Personality
and
Social
Psychology
Review
2001,
Vol.
5,
No.
3,
184-200
Copyright
©
2001
by
Lawrence
Erlbaum
Associates,
Inc.
A
Social
Identity
Theory
of
Leadership
Michael
A.
Hogg
School
of
Psychology
University
of
Queensland
A
social
identity
theory
of
leadership
is
described
that
views
leadership
as
a
group
process
generated
by
social
categorization
and
prototype-based
depersonalization
processes
associated
with
social
identity.
Group
identification,
as
self-categorization,
constructs
an
intragroup
prototypicality
gradient
that
invests
the
most
prototypical
member
with
the
appearance
of
having
influence;
the
appearance
arises
because
members
cognitively
and
behaviorally
conform
to
the
prototype.
The
appearance
of
influence
becomes
a
reality
through
depersonalized
social
attraction
processes
that
makefollowers
agree
and
comply
with
the
leader's
ideas
and
suggestions.
Consensual
social
attraction
also
imbues
the
leader with
apparent
status
and
creates
a
sta-
tus-based
structural
differentiation
within
the
group
into
leader(s)
and
followers,
which
has
characteristics
of
unequal
status
intergroup
relations.
In
addition,
afunda-
mental
attribution
process
constructs
a
charismatic
leadership
personality
for
the
leader,
which
further
empowers
the
leader
and
sharpens
the
leader-follower
status
differential.
Empirical
support
for
the
theory
is
reviewed
and
a
range
of
implications
discussed,
including
intergroup
dimensions,
uncertainty
reduction
and
extremism,
power,
and
pitfalls
of
prototype-based
leadership.
Over
the
past
25
years
social
psychology
has
placed
relatively
little
emphasis
on
the
study
of
leadership.
This
is
probably
associated
with
the
well-documented
decline
during
the
1960s
and
1970s
of
interest
in
small
group
research,
the
associated
ascendency
of
social
cognition,
the
European
emphasis
on
large
scale
inter-
group
relations,
and
the
"outsourcing"
of
small
group
and
leadership
research
to
organizational
and
manage-
ment
departments
(for
historical
overviews,
see
Abrams
&
Hogg,
1998;
Hogg
&
Abrams,
1999;
Jones,
1998;
McGrath,
1997;
Moreland,
Hogg,
&
Hains,
1994;
Sanna
&
Parks,
1997).
In
recent
years,
however,
the
social
identity
perspec-
tive
has
begun
to
provide
an
integrative
conceptual
fo-
cus
for
a
revival
of
interest
among
social
psychologists
in
group
phenomena
and
for
a
dynamic
linkage
of
social
cognitive
and
intergroup
processes
(Hogg
&
Abrams,
1999;
Moreland
et
al.,
1994).
This
has
generated
social
identity
analyses
of
a
diverse
range
of
phenomena
(e.g.,
Abrams
&
Hogg,
1990,1999;
Capozza
&
Brown,
2000;
Ellemers,
Spears,
&
Doosje,
1999;
Hogg
&
Abrams,
This
article
was
made
possible
by
a
research
grant
from
the
Aus-
tralian
Research
Council.
I
thank
Kelly
Fielding,
Margaret
Foddy,
Sarah
Hains,
Leigh
Mor-
ris,
and
Sherry
Schneider
for
their
various
intellectual
contributions
to
the
development
of
some
of
the
ideas
presented
in
this
article.
Requests
for
reprints
should
be
sent
to
Michael
A.
Hogg,
School
of
Psychology,
University
of
Queensland,
Brisbane,
QLD
4072,
Austra-
lia.
E-mail:
m.hogg@psy.uq.edu.au
1988;
Robinson,
1996;
J.
C.
Turner,
Hogg,
Oakes,
Reicher,
&
Wetherell,
1987;
Worchel,
Morales,
Paez,
&
Deschamps,
1998)
and
has
provided
a
social
cognitive
framework
for
social
psychology
to
reexamine
leader-
ship
as
a
group
process.
In
this
article
I
describe
a
social
identity
theory
of
group
leadership.
I
develop
the
theory
with
an
introduction
to
relevant
aspects
of
social
identity
and
self-categorization
the-
ory,
and
then
I
contextualize
by
a
short
dissertation
on
leadership
research.
I
describe
in
some
detail
the
em-
pirical
support
for
core
components
of
the
leadership
theory
with
the
emphasis
on
direct
tests.
In
the
latter
part
of
the
article,
I
draw
out
and
describe
some
direct
implications
and
extensions
of
the
theory
for
a
range
of
aspects
of
leadership.
Brief
Overview
of
Leadership
Research
Leadership
has
long
been
a
focus
of
research
for
so-
cial
psychology
and
the
wider
social
sciences
(e.g.,
Bass,
1990a;
Chemers,
1987,
2001;
Eagly,
Karau,
&
Makhijani,
1995;
Fiedler
&
House,
1994;
Graumann
&
Moscovici,
1986;
Hollander,
1985;
Stogdill,
1974;
Yukl,
1981).
In
recent
years,
however,
it
has
waned
in
popularity
for
mainstream
social
psychology
and
is
in-
stead
more
commonly
researched
in
other
disciplines,
184
at Bobst Library, New York University on March 23, 2015psr.sagepub.comDownloaded from
SOCIAL
IDENTITY
THEORY
OF
LEADERSHIP
particularly
organizational
psychology
(e.g.,
Wilpert,
1995;
Yukl
&
Van
Fleet,
1992).
Although
personality
perspectives
identify
some
per-
sonality
correlates
of
leadership
(e.g.,
talkativeness;
Mullen,
Salas,
&
Driskell,
1989),
and
personality
ex-
plains
some
variance
in
the
emergence
of
leaders
in
ini-
tially
leaderless
groups
(Kenny
&
Zaccaro,
1983),
scholars
agree
that
personality
alone
is
a
relatively
poor
correlate
of
leadership
(Stogdill,
1974;
Yukl,
1981;
but
see
Albright
&
Forziati,
1995).
An
alternative,
situational
perspective
is
that
almost
anyone
can
be
an
effective
leader
if
the
circumstances
are
right
(e.g.,
Bales,
1950;
Sherif,
1966).
In
the
1960s,
drawing
on
a
third
strand
of
research
that
described
the
actual
behavior
of
leaders
(e.g.,
Bales,
1950;
Lippitt
&
White,
1943;
Stogdill,
1974),
Fiedler
(1965,
1971)
introduced
an
interactionist
model;
the
leadership
effectiveness
of
a
particular
behav-
ioral
style
is
contingent
on
the
favorability
of
the
situation
to
that
behavioral
style.
Fiedler's
contingency
theory
is
generally
reasonably
well
supported
(e.g.,
Strube
&
Gar-
cia,
1981)
despite
some
continuing
controversy
(e.g.,
Pe-
ters,
Hartke,
&
Pohlmann,
1985).
Another
perspective
focuses
on
leadership
as
a
dy-
namic
product
of
transactions
between
leaders
and
follow-
ers
(Bass,
1990b;
Hollander,
1985;
Lord
&
Maher,
1991;
Nye
&
Simonetta,
1996).
For
example,
because
leaders
play
a
significant
role
in
helping
followers
achieve
their
goals,
followers
bestow
power
and
status
on
leaders
to
re-
store
equity.
Relatedly,
followers
may
try
to
redress
the
power
imbalance
in
groups
by
gaining
personal
informa-
tion
about
the
leader.
This
is
an
attributional
process
that
imbues
the
leader
with
charisma
and
thus
additional
power
(Fiske,
1993;
Fiske
&
Depret,
1996).
Leaders
may
also
ac-
cumulate
"idiosyncrasy
credit"
with
the
group
by
con-
forming
to
group
norms.
This
subsequently
allows
them
to
be
innovative
and
effective
leaders
(Hollander,
1958;
Hol-
lander
&
Julian,
1970).
Recent
transactional
leadership
perspectives,
mainly
in
organizational
psychology
(Wilpert,
1995),
focus
on
transformational
leadership.
Charismatic
leaders
are
able
to
motivate
followers
to
work
for
col-
lective
goals
that
transcend
self-interest
and
transform
organizations
(Bass,
1990b;
Bass
&
Avolio, 1993;
see
Mowday
&
Sutton,
1993,
for
critical
comment).
This
focus
on
charisma
is
particularly
evident
in
"new
lead-
ership"
research
(e.g.,
Bass,
1985,
1990b,
1998;
Bryman,
1992;
Burns,
1978;
Conger
&
Kanungo,
1987,
1988),
which
proposes
that
effective
leaders
should
be
proactive,
change
oriented,
innovative,
mo-
tivating
and
inspiring,
and
have
a
vision
or
mission
with
which
they
infuse
the
group.
They
should
also
be
interested
in
others
and
be
able
to
create
commitment
to
the
group
and
extract
extra
effort
from,
and
gener-
ally
empower,
members
of
the
group.
Social
psychology's
emphasis
on
social
cognition
has
produced
an
extension
of
implicit
leadership
the-
ory
(Hollander
&
Julian,
1969)
called
leader
categori-
zation
theory
(e.g.,
Lord,
Foti,
&
DeVader,
1984;
Nye
&
Forsyth,
1991;
Palich
&
Hom,
1992;
Rush
&
Rus-
sell,
1988;
also
see
Nye
&
Simonetta,
1996).
People
have
preconceptions
about
how
leaders
should
behave
in
general
and
in
specific
leadership
situations.
These
preconceptions
are
cognitive
schemas
of
types
of
leader
(i.e.,
categories
of
leader
that
are
represented
as
person
schemas)
that
operate
in
the
same
way
as
other
schemas
(see
Fiske
&
Taylor,
1991).
When
someone
is
categorized
on
the
basis
of
their
behavior
as
a
leader,
the
relevant
leadership
schema
comes
into
play
to
gen-
erate
further
assumptions
about
behavior.
Leadership
schemas
vary
in
situational
inclusiveness.
Subordinate
schemas
apply
only
to
specific
situations
(they
may
re-
late
to
specific
status
characteristics,
as
described
by
expectation
states
theory;
e.g.,
Berger,
Fisek,
Norman,
&
Zelditch,
1977;
Ridgeway,
2001),
whereas
superordinate
schemas
apply
to
a
wide
range
of
situa-
tions
and
embody
very
general
leadership
characteris-
tics.
Good
leaders
are
people
who
have
the
attributes
of
the
category
of
leader
that
fits
situational
requirements.
This
perspective
treats
leader
categories
as
nominal
categories-that
is,
cognitive
groupings
of
instances
that
share
attributes
but
do
not
have
any
psychological
existence
as
a
real
human
group.
Indeed,
the
notion
of
a
social
group
only
of
leaders
makes
little
sense;
who
would
lead
and
who
would
follow?
Leadership
is
viewed
as
a
product
of
individual
information
process-
ing,
not
as
a
structural
property
of
real
groups
nor
as
an
intrinsic
or
emergent
property
of
psychological
ingroup
membership
(see
Hogg,
1996a).
Commentary
on
Leadership
Research
and
a
New
Direction
With
only
a
few
notable
exceptions,
then,
the
recent
study
of
leadership
has
been
conducted
outside
of
con-
temporary
mainstream
social
psychology
and
so
has
not
benefitted
from
some
of
the
recent
conceptual
ad-
vances
made
within
social
psychology.
Although
most
perspectives
now
recognize
that
leadership
is
a
rela-
tional
property
within
groups
(i.e.,
leaders
exist
be-
cause
of
followers
and
followers
exist
because
of
leaders),
there
is
no
analysis
of
leadership
that
de-
scribes
how
leadership
may
emerge
through
the
opera-
tion
of
ordinary
social
cognitive
processes
associated
with
psychologically
belonging
to
a
group.
In
contrast,
the
most
recent
analytic
emphasis
is
mainly
on
(a)
individual
cognitive
processes
that
cat-
egorize
individuals
as
leaders-the
social
orientation
between
individuals
is
not
considered,
and
thus
group
processes
are
not
incorporated;
or
(b)
whether
indi-
viduals
have
the
charismatic
properties
necessary
to
meet
the
alleged
transformational
objectives
of
lead-
185
at Bobst Library, New York University on March 23, 2015psr.sagepub.comDownloaded from
HOGG
ership-leadership
is
a
matter
of
situationally
attrac-
tive
individual
characteristics
rather
than
group
processes.
Both
these
perspectives
have
recently
in-
vited
some
criticism
for
neglecting
the
effects
of
larger
social
systems
within
which
the
individual
is
embedded
(e.g.,
Hall
&
Lord,
1995;
Haslam
&
Platow,
2001;
Lord,
Brown,
&
Harvey,
2001;
Pawar
&
Eastman,
1997).
Lord
et
al.
(2001)
explained
that
leadership
cannot
be
properly
understood
in
terms
of
a
leader's
actions
or
in
terms
of
abstract
perceptual
categories
of
types
of
leader,
and
that
a
paradigm
shift
in
how
we
understand
leadership
is
called
for.
Their
solution
is
to
explore
a
connectionist,
or
parallel
con-
straint
satisfaction,
level
model.
However,
if
leadership
is
indeed
a
structural
feature
of
ingroups,
then
leaders
and
followers
are
interdepen-
dent
roles
embedded
within
a
social
system
bounded
by
common
group
or
category
membership.
Thus,
leadership
dynamics
may
be
significantly
affected
by
the
social
cognitive
processes
associated
with
group
membership
(and
group
behaviors),
specifically
the
processes
of
self-categorization
and
depersonalization
now
believed
to
be
responsible
for
social
identity
pro-
cesses,
group
behavior,
and
intergroup
relations.
Leaders
may
emerge,
maintain
their
position,
be
effec-
tive,
and
so
forth,
as
a
result
of
basic
social
cognitive
processes
that
cause
people
1.
To
conceive
of
themselves
in
terms
of
the
defin-
ing
features
of
a
common
and
distinctive
ingroup
(i.e.,
self-categorization,
or
identification,
in
terms
of
the
ingroup
prototype).
2.
To
cognitively
and
behaviorally
assimilate
themselves
to
these
features
(i.e.,
cognitive
and
behav-
ioral
depersonalization
in
terms
of
the
ingroup
proto-
type
producing
ingroup
stereotypic
or
normative
perceptions,
attitudes,
feelings,
and
behaviors).
3.
To
perceive
others
not
as
unique
individuals
but
through
the
lens
of
features
that
define
relevant
ingroup
or
outgroup
membership
(i.e.,
perceptual
depersonal-
ization
of
others
in
terms
of
the
ingroup
or
outgroup
prototype,
producing
stereotypical
homogenization).
If
leadership
is
produced
by
these
group
processes
contingent
on
psychologically
belonging
to
the
group,
then
having
the
prototypical
or
normative
characteris-
tics
of
a
psychologically
salient
ingroup
(i.e.,
being
a
prototypical
ingroup
member)
may
be
at
least
as
im-
portant
for
leadership
as
being
charismatic
or
having
schema-consistent
characteristics
of
a
particular
type
or
category
of
leader
(i.e.,
being
schematic
of
a
nomi-
nal
leader
category).
My
aim
in
this
article
is
to
show
that
a
social
iden-
tity
analysis
can
provide
this
sort
of
new,
group
mem-
bership
oriented
analysis
of
leadership
processes.
First,
I
briefly
review
social
identity
and
self-categori-
zation
theory,
then
describe
the
proposed
leadership
theory
followed
by
empirical
support
for
its
core
ten-
ets,
and
finally
I
explore
a
range
of
specific
implica-
tions,
extensions,
and
future
directions.
Social
Identity
and
Self-Categorization
The
social
identity
perspective
contains
a
number
of
compatible
and
interrelated
components
and
empha-
ses,
in
particular
an
original
emphasis
by
Tajfel
and
Turner
and
their
associates
on
social
identity,
social
comparison,
intergroup
relations,
and
self-enhance-
ment
motivation
(often
simply
called
social
identity
theory;
e.g.,
Tajfel
&
Turner,
1979)
and
a
later
cogni-
tive
emphasis
by
J.
C.
Turner
and
his
associates
on
the
categorization
process
(called
self-categorization
the-
ory;
e.g.,
J.
C.
Turner
et
al.,
1987).
Social
identity
the-
ory
and
self-categorization
theory
have
been
extensively
overviewed
as
an
integrated
whole
else-
where
(e.g.,
Hogg,
1996a,
2000a,
2001;
Hogg
&
Abrams,
1988,
1999;
Hogg,
Terry,
&
White,
1995;
J.
C.
Turner,
1999).
Social
Identity
and
Intergroup
Relations
Tajfel
(1972) introduced
the
idea
of
social
identity
to
theorize
how
people
conceptualize
themselves
in
in-
tergroup
contexts,
how
a
system
of
social
categoriza-
tions
"creates
and
defines
an
individual'
s
own
place
in
society"
(p.
293).
He
defined
social
identity
as
"the
in-
dividual'
s
knowledge
that
he
belongs
to
certain
social
groups
together
with
some
emotional
and
value
signif-
icance
to
him
of
this
group
membership"
(Tajfel,
1972,
p.
292).
Because
groups
only
exist
in
relation
to
other
groups,
they
derive
their
descriptive
and
evaluative
properties,
and
thus
their
social
meaning,
in
relation
to
these
other
groups.
Furthermore,
because
social
identity
is
self-evaluative
and
derives
its
value
from
the
evaluative
properties
of
the
ingroup,
social
compari-
sons
between
groups
are
focused
on
establishing
evaluatively
positive
distinctiveness
for
one's
own
group.
Intergroup
relations
involve
a
process
of
com-
petition
for
positive
identity
(J.
C.
Turner,
1975)
in
which
groups
and
their
members
strive
to
protect
or
en-
hance
positive
distinctiveness
and
positive
social
iden-
tity.
The
specific
way
this
occurs
is
governed
by
people's
subjective
understanding
of
the
psychological
permeability
of
group
boundaries
and
the
stability
and
legitimacy
of
status
relations
between
groups
(Tajfel
&
Turner,
1979).
This
aspect
of
social
identity
theory
has
had
a
significant
impact
on
social
psychology
(see
Ellemers,
1993;
Hogg
&
Abrams,
1988).
186
at Bobst Library, New York University on March 23, 2015psr.sagepub.comDownloaded from
SOCIAL
IDENTITY
THEORY
OF
LEADERSHIP
Self-Esteem
Hypothesis
The
drive
for
evaluatively
positive
social
identity
through
positive
distinctiveness
is
underpinned
by
a
ba-
sic
human
need
for
positive
self-esteem
(e.g.,
J.
C.
Turner,
1982;
J.
C.
Turner,
Brown,
&
Tajfel,
1979),
a
self-enhancement
motive.
The
implication
is
that
self-esteem
motivates
social
identification
and
group
behavior,
and
social
identification
satisfies
the
need
for
self-esteem
(Abrams
&
Hogg,
1988).
Reviews
of
re-
search
on
this
self-esteem
hypothesis
reveal
inconsistent
findings
that
suggest
a
distinction
between
individual
and
group
membership
based
self-esteem,
and
that
the
relationship
between
self-esteem
and
group
behavior
may
be
affected
by
other
variables
such
as
self-esteem
extremity,
identity
strength,
and
group
threat
(see
Abrams
&
Hogg,
1988;
Hogg
&
Abrams,
1990;
Long
&
Spears,
1997;
Rubin
&
Hewstone,
1998).
Crocker
and
her
colleagues
(e.g.,
Crocker,
Blaine,
&
Luhtanen,
1993;
Crocker
&
Luhtanen,
1990;
Luhtanen
&
Crocker,
1992)
have
explored
self-esteem
processes
in
inter-
group
behavior
extensively
and
have
developed
a
col-
lective
self-esteem
scale
that
some
researchers
employ
as
a
measure
of
social
identity.
Self-Categorization,
Prototypicality,
and
Depersonalization
The
cognitive
dimension
of
social
identity
is
speci-
fied
by
self-categorization
theory
(J.
C.
Turner,
1985;
J.
C.
Turner
et
al.,
1987)
in
terms
of
the
causes
and
con-
sequences
of
social
categorization
of
self
and
others
(Hogg,
2001).
The
process
of
social
categorization
perceptually
segments
the
social
world
into
ingroups
and
outgroups
that
are
cognitively
represented
as
pro-
totypes.
These
prototypes
are
context
specific,
multidi-
mensional
fuzzy
sets
of
attributes
that
define
and
prescribe
attitudes,
feelings,
and
behaviors
that
charac-
terize
one
group
and
distinguish
it
from
other
groups.
Social
categorization
of
other
people
perceptually
as-
similates
them
to
the
relevant
ingroup
or
outgroup
pro-
totype
and
thus
perceptually
accentuates
prototypical
similarities
among
people
in
the
same
group
and
prototypical
differences
between
people
from
different
groups;
there
is
an
accentuation
effect
(e.g.,
Tajfel,
1959,
1969)
that
underpins
stereotyping.
This
overall
process
is
called
depersonalization
because
people
are
not
viewed
as
unique
and
multifaceted
individuals
but
as
matches
to
the
relevant
ingroup
or
outgroup
proto-
type;
prototypicality,
not
individuality,
is
the
focus
of
attention.
Depersonalization
refers
to
change
in
the
ba-
sis
of
perception;
it
does
not
have
the
negative
conno-
tations
of
deindividuation
or
dehumanization.
Social
categorization
of
self,
self-categorization,
has
the
same
effect
but
more
so.
It
not
only
depersonal-
izes
self-perception
but
goes
further
in
actually
trans-
forming
self-conception
and
assimilating
all
aspects
of
ones
attitudes,
feelings,
and
behaviors
to
the
ingroup
prototype;
it
changes
what
people
think,
feel,
and
do.
Depersonalization
is
the
basic
process
underlying
group
phenomena;
it
perceptually
differentiates
groups
and
renders
perceptions,
attitudes,
feelings,
and
behav-
iors
stereotypical
and
group
normative.
Prototypes
are
often
stored
in
memory
to
be
"called
forth"
by
social
categorization
in
a
particular
context
to
guide
perception,
self-conception,
and
action.
How-
ever,
they
are
almost
always
modified
to
varying
de-
grees,
and
they
can
be
entirely
constructed
by
specifics
of
a
particular
social
context.
Prototypes
are
contextu-
ally
responsive,
and
the
principle
governing
this
con-
textual
sensitivity
is
metacontrast.
New
prototypes
form,
or
existing
ones
are
modified,
in
such
a
way
as
to
maximize
the
ratio
of
perceived
intergroup
differences
to
intragroup
similarities;
prototypes
form
to
accentu-
ate
similarities
within
a
category
and
differences
be-
tween
categories.
Social
Attraction
Hypothesis
Depersonalization
affects
people's
feelings
about
one
another.
They
become
based
on
perceived
prototypicality
(called
social
attraction)
rather
than
idiosyncratic
prefer-
ences
or
personal
relationships
(called
personal
attrac-
tion),
which
is
the
social
attraction
hypothesis
(Hogg,
1992,
1993;
for
direct
empirical
tests,
see
Hogg,
Coo-
per-Shaw,
&
Holzworth,
1993;
Hogg
&
Hains,
1996,
1998;
Hogg
&
Hardie,
1991;
Hogg,
Hardie,
&
Reynolds,
1995).
Ingroup
members
are
liked
more
than
outgroup
members
because
the
former
are
perceptually
assimilated
to
a
relatively
positive
ingroup
prototype,
or
because
their
prototypical
similarity
to
self
is
perceptually
accentuated,
or
because
self-liking
(self-esteem)
is
extended
to
em-
brace
people
who
are
to
some
extent
now
viewed
as
prototypical
extensions
of
self.
Within
the
ingroup
there
is
consensual
liking
that
is
stronger
for
more
prototypical
than
for
less
prototypical
members;
there
is
a
proto-
type-based
social
attraction
gradient.
Uncertainty
Reduction
Hypothesis
It
has
recently
been
suggested
that
social
identity
processes
are
not
only
motivated
by
self-enhancement
(the
self-esteem
hypothesis)
but
also
by
an
epistemic
or
self-evaluative
motive
to
reduce
subjective
uncer-
tainty
(the
uncertainty
reduction
hypothesis;
Hogg,
2000b;
Hogg
&
Abrams,
1993;
Hogg
&
Mullin,
1999;
for
direct
empirical
evidence,
see
Grieve
&
Hogg,
1999;
Hogg
&
Grieve,
1999;
Jetten,
Hogg,
&
Mullin,
187
at Bobst Library, New York University on March 23, 2015psr.sagepub.comDownloaded from
HOGG
2000;
Mullin
&
Hogg,
1998,
1999).
This
motive
is
closely
tied
to
social
categorization.
Subjective
uncertainty
about
important,
usually
self-conceptually
relevant
matters
is
aversive;
thus,
subjec-
tive
uncertainty
reduction
is
a
powerful
human
motive.
The
processes
of
self-categorization
and
prototype-based
depersonalization
reduce
uncertainty
because
perceptions,
attitudes,
feelings,
and
behavior
are
now
prescribed
by
an
ingroup
prototype
that
usually
has
consensual
validation
from
other
group
members.
Groups
with
high
entitativity
(e.g.,
Campbell,
1958;
see
Brewer
&
Harasty,
1996;
Ham-
ilton
&
Sherman,
1996),
with
clear,
unambiguous,
and
consensual
prototypes,
and
which
are
located
in
a
clearly
delineated
intergroup
structure,
are
most
effective
at
re-
ducing
subjective
uncertainty.
Salience
A
key
question
for
social
identity
theory
is
what
causes
social
identity
(as
defined
earlier)
as
opposed
to
personal
identity
(self-conception
in
terms
of
unique
properties
of
self
or
of
one's
personal
relationships
with
specific
other
individuals),
or
one
social
identity
rather
than
another,
to
become
the
contextually
salient
basis
of
perception,
thought,
and
behavior?
Theory
and
research
suggest
that
there
is
an
interaction
between
category
accessibility
and
category
fit
(e.g.,
Oakes,
Haslam,
&
Turner,
1994)
that
operates
within
the
moti-
vational
framework
provided
by
self-esteem
and
un-
certainty
reduction
(see
Hogg,
1996a,
2001).
People,
influenced
by
self-enhancement
and
uncer-
tainty
reduction
motives,
categorize
the
social
context
in
terms
of
categories,
represented
as
prototypes,
which
are
chronically
accessible
in
memory
or
rendered
acces-
sible
by
the
immediate
context,
or
both.
That
categoriza-
tion
becomes
salient,
which
best
accounts
for
relevant
similarities
and
differences
among
people
in
the
context
(structural
or
comparative
fit),
which
best
accords
with
the
social
meaning
of
the
context
(normative
fit),
and
which
best
satisfies
self-enhancement
and
self-evaluative
concerns.
Situational
cues
or
personal
agenda,
or
both,
cause
people
to
"try
out"
different
cate-
gories
or
prototypes
to
make
sense
of
the
social
field
in
ways
that
also
evaluate
self
relatively
favorably.
This
is
a
fast
and
relatively
automatic
cognitive
perceptual
iter-
ative
process
that
stabilizes
when
fit
is
optimized.
Only
then
does
self-categorization
occur,
in
the
sense
of
de-
personalized
self-conception,
cognition,
and
behavior.
Social
Identity
and
Intragroup
Processes
and
Structures
Social
identity
theory
originally
focused
on
inter-
group
behavior
in
the
context
of
large
social
categories:
intergroup
social
comparisons,
positive
distinctiveness,
stereotypes,
discrimination,
and
intergroup
relations.
Intragroup
behavior
was
generally
treated
as
an
unproblematic
by-product
of
intergroup
relations;
ingroups
were
largely
treated
as
homogenous
and
undif-
ferentiated.
In
contrast,
the
recent
emphasis
on
self-cate-
gorization
and
depersonalization
has
prompted
recognition
that
groups
are
internally
structured
with
re-
spect
to
prototypicality.
Within
a
group,
some
people
are
more
prototypical
than
others.
Therefore,
deperson-
alized
social
identity
processes
within
groups
may
in-
volve
differentiation
among
people
that
is
not
interpersonal
but
is
genuinely
grounded
in
common
cat-
egory
membership.
Initially,
this
idea
was
explored
in
the
context
of
group
cohesion
and
social
attraction
(Hogg,
1992,
1993),
and
group
polarization
(e.g.,
Abrams,
Wetherell,
Cochrane,
Hogg,
&
Turner,
1990;
McGarty,
Turner,
Hogg,
David,
&
Wetherell,
1992;
see
J.
C.
Turner,
1991),
but
it
has
subsequently
pro-
duced
other
research
on
social
identity-based
intragroup
processes
(Hogg,
1996a,
1996b):
for
ex-
ample,
structural
differentiation
into
subgroups
within
groups
(e.g.,
Hornsey
&
Hogg,
1999,
2000a,
2000b),
and
deviance
and
the
"black
sheep
effect"
(e.g.,
Marques
&
Paez,
1994).
However,
perhaps
the
most
basic
and
pervasive
structural
differentiation
within
groups
is
into
leader(s)
and
followers.
Social
Identity,
Self-Categorization,
and
Leadership
Building
on
preliminary
ideas
in
a
book
chapter
(Hogg,
1996a),
I
describe
how
social
identity
pro-
cesses
influence
leadership,
followed
in
the
next
sec-
tion
by
discussion
of
empirical
support
for
these
ideas.
Leadership
is
about
how
some
individuals
or
cliques
have
disproportionate
power
and
influence
to
set
agenda,
define
identity,
and
mobilize
people
to
achieve
collective
goals.
The
differential
ability
of
some
peo-
ple
to
stamp
their
mark
on
attitudes,
practices,
deci-
sions,
and
actions
is
endemic
to
all
social
groups-for
example,
nations,
communities,
organizations,
com-
mittees,
cliques,
and
families.
Leaders
are
people
who
have
disproportionate
influence,
through
possession
of
consensual
prestige
or
the
exercise
of
power,
or
both,
over
the
attitudes,
behaviors,
and
destiny
of
group
members.
Leadership
is
very
much
a
group
process
(Chemers,
2001).
I
propose
that
there
are
three
core
processes
that
operate
in
conjunction
to
make
prototypicality
an
increasingly
influential
basis
of
leadership
processes
as
a
function
of
increasing
social
identity
salience:
prototypicality,
social
attraction,
and
attribution
and
information
processing.
These
pro-
cesses
are
described
mainly
in
terms
of
emergent
lead-
188
at Bobst Library, New York University on March 23, 2015psr.sagepub.comDownloaded from