suggest that
one
relatively
neglected part
of
per-

TACTICS
OF
MANIPULATION
1227
Table
8
Correlations
Between
Relationship Variables
and
Manipulation Tactics (Instigation)
Well
matched
Tactic
Charm
Silent
treatment
Coercion
Reason
Regression
Debasement
Total
score
Self
data
-.11
-.25"
-.13
-.20*
-.27"
-.17
-.27"
Observer
data
-.33"*
-.18*
-.12
-.35*"
-.10
-.34***
-.39***
Similarity
Self
data
-.23*
-.23*
-.14
-.16
-.39***
-.18*
-.31"
Observer
data
-.30**
-.19*
-.07
-.18*
-.10
-.33"*
-.31"
Power
(she
or he)
Self
data
.00
-.07
-.02
.00
-.27"
.00
-.07
Observer
data
-.10
-.20
-.12
-.25*
-.26**
-.25*
-.27*
Probability
of
termination
Self
data
.07
.35"*
.25**
.14
.16
•>•?**
.26"
Observer
data
,7,
.16
.18*
.19
^2*
.36*
.33***
Note.
Relationship
variables
are
based
on
interviewer
judgments
composited
with
unit
weighting
across
the
male
and
female interviewers.
*;?<.05.
"ex.01.
*"p<.001.
sonality
consists
of the
tactics
used
to
shape
our
interpersonal
world.
In
addition
to
clear links
to
personality
variables,
manipula-
tion
tactics covaried
with
independently
assessed
characteristics
of
the
couple's relationship.
The
less similar
and
less
well
matched
the
couple,
the
more
frequently
they deployed
tactics
of
manipulation.
This
rinding
was
especially
robust
for the si-
lent treatment, debasement,
and
reason
tactics.
Couples
judged
by
interviewers
to be
relatively well matched
and
similar
tended
to use
these
manipulation tactics less
frequently.
Judgments
of the
relative power
balance
within
the
relation-
ship (man
vs.
woman) were
correlated
with
use of the
regression
tactic. When women were judged
by
interviewers
to
have
more
power,
regression tactics were displayed
more
frequently
in the
relationship.
Judged probability
of
relationship
termination
was
correlated positively with
the use of
both coercion
and de-
basement tactics.
The final set of
results center around links
to the
manipula-
tion
environment
to
which
one is
exposed.
Two
general
findings
emerged. First, elevated
use of any
manipulation tactic
was as-
sociated
with
elevated receipt
of
manipulative
acts
by the
part-
ner.
Second, couples showed correspondence with respect
to
their tactics.
These
results have implications
for
future research
con-
ducted
within
the
person-environment
correspondence
frame-
work
(Buss, 1984b, 1985a),
for the
taxonomic
task
that
faces
personality
psychology,
and for
links
between personality
and
other disciplines.
Implications
and
future
Research Directions
As
outlined
in the
introduction, there
are
three basic mecha-
nisms
in the
present interactionism
framework
by
which
links
between features
of
persons
and
features
of
their environments
are
produced.
One can
select environments
nonrandomly,
such
as
habitats, climates,
and
locations
in the
physical
realm
or
friends,
mates,
and
colleagues
in
the
social sphere.
One can
evoke
responses
unintentionally,
such
as
eliciting
landslides
through incautious motion
or
provoking
hostility
through
a
high activity
level.
One can
also manipulate inanimate
and
liv-
ing
objects.
This
study
offers
an
empirical
probe
into
the
tactics that peo-
ple
use to
manipulate
one
another,
the
personality characteris-
tics
of
those
who use
them,
and
links
between
use
ofthese
tactics
and
certain
features
of the
social environment.
Results
suggest
that personality
does
not
consist
simply
of
intrapsychic
struc-
tures
or
even adjustments
to the
environment,
which
implies
organismic
change
to a fixed
environment.
A
central part
of
personality
consists
of the
ways
in
which
we
shape
the
world
we
inhabit. Future studies could examine
manipulation
tactics
across
a
broader
array
of
social relationships,
including
those
used
with
friends,
parents, children, allies,
and
competitors.
Table
9
Person-Environment
Links
Tactic
1.
Charm
2.
Silent
treatment
3.
Coercion
4.
Reason
5.
Regression
6.
Debasement
1
.41*"
(.29*")
.13
.17
.01
22*
.16
2
.13
.27* (.11)
.16
.16
.25*
.21*
3
.16
.41*"
.48"*
(.35")
.20*
.42***
.34"
4
.14
.01
.05
.25"
(.07)
.25*
.18*
^
.14
.26
.29**
.03
.09
(.30")
.30**
6
.19
-n
.22*
.07
.16
.26*
(.
12)
Note.
Numbers
in
parentheses
and
below
the
diagonal
are
correlations
between
self-reported
tactics
and the
observer-reported
environment,
Those
in
and
above
the
diagonal
are
between
observer-reported
tactics
and the
self-reported
environment.
*p<
.05.**;><.01.***p<-001.

1228
BUSS, GOMES,
HIGGINS,
AND
LAUTERBACH
This
framework
calls
for an
expanded
view
of the
taxonomic
task
that
faces
personality
psychology.
The
taxonomic
task
ex-
pands
to
include
not
just
dispositions
identified
through
lexical
or
statistical
analyses
(Buss
&
Craik,
1985)
such
as
those
of
Nor-
man
(1963),
Goldberg
(1981),
Gough
(1968),
Cattell
(1946),
Gmlford
(1975),
Eysenck
(1947),
or
Costa
and
McCrae
(1980).
Personality
taxonomies
should
include
assessment
of the
rela-
tively
enduring
ways
in
which
people
select,
evoke,
and
manipu-
late
the
environments
they
inhabit,
as
well
as
the
projects
(Lit-
tle, 1983)
and
life
tasks
(Cantor
&
Kihlstrom,
1986)
toward
which
these
mechanisms
are
directed.
In
this
way,
personality
psychology
can
move
beyond
trait
identification
and
advance
toward
a
more
dynamic
interactionism.
A
closer
rapprochement
between
personality
and
social
psy-
chology
is
also
suggested
by
this
framework.
Influence
has
been
a
central
topic
in
social
psychology
(Cialdini,
1985),
but
studies
have
omitted
the
role
of
consistent
individual
differences
in the
tactics
by
which
influence
occurs.
Attention
to the
manner
in
which
dispositions
are
played
out in the
social
sphere
benefits
both
disciplines.
Toward
this
end,
future
work
could
assess
the
effectiveness
of
each
tactic
of
manipulation
as
well
as
which
people
display
them
potently.
A
final
implication
centers
around
the
integration
of
evolu-
tionary
biology
with
personality
psychology
(Buss,
1984a).
Re-
cent
work
in
evolutionary
theory
(e.g.,
Dawkins
&
Krebs,
1978;
Krebs
&
Dawkins,
1984)
suggests
that
the
manipulation
of
competitors,
allies,
parents,
offspring,
friends,
and
lovers
is
cen-
tral
to
reproductive
success.
Existing
humans
have
ancestors
who
were
especially
adept
at
influencing
others.
This
implies
that
there
are
evolutionarily
relevant
proximate
goals
toward
which
tactics
of
manipulation
are
directed,
such
as
resource
acquisition,
alliance
formation,
intrasexual
competition,
mate
selection,
and
nepotistic
investment
(Buss,
1986).
Future
re-
search
could
profitably
examine
the
proximate
goals
toward
which
human
manipulative
tactics
are
directed.
References
Buss,
D. M.
(1981).
Predicting
parent-child
interactions from chil-
dren's
activity
level.
Developmental
Psychology,
17,
59-65.
Buss,
D. M.
(1984a). Evolutionary biology
and
personality
psychology:
Toward
a
conception
of
human nature
and
individual
differences.
American
Psychologist,
39,
1135-1147.
Buss,
D. M.
(1984b).
Toward
a
psychology
of
person-environment
(PE)
correspondence:
The
role
of
spouse
selection. Journal
of
Personality
and
Social
Psychology,
47,
361-377.
Buss.
D. M.
(19ioa).
Human mate selection.
American
Scientist,
73,
47-51.
Buss,
D. M.
(I985b).
The
temporal
stability
of
acts,
trends,
and
pat-
terns.
InC.
D.
Spielberger&J.
N.
Butcher
(Eds.
\
Advances
in
person-
ality
assessment
(Vol.
5. pp.
165-196).
Hillsdale.
NJ:
Erlbaum.
Buss,
D. M.
(1986).
Can
social science
be
anchored
in
evolutionary
biology?
Four problems
and a
strategic solution.
Revue
Europeenne
des
Sciences
Sociales,
24.
41-50.
Buss,
D.
M.,
&
Craik,
K.
H.
(1983).
The act
frequency
approach
to
personality. Psychological
Review,
90,
105-126.
Buss.
D.
M..
&
Craik,
K. H.
(1984). Acts,
dispositions,
and
personality.
In
B. A.
Maher
& W. B.
Maher
(Eds.),
Advances
in
experimental
per-
sonality
research
(Vol.
13,
pp.
241-301).
New
York:
Academic
Press.
Buss.
D.
M.,
&
Craik,
K. H.
(1985).
Why not
measure that trait? Alter-
native criteria
for
identifying
important
disposiiions.
Journal
0!
Per-
sonality
and
Social
Psychology,
48,
934-946.
Cantor,
N.,
&
Kihlstrom,
J. F.
(1986).
Social
intelligence:
The
cognitive
basis
of
personality.
In P.
Shaver
(Ed.).
Review
ol
personality
ami
so-
cial psychology. Beverly Hills,
CA:
Sage.
Cattell,
R. B.
(1946).
Description
and
measurement
of
personality
New
York:
World Book.
Christie,
R.,
&
Geis,
F. L.
(1970).
Studies
in
Machiavellianism
New
York:
Academic
Press.
Cialdini,
R. B.
(1985).
Influence:
Science
and
practice.
Glenview.
1L:
Scott,
Foresman.
Costa,
P.
T,
Jr.,
&
McCrae,
R. R.
(1980).
Still stable
after
all
these
years:
Personality
as the key to
some
issues
in
aging.
In
P. B.
Baltes
& O.
G.
Brim
(Eds.),
Life
span
development
and
behavior
(Vol.
3).
New
York:
Academic
Press.
Dawkins,
A..
&
Krebs,
J. R.
(1978).
Animal signals:
Information
or
manipulation?
In J. R.
Krebs
& N. B.
Davies
(F.ds.),
Behavioral ecol-
ogy:
An
evolutionary
approach
Oxford.
England:
Blackwell
Scien-
tific.
Ekehammar,
B.
(1974).
Interaclionism
in
personality
from
a
historical
perspective.
Psychological
Bulletin,
81.
1026-1048.
Eysenck,
H. J.
(1947).
Dimensions
of
personality. London: Routledge
&
Kegan Paul.
Eysenck,
H.
J-,
&
Eysenck,
S. B.
(1975).
Eysenck
Personality
Question-
naire
manual.
San
Diego,
CA:
Educational Testing Service.
Goldberg,
L. R.
(1981). Language
and
individual
differences:
The
search
for
universals
in
personality lexicons.
In L.
Wheeler
(Ed.),
Re-
view
of
personality
and
social psychology (Vol.
1. pp.
141-165).
Bev-
erly
Hills,
CA:
Sage.
Golding,
S. L.
(1975).
Flies
in the
ointment:
Methodological
problems
in
the
analysis
of the
percentage
of
variance
due to
persons
and
situa-
tions.
Psychological
Bulletin.
82,
278-288.
Gough,
H. G.
(1968).
An
interpreter's syllabus
for the
California
Psy-
chological Inventory.
In P.
McReynolds
(Ed.),
Advances
in
psycholog-
ical
assessment
(Vol.
I,
pp.
55-79).
Palo
Alto,
CA:
Science
&
Behav-
ior
Books.
Guilford,
J. P.
(1975).
Factors
and
factors
of
personality.
Psychological
Bulletin,
82,
802-814.
Krebs,
J.
R.,
&
Dawkins,
R.
(1984).
Animal
signals: Mind-reading
and
manipulation.
In J. R.
Krebs
& N. B.
Davies
(Eds.),
Behavioral
ecol-
ogy:
An
evolutionary approach (2nd ed.). Sunderland,
MA:
Smauer.
Little,
B. R.
(1983).
Personal
projects:
A
rationale
and
method
for
inves-
tigation.
Environment
and
Behavior,
15,
273-309.
Magnusson,
D.,
&
Endlcr,
N. S.
(1977).
Personality
at
the
crossroads:
Current
issues
in
interactional
psychology
Hillsdale,
NJ:
Erlbaum.
Norman,
W. T.
(1963).
Toward
an
adequate taxonomy
of
personality
attributes:
Replicated factor structure
in
peer
nomination personality
ratings. Journal
of
Abnormal
and
Social
Psychology
66.
574-583.
Patterson,
G.
R.,
&
Bechtel,
G. G.
(1977).
Formulating
the
situational
environment
in
relation
to
states
and
traits.
In R. B.
Cattell
& R. M.
Dreger
(Eds.).
Handbook
of
modern
personality
theorv
(pp. 254-
286).
New
York: Halstead
Press.
Plomin,
R.,
DeFries,
J.
C,
&
Loehlin.
J. C.
(1977).
Genotype-environ-
ment interaction
and
correlation
in
the
analysis
of
human
behavior
Psychological
Bulletin.
88,
245-258.
Scarr,
S..
&
McCartney,
K.
(1983).
How
people
make their
own
environ-
ments:
A
theory
of
genotype
-^
environment
effects.
Child Develop-
ment,
54,
424-4K.
Snyder,
M.
(1981).
On the
influence
of
individuals
on
situations.
In N.
Cantor
& J. F.
Kilstom
(Eds.),
Personality,
cognition,
and
social
inter-
action.
Hillsdale,
NJ:
Erlbaum.
Snyder,
M.,
&
Gangestad.
S.
(1982).
Choosing
social situations:
Two
investigations
of
self-monitoring
processes.
Journal
of
Personality
and
Social
Psychology.
43,
123-135.
Wiggins,
J. S.
(1979).
A
psychological taxonomy
of
interpersonal
behav-
ior.
Journal
of
Personality
and
Social Psychology,
37
395-412.

TACTICS
OF
MANIPULATION1229
Appendix
Tactics
of
Manipulation
Instructions: When
you
want your partner
to do
something
for
you.
era!,
to do
what
is
described.
If you are
extremely
likely
10
do
il,
circle
what
are you
likely
to do?
Look
at
each
of the
items listed below
and
rate
how
likely
you
are to do
each when
you are
trying
to get
your
partner
to do
something. None
of
them
will
apply
to all
situations
in
which
you
want
your partner
to do
something,
so
rate
how
likely
you
are,
in
gen-
1. If you are not at all
likely
to do il,
circle
1.
If
you are
somewhat
likely
to do it,
circle
4.
Give intermediate ratings
for
intermediate
likelihoods
of
performing
the
behaviors.
1.
~)
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
I
ask him to do it.
I
pout
until
he
docs
it.
I
demand that
he do it.
1
compliment
him so
he'll
do it.
1
beg him to do it.
I
explain
why I
want
him
to do it.
I
sulk until
he
does
it.
I
embarrass
him
into
doing
it.
I
act
charming
so
he'll
do it.
I
whine
until
he
does
it.
I
give
him
reasons
for
doing
it.
I
ignore
him
until
he
agrees
to do it.
I
criticize
him for not
doing
it.
I
tell
him how
happv
I'll
be
if
he
does
it. '
I
act
humble
so
he'll
doit.
I
show
him
that
I
would
be
willing
to do
it
for
him.
I
am
silent
until
he
agrees
to do it.
I
compare
him to
someone
who
would
doit.
I
give
him a
small
gift
or
card before
I ask
him
to do it.
I
give
him
something
so
he'll
do it.
Not
at
all
likely
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Somewhat
likely
3
4
3
4
3
4
3
4
3
4
3
4
3 4
3 4
3 4
3 4
3 4
3
4
3
4
3
4
3
4
3
4
3
4
3 4
3 4
3
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
Ex-
tremely
likely
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Not
at
all
likely
I
repeat
the
request
from
different
angles.
1 2
I
don't
respond
to him
until
he
does
it. 1 2
I
yell
at him so
he'll
do
it. 1 2
I
try to be
loving/
romantic
when
I ask
him to do it. 1 2
I
let him do
something
that
I
don't
approve
of
so
that
he'll
do it. 1 2
1
draw analogies
between
what
I
want
done
and
cases
in our
past.
1 2
I
tell
him
I'll leave
if he
doesn't
do il. 1 2
I
threaten
him
with
something
if he
doesn't
do
it.
1 2
I
tell
him
I'll
do him a
favor
if
he'll
do
it.
1 2
I
lower myself
so
he'll
do it. 1 2
I
ask why he
doesn't
do
it. 1 2
I
refuse
to do
something
that
he
likes
until
he
does
it. 1 2
I
curse
at him
until
he
does
it. 1 2
I
point
out all of the
good
things that
will
come
from
doing
it. 1 2
I
allow myself
to be
debased
so
he'll
do it.
1
2
Somewhat
likel\
3
4 5
3 4 5
3 4 5
3 4 5
3 4 5
3 4 5
3
4 ?
345
3
4 5
345
345
345
345
3 4 5
3 4 5
Ex-
tremely
likely
6 7
*
7
6 7
6 7
6 7
6 7
6 7
6 7
6
1
6 7
6
7
6 7
fi
7
6 7
6 7
Received
February
7.
1986
Revision
received
August
8.
1986
Accepted
December
12,
1986