ArticlePDF Available

Punished by Misunderstanding: A Critical Evaluation of Kohn’s Punished by Rewards and Its Implications for Behavioral Interventions with Children

Authors:

Abstract

Despite the growth of behavior analysis over the past 30 years, misunderstandings of behavioral theory and practice may threaten its continued growth and application. Alfie Kohn's Punished by Rewards: The Trouble with Gold Stars, Incentive Plans, A's, Praise, and Other Bribes (1993) offers a view of behaviorism that, if accepted uncritically, could hinder efforts to disseminate behavioral interventions, particularly those involving children. A comprehensive review of Punished by Rewards revealed the following assumptions: (a) Behavior analysis is exclusively derived from animal research; (b) behavior analysts characterize children as passive agents to be manipulated; (c) behavior analysis is a behavior control technology; (d) behavior analysts are uninterested in and ignorant of the causes of behavior; and (e) behavior analysts are unconcerned about the protection and welfare of children. For each class of assumptions, critique and commentary are offered. Empirical and philosophical aspects of Kohn's arguments are scrutinized, and the implications of continued misunderstandings for the future of behavior analysis and the need for more accurate dissemination of contemporary applied behavioral research are also discussed. Despite Kohn's apparent biases against behaviorism, it is concluded that a better informed public would find much to like about contemporary behavior analysis.
The
Behavior
Analyst
1998,
21,
143-157
No.
1
(Spring)
On
Books
Punished
by
Misunderstanding:
A
Critical
Evaluation
of
Kohn's
Punished
by
Rewards
and
Its
Implications
for
Behavioral
Interventions
with
Children
David
Reitman
Louisiana
State
University
Despite
the
growth
of
behavior
analysis
over
the
past
30
years,
misunderstandings
of
behavioral
theory
and
practice
may
threaten
its
continued
growth
and
application.
Alfie
Kohn's
Punished
by
Rewards:
The
Trouble
with
Gold
Stars,
Incentive
Plans,
A's,
Praise,
and
Other
Bribes
(1993)
offers
a
view
of
behaviorism
that,
if
accepted
uncritically,
could
hinder
efforts
to
disseminate
behavioral
interventions,
particularly
those
involving
children.
A
comprehensive
review
of
Punished
by
Re-
wards
revealed
the
following
assumptions:
(a)
Behavior
analysis
is
exclusively
derived
from
animal
research;
(b)
behavior
analysts
characterize
children
as
passive
agents
to
be
manipulated;
(c)
be-
havior
analysis
is
a
behavior
control
technology;
(d)
behavior
analysts
are
uninterested
in
and
ignorant
of
the
causes
of
behavior;
and
(e)
behavior
analysts
are
unconcerned
about
the
protection
and
welfare
of
children.
For
each
class
of
assumptions,
critique
and
commentary
are
offered.
Em-
pirical
and
philosophical
aspects
of
Kohn's
arguments
are
scrutinized,
and
the
implications
of
con-
tinued
misunderstandings
for
the
future
of
behavior
analysis
and
the
need
for
more
accurate
dis-
semination of
contemporary
applied
behavioral
research
are
also
discussed.
Despite
Kohn's
apparent
biases
against
behaviorism,
it
is
concluded
that
a
better
informed
public
would
find
much
to
like
about
contemporary
behavior
analysis.
Key
words:
behaviorism,
behavior
therapy,
critical
review,
children,
intrinsic
motivation
Is
image
everything?
If
so,
behavior
analysis
may
have
a
problem.
Behavior
analysis,
and
more
generally,
behavior-
ism,
have
been
misrepresented
in
both
the
academic
(Todd
&
Morris,
1983;
1992)
and
popular
press
(Eisenberger
Kohn,
A.
(1993).
Punished
by
rewards:
The
trouble
with
gold
stars,
incentive
plans,
A's.
praise,
and
other
bribes.
New
York:
Houghton
Mifflin.
An
earlier
version
of
this
article
was
presented
in
R.
Gallagher
(Chair),
Externalizing
behavior
disorders
in
youth:
Where
we
stand
and
where
we
should
be
heading,
at
the
30th
annual
meet-
ing
of
the
Association
for
the
Advancement
of
Behavior
Therapy,
New
York,
November
1996.
I
thank
Stephen
R.
Flora,
Robert
P.
Hawkins,
and
Joseph
C.
Witt
for
their
comments
on
an
earlier
draft
of
this
paper
and
acknowledge
re-
search
assistance
from
Steven
Hupp,
Paula
Rhode,
Rebecca
Currier,
and
Jordan
Abshire.
Correspondence
concerning
this
article
should
be
addressed
to
David
Reitman,
Department
of
Psychology,
Louisiana
State
University,
236
Au-
dubon
Hall,
Baton
Rouge,
Louisiana
70803
(E-
mail:
dreitma@unixl.sncc.lsu.edu).
&
Cameron,
1996).
Even
those
more
familiar
with
applied
behavioral
inter-
ventions
may
oppose
the
approach
on
philosophical
grounds
(Axelrod,
1996).
For
example,
Woolfolk,
Woolfolk,
and
Wilson
(1977)
found
that
preservice
teachers
who
watched
a
videotape
of
a
teacher
using
behavior
modification
techniques
rated
the
procedures
more
positively
when
the
intervention
was
described
as
"humanistic
education"
than
when
it
was
described
as
"behav-
ior
modification."
Among
undergrad-
uates
surveyed,
79%
disagreed
with
the
statement,
"It
is
appropriate
to
use
behavior
modification
with
normal
children"
(Turkat,
Harris,
&
Forehand,
1979).
In
addition,
Favell's
(1977)
in-
troduction
to
a
volume
on
behavior
modification
notes
that
behavioral
in-
terventions
have
been
unfairly
associ-
ated
with
psychosurgery
and
a
reliance
on
aversive
methods.
After
30
years,
143
144
DAVID
REITMAN
behaviorism
and
behavior
analysis
seem
to
be
in
need
of
a
makeover.
Purpose
The
intent
of
this
paper
is
to
evalu-
ate
the
basic
assumptions
and
claims
made
in
Kohn's
Punished
by
Rewards:
The
Trouble
with
Gold
Stars,
Incentive
Plans,
A's,
Praise,
and
Other
Bribes
(1993).
Where
distortions
or
misrep-
resentations
of
behavior
analysis
or
be-
haviorism
are
revealed,
clarification
or
contrary
evidence
will
be
presented.
Points
of
convergence
between
Kohn's
views
and
those
of
contemporary
be-
havior
analysts
will
also
be
highlight-
ed.
As
in
Maurice's
(1996)
emotional
address
to
the
annual
convention
of
the
Association
for
Behavior
Analysis
(ABA),
it
is
argued
that
the
misrepre-
sentations
and
philosophical
arguments
presented
against
behaviorism
in
Kohn's
text
could
add
to
existing
neg-
ative
perceptions
of
behavioral
inter-
ventions
and
thus
prevent
many
chil-
dren
from
receiving
the
care
best
suited
to
promoting
their
healthy
growth
and
development.
Punished
by
Rewards
should
be
of
interest
to
behavior
analysts
and
others
who
are
responsible
for
the
protection
and
welfare
of
children
for
several
rea-
sons.
First,
the
text
brings
together
a
large
body
of
research
and
philosophy
that
challenges
some
of
behaviorism's
and,
according
to
Kohn,
the
lay
pub-
lic's
core
assumptions
about
education
and
parenting.
Second,
Kohn
reproduc-
es
a
host
of
controversial
assertions
that
echo
those
frequently
encountered
by
behaviorists
in
applied
and
academ-
ic
settings.
Third,
in
contrast
to
more
academically
targeted
critiques
(e.g.,
Deci
&
Ryan,
1985),
Punished
by
Re-
wards
was
written
for
lay
audiences
and
has
been
widely
read
(over
63,000
copies
sold
to
date).
Other
Reviews
Previous
reviews
of
Punished
by
Re-
wards
have
been
mixed.
Organization-
al
psychologist
Montemayor
(1995)
re-
viewed
the
book
for
Personnel
Psy-
chology
and
concluded
that
Kohn's
criticisms
were
not
relevant
to
appli-
cations
of
behavioral
principles
in
vo-
cational
settings.
Clifford
(1996)
re-
viewed
the
text
for
Contemporary
Psy-
chology
and,
although
generally
sup-
portive
of
Kohn's
arguments,
questioned
the
feasibility
of
completely
abandon-
ing
contingent
reward
in
the
classroom.
Neither
author
seriously
challenged
Kohn's
broad
claims
or
the
body
of
re-
search
alleged
to
support
those
claims.
Moreover,
neither
review
examined
the
philosophical
basis
of
Kohn's
argu-
ments.
Interestingly,
no
substantive
re-
view
of
Punished
by
Rewards
has
yet
appeared
in
behaviorally
oriented
jour-
nals,
despite
its
apparent
relevance
to
behavior
analysts.
As
both
Montemayor
(1995)
and
Clifford
(1996)
recognize,
Kohn
finds
empirical
support
for
his
views
in
lab-
oratory
research
employing
both
adults
(often
college
students)
and
children
as
participants
(e.g.,
Deci
&
Ryan,
1985;
Lepper
&
Greene,
1975,
1978).
The
purpose
of
the
present
paper
is
not
to
critique
earlier
laboratory
work
on
in-
trinsic
motivation
(for
critical
reviews,
see
Dickinson,
1989;
Flora,
1990),
but
to
evaluate
Kohn's
portrayal
of
behav-
ior
analysis
and
contrast
it
with
a
sur-
vey
of
applied
research
with
children
and
adolescents.
Such
an
analysis
is
es-
pecially
critical
given
Kohn
and
oth-
ers'
concerns
about
the
potentially
harmful
impact
of
systematic
applica-
tions
of
contingent
reward
in
our
homes
and
schools.
Punished
by
Misunderstanding:
Kohn's
View
of
Behavior
Analysis
The
central
thesis
of
Kohn's
book
is
that
contingent
reward
is
undesirable
on
both
practical
and
philosophical
grounds.
Divided
into
three
parts,
Part
1
contains
Kohn's
philosophical
and
empirical
arguments
against
behavior-
ism.
Next,
Kohn
exposes
what
he
terms
the
"core
philosophy"
of
Skin-
nerian
behaviorism:
"Do
this
and
you'll
get
that."
Kohn
then
argues
against
the
allegedly
widespread
appli-
PUNISHED
BY
MISUNDERSTANDING
145
cation
of
operantly
influenced
inter-
ventions
in
school,
home,
and
voca-
tional
settings.'
In
contrast
to
popular
and
professional
notions
about
the
ef-
fectiveness
and
desirability
of
reward-
ing
children's
behavior,
Kohn
charges
that
contingent
rewards
reduce
creativ-
ity
and
risk
taking,
diminish
perfor-
mance
quality,
and
cause
children
and
adults
to
lose
intrinsic
interest
in
tasks
for
which
they
are
rewarded
(for
a
de-
tailed
response
to
these
criticisms,
see
Eisenberger
&
Cameron,
1996).
In
Part
2,
Kohn
describes
the
alleg-
edly
harmful
impact
of
rewards
in
school
and
home
settings.
Drawing
se-
lectively
from
both
published
research
and
anecdotal
accounts,
Kohn
paints
a
bleak
picture
of
how
rewards
fail
to
produce
the
favorable
outcomes
often
ascribed
to
them.
For
example,
Kohn's
critique
of
competitive grading
policies
is
directed
at
administrators,
teachers,
and
an
educational
system
that
he
per-
ceives
as
imbued
with
a
behaviorist
philosophy.
Kohn
is
especially
critical
of
what
he
considers
to
be
the
wide-
spread
thoughtless
introduction
of
ar-
bitrary
rewards
for
performance
in
home
and
classroom
settings.
Later
in
Part
2,
Kohn
presents
the
work
of
Lepper
and
Greene
(1975)
and
Deci
and
Ryan
(1985),
who,
among
others,
promote
the
view
that
both
pun-
ishing
and
rewarding
children,
although
sometimes
effective
in
generating
short-
term
compliance,
do
little
to
foster
long-term
growth
and
development.
Drawing
from
this
research
base,
Kohn
charges
that
contingent
reward
ulti-
mately
yields
children,
adolescents,
and
adults
who
lack
intrinsic
interest
in
learning
or
work
and
whose
behavior
is
controlled
nearly
exclusively
by
extrin-
sic
motivators
such
as
money
and
praise.
Unfortunately,
long-term
(e.g.,
longer
than
1
month)
controlled
studies
of
the
impact
of
contingent
reward
on
'
Kohn's
comments
on
behavioral
interven-
tions
in
the
workplace
are
not
considered
here,
but
see
Montemayor
(1995)
for
relevant
criti-
cisms
of
Kohn's
views
as
they
apply
to
voca-
tional
settings.
either
achievement
or
task
interest
are
rare,
and
the
few
existing
studies
appear
to
be
inconsistent
(see
Kohn,
1993,
note
13,
p.
299).
No
evidence
of
detrimental
long-term
effects
of
the
sort
necessary
to
justify
Kohn's
concerns
is
readily
available.
However,
neither
is
there
compelling
evidence
to
dismiss
Kohn's
claims.
Specifically,
all
known
studies
of
the
alleged
detrimental
effects
of
re-
ward
involve
only
brief
exposure
to
re-
wards,
so
conclusions
about
their
poten-
tial
harmful
effects
await
studies
in
which
the
participants
are
exposed
to
contingent
reward
for
much
longer
pe-
riods
(e.g.,
weeks
or
months)
(Bern-
stein,
1990).
In
Part
3,
Kohn
describes
alterna-
tives
to
methods
based
on
what
he
terms
the
"carrot
and
stick"
approach.
As
noted
by
Clifford
(1996),
these
ideas
are
less
well
developed
than
Kohn's
exhaustive
criticisms
of
behav-
iorism.
Kohn's
introduction
of
"the
three
Cs"-collaborative
learning
en-
vironments,
modified
course
content,
and
increased
student
choice-is
fol-
lowed
by
the
suggestion
that
teachers
adopt
this
approach
and
abandon
the
use
of
contingent
reward
to
motivate
children.
Unfortunately,
however
at-
tractive
Kohn's
recommendations
may
be,
little
empirical
evidence
suggests
that
his
solutions
will
yield
better
ed-
ucational
outcomes
than
existing
prac-
tices
(Clifford,
1996).
For
example,
Slavin's
(1991)
critical
examination
of
the
cooperative
learning
literature
led
to
the
conclusion
that
achievement
gains
were
produced
only
when
both
group
and
individual
rewards
were
programmed.
Commenting
on
the
somewhat
paradoxical
popularity
of
cooperative
learning,
Slavin
noted,
The
cooperative
learning
movement
has
created
an
interesting
phenomenon,
in
which
humanistic
educators
and
psychologists
are
championing
classroom
methods
that
could
be
described
com-
pletely
in
behavioral
language....
However,
the
attraction
of
cooperative
learning
for
many
hu-
manistic
educators
probably
lies
not
so
much
in
accelerating
student
achievement
as
in
the
con-
sistently
found
positive
effects
on
such
variables
as
race
relations,
attitudes
toward
mainstreamed
classmates,
self-esteem,
and
other
non-academic
146
DAVID
REITMAN
outcomes....
In
contrast
to
achievement
effects,
these
important
outcomes
do
not
appear
to
de-
pend
on
the
use
of
group
rewards
for
individual
learning.
(p.
113,
italics
added)
Thus,
although
the
promotion
of
col-
laboration,
choice,
and
changes
in
les-
son
content
may
enhance
some
aspects
of
children's
educational
experiences,
it
does
not
follow
from
the
existing
re-
search
that
we
ought
to
abandon
the
use
of
reward
in
the
classroom.
Prop-
erly
employed,
extrinsic
rewards
can
be
important
and
useful
motivational
tools
(Travers,
Elliot,
&
Kratochwill,
1993).
At
the
same
time,
the
delivery
of
arbitrary
rewards
for
good
perfor-
mance
does
not,
by
itself,
constitute
a
well-conceived
educational
plan.
As
will
be
described
later,
contemporary
behavioral
approaches
to
teaching
con-
centrate
more
on
instructional
design
than
on
arbitrary
schemes
for
the
de-
livery
of
extrinsic
rewards.
An
Alternative
Perspective
on
Behavior
Analysis:
A
Reply
to
Kohn
Far
from
being
comprehensive,
Kohn's
selected
literature
review
high-
lights
failures
while
ignoring
the
many
successes
associated
with
behavioral
interventions.
In
particular,
Kohn
seems
to
be
oblivious
to
behavioral
in-
terventions
that
do
not
hold
compli-
ance
as
an
explicit
goal
(see
Milten-
berger,
1997,
for
many
examples
of
the
successful
application
of
behavioral
principles
that
do
not
involve
compli-
ance
training)
and
is
largely
unaware
of
the
behavioral
literature
debating
the
ethics
of
reward
and
punishment
(e.g.,
Van
Houten
et
al.,
1988;
Wolpe,
1978).
On
the
whole,
Punished
by
Rewards
presents
a
rather
unattractive
behavior-
al
straw
man
that
vastly
underrates
the
utility
of
behavioral
interventions
and
disingenuously
portrays
the
motives
of
parents,
teachers,
and
behavior
modi-
fiers.
As
will
later
be
shown,
there
are
many
points
of
correspondence
be-
tween
Kohn's
views
and
those
of
ap-
plied
behaviorists.
Still,
the
misrepre-
sentations
of
contemporary
behavior-
ism
present
in
Punished
by
Rewards
may
cause
considerable
concern
among
school
administrators,
teachers,
parents,
and
others
who,
up
until
now,
have
been
receptive
to
behavioral
in-
terventions.
In
citing
Charles
Dar-
win-"Great
is
the
power
of
steady
misrepresentation;
but
the
history
of
science
shows
that
fortunately
this
power
does
not
long
endure"-Todd
and
Morris
(1992,
p.
1450)
hoped
that
misrepresentations
of
behaviorism
would
ultimately
give
way
to
more
even-handed
accounts.
Punished
by
Rewards
is
formidable
evidence
that
prejudiced
accounts
have
yet
to
give
way
to
fair
representation.
Consequent-
ly,
a
more
complete
examination
of
the
most
consistently
voiced
misunder-
standings
and
misrepresentations
re-
flected
in
Kohn's
text
may
be
neces-
sary
to
assist
behavior
analysts
in
re-
sponding
to
the
concerns
about
behav-
iorism
raised
in
Punished
by
Rewards
and
to
foster
more
accurate
represen-
tation
of
behavior
analysis
in
the
fu-
ture.
Using
direct
quotations
from
Pun-
ished
by
Rewards,
the
following
class-
es
of
misrepresentation
or
misunder-
standing
were
identified:
(a)
Behavior
analysis
is
exclusively
derived
from
animal
research;
(b)
behavior
analysts
characterize
children
as
passive
agents
to
be
manipulated;
(c)
behavior
analy-
sis
is
a
behavior
control
technology;
(d)
behavior
analysts
are
uninterested
in
and
ignorant
of
the
causes
of
behav-
ior;
and
(e)
behavior
analysts
are
un-
concerned
about
the
protection
and
welfare
of
children.
Following
each
as-
sumption
and
representative
quotations
from
the
text,
critique
and
commentary
are
offered.
Behavior
analysis,
and
more
broad-
ly
behaviorism,
is
based
primarily
upon
work
that
has
been
conducted
with
animals,
in
particular,
rats
and
pigeons.
As
behaviorists
cheerfully
admit,
theories
about
rewards
and
various
practical
programs
of
be-
havior
modification
are
mostly
based
on
work
with
rats
and
pigeons.
...
Yet
it
is
not
an
acci-
dent
that
the
theory
behind
"Do
this
and
you'll
PUNISHED
BY
MISUNDERSTANDING
147
get
that"
derives
from
work
with
other
species,
or
that
behavior
management
is
frequently
de-
scribed
in
words
better
suited
to
animals.
(p.
24)
Although
largely
true
in
1960,
it
is
no
longer
the
case
that
the
empirical
base
of
behavior
analysis
derives
exclusive-
ly
from
animal
research.
Over
the
past
30
years,
basic
research
and
innovation
in
therapy
have
extended
the
principles
of
applied
behavior
analysis
to
human
participants.
Basic
research
examining
phenomena
such
as
stimulus
equiva-
lence
and
instructional
control
has
been
conducted
almost
exclusively
with
humans
(e.g.,
Hayes
&
Hayes,
1992;
Sidman,
1994).
Since
the
early
1980s,
there
has
been
a
behavioral
journal
dedicated
to
the
examination
of
human
verbal
processes
(The
Analysis
of
Verbal
Behavior),
and
the
Journal
of
the
Experimental
Analysis
of
Behav-
ior,
a
premier
behavioral
experimental
journal,
has
published
an
ever-increas-
ing
number
of
articles
involving
hu-
mans
(Nevin,
1982).
Similarly,
the
Journal
of
Applied
Behavior
Analysis