BookPDF Available
Mindset
Theo
of
Action
Phases
ABSTRACT
Mindset theory of action
phases
is
based
on
the
distinction between motivation
and
volition
as
proposed
by
the Rubicon model which claims that
prior to
crossing
the
Rubicon
(i.e., making a goal
decision) motivational principles apply whereas
thereafter volitional principles
set
in.
The
latter
are
concerned with goal implementation, whereas the
former relate
to
the choosing of
goals.
Mindset
theory
of
action
phases
proposes that different
cognitive procedures
are
activated when people
tackle the task
of
choosing goals
versus
imple-
menting them.
The
respective task demands
determine the features that characterize the delib-
erative
versus
implemental mindset.
These
pertain
to wh at type
of
information
is
preferably
processed
and
how
it
is
analyzed. Mindset
research
has
produced findings that not only support the
motivation
versus
volition distinction but
also
enlighten various debates
and
theories
in
sodal
psychology (e.g., optimism
versus
realism
debate,
dual
process
theories, goal theory). Mindset theory
of action
phases
has
also
spurred
research
on
effective planning
by
pointing to implementation
intentions (i.e., if-then
plans).
This
research
has
had
much applied impact. When it
was
linked up
to
research
on
strategies of motivationally smart
goal setting (i.e., mental contrasting), it initiated
the development
of
a time-
and
cost-effective
behavior change intervention.
Peter
M.
Gollwitzer
INTRODUCTION
During my graduate education in the late
1970s at the University
of
Texas at Austin,
my mentor, Robert Wicklund, and I started to
conceive
of
people's selves or identities
as
goals.
We
thought that people can very weil
set themselves goals to become a good
parent, a brilliant scientist, or a great athlete.
If
one takes this perspective, the self
of
a
person
is
no longer just something to under-
stand (self-concept) and like (self-esteem)
but something
to
be achieved (identity goal).
We
turned to the writings
of
Kurt Lewin
(1926) and his students, whose tension
system theory of goal pursuit, with its notion
of substitution,
was
very helpful to develop-
ing our theory of symbolic self-completion
(Gollwitzer and Kirchhof, 1998; Wicklund
and Gollwitzer, 1982). The main proposition
of
self-completion theory is that once people
have set themselves certain identity or self-
defining goals, they respond
to
failure expe-
riences, shortcomings, or barriers not with
retreat but instead with intensified efforts to
reach the goal. These efforts, however,
do
not
f
First publ. in: Theories of social psychology / Lange, Paul A. van (Ed.). - Los Angeles [u.a.] : Sage,
2012. - pp. 526-545. - (Handbook of theories of social psychology ; 1). - ISBN 978-0-85702-960-7
Konstanzer Online-Publikations-System (KOPS)
URL: http://nbn-resolving.de/urn:nbn:de:bsz:352-179905
MINDSET THEORY
OF
ACTION
PHASES
527
have to alleviate the problem
at
hand but may
involve resorting to any substitute that
indicates goal attainment
Ce.g.,
showing off
relevant status symbols, engaging in identity-
relevant activities, describing oneself as
having the required personal attributes; for
recent research on self-completion theory see
Gollwitzer et al., 2009; Harmon-Jones et al.,
2009; Ledgerwood et
a1.,
2007).
In
the early 1980s, Heinz Heckhausen
invited
me
to join the newly founded Max
Planck Institute for Psychological Research
at Munich to start a research unit called
Motivation and Action.
We
quickly realized
that we had a directly opposed conceptual
view of motivation. Whereas Heinz
Heckhausen 's motivation was that
of
an
expectancy-value theorist in the tradition
of
Atkinson (1957) and Heckhausen (1977)
and was thus fueled by the perceived feasibil-
ity and desirability
of
a given action,
my motivation was that
of
Lewin's (1926)
tension system and was resting in the deter-
mination or commitment a person holds with
respect
to
the action goal at hand. Apparently,
in the research on self-completion (Wicklund
and Gollwitzer, 1982) I had been studying
issues of goal striving (i.e., thoughts and
behavior directed toward existing goals),
whereas Heinz Heckhausen
in
his work on
achievement motivation (Heckhausen, 1977)
had focused on issues
of
goal setting (i.e.,
what goals people find attractive and feasi-
ble, and thus choose for themselves).
THE
RUBICON
MODEL
OF
ACTION
PHASES
To highlight this insight, we suggested
making a distinction between motivation and
volition. Following the conceptual terms
used by Lewin (1926) and Narziß Ach (1935),
we dubbed the goal-striving-related motiva-
tion with the term volition, and kept the term
motivation for the goal-setting-related
motivation. More importantly,
in
an attempt
to integrate these two kinds of phenomena
(i.eo,
motivation and volition) we developed
the Rubicon model
of
action phases
(Heckhausen, 1987;
Heckhausen
and
Gollwitzer, 1987). This model suggests that
the course
of
action can be segmented into
four different, consecutive phases that differ
in terms
of
the tasks that are to be solved by
the individual given that s/he wants to exe-
cute a given course of action successfuIly.
The first phase (predecision phase) is said to
pose the task
of
setting preferences among
wishes and desires by deliberating their
desirability and feasibility. As people's
motives and needs produce more wishes and
desires than can possibly be realized, the
individual is forced to choose among these
desires and by doing so turn them into goals.
Once goals are set (i.e., the Rubicon has been
crossed), the individual faces the second task
(preaction phase), which is getting started
with goal-directed behaviors. This may be
simple
if
the necessary goal-directed actions
are weil practiced and routine but complex if
the individual
is
still undecided about where,
when, and how
to
act.
In
such complex cases,
the execution
of
goal-directed action has to
be planned by deciding on when, where, and
how to act. The third task (action phase)
is
bringing the initiated goal-directed action to
a successful ending, and this is best achieved
by determined and persistent pursuit
of
goal
completion. Finally, in the fOlllth task (postac-
tion phase), the individual needs to decide
whether the desired goal has indeed been
achieved or wh ether further striving is
needed.
The Rubicon model
of
action phases
postulates that a person's psychological func-
tioning in each
of
these phases is governed
by different principles. Classic theories
of
motivation (adhering to the restricted defini-
tion
of
motivation as determined by feasibil-
ity and desirability; Atkinson, 1957; Feather
and Newton, 1982; Heckhausen, 1977) are
said to be weIl suited to explicate the psycho-
logical processes associated with the prede-
cision and postaction phases, whereas
theories
of
volition (i.e., theories on the self-
regulation
of
goal attainment; Lewin, 1926;
528
HANDBOOK
OF THEORIES OF SOClAL PSYCHOLOGY
Mischel, 1974; Mischel and Patterson, 1978)
are most appropriate to explaining the psy-
chological processes that characterize the
preaction and action phase. In other words,
the predecision and postaction phases are
expected to encompass motivation al phe-
nomena and processes in the classic sense
of
the term, whereas in the phases in between
volitional phenomena and processes are
thought to occur.
This radical statement needed empirical
support, and therefore Heckhausen and I con-
ducted an early experiment aimed at demon-
strating that individuals placed in the
predecision phase evidence different cogni-
tive functioning than do individuals in the
preaction phase (Heckhausen and Gollwitzer,
1987, Study 2). Assuming that deliberation
of
the desirability and feasibility
of
wishes and
desires (the task
of
the predecision phase)
is
cognitively more demanding than commit-
ting to a plan that specifies, when, where, and
how one wants to perform goal-directed
actions (the task
of
the preaction phase), we
expected that deliberating individuals experi-
enced a higher cognitive load than planning
(i.e., preaction) individuals.
We
therefore
interrupted experimental participants who
were either in the middle
of
deliberating a
choice between two different tests that pre-
sumably measured their creative potential or
in the middle
of
planning how to perform the
test they had just chosen and then asked them
to take a short-term memory test (i.e., a noun
span test that presented nouns ilTelevant
to
the creativity tests at hand).
We
expected that
deliberating participants, because
of
height-
ened cognitive load, would evidence a reduced
noun span, compared with their span
as
measured at the beginning
of
the experiment.
We
also expected that deliberating partici-
pants would evidence a comparatively more
reduced noun span than planning participants
because laying down a plan on how
to
act was
expected to take
up
less cognitive resources
than deliberating the pros and cons
of
a goal
decision.
To
our surprise, the results were just oppo-
site to what we had predicted (Heckhausen
and Gollwitzer, 1987, Study 2). The deliber-
ating participants showed an increase in their
short-term memory capacity, compared with
both their
own
prior span and the span
of
the
planning participants.
In
an effort to reduce
our confusion about these unexpected
findings, I turned to Gerhard Strube, at the
time a cognitive psychologist at the Max-
Planck-Institute for Psychological Research,
and he pointed me to the classic concept
of
mindset
as
originally advanced at the turn
of
the century
by
the German psychologists
Külpe (1904), Marbe (1915), Orth (1903),
and Watt (1905), all members
of
the Würzburg
school. These early cognitive psychologists
had discovered that becoming intensively
involved with performing a given task acti-
vates exactly those cognitive procedures that
help task completion. The created mindset
(i.e., the sum total
of
the activated cognitive
procedures)
is
the cognitive orientation most
conducive
to
successful task performance.
The mindset notion allows interpreting
the observed noun span data
as
folIows:
deliberating between potential action goals
activates cognitive procedures (the delibera-
tive mindset) that facilitate the task
of
the
predecision phase, which is to set prefer-
ences.
As
undecided individuals do not know
yet in which direction their decisions will
finally take them, a heightened receptiveness
to all kinds
of
information (open-minded-
ness) seems appropriate and functional
to
task solution. Similarly, planning out the
implementation
of
a chosen goal should
activate cognitive procedures (the implemen-
tal mindset) that facilitate the task
of
the
preaction phase (i.e., getting started on the
chosen goal).
As
this requires a more focused
and selective orientation to processing
information, closed-mindedness rather than
open-mindedness with respect
to
available
information seems called
for.
This postulated
difference in receptiveness between deliber-
ating and planning individuals
is
expressed
in the fact that the experimental participants
in the Heckhausen and Gollwitzer (1987,
Study
2)
noun span study processed the
presented information in the noun span task
MINDSET
THEORY
OF
ACTION
PHASES
529
faster than planning partieipants (i.e., the delib-
erating participants demonstrated a broader
noun span than the planning participants).
MINDSET
THEORY
Of
ACTION
PHASES
However, isn't all
of
this post hoc? This
is exactly the type
of
worry that made
me
use the mindset notion
as
a hypothesis-
generating device for subsequent research
and thus for developing a comprehensive
mindset theory
of
action phases
in
my
Habilitationsschrift (i.e., a second, more
extensive doctoral thesis that in Germany
is
a
prerequisite for attaining a tenured professor-
ship; Gollwitzer, 1987). A summary
of
this
thesis (Gollwitzer, 1990) can be found in a
chapter in Motivation and Cognition edited
by
E.
Tory Higgins and Richard M. Sorrentino
(1990), and a more extensive version
in
a
German book, Abwägen
und
Planen
[Deliberating and Planning] (Gollwitzer,
1991).
If
one analyzes the unique demands
of
the task
of
choosing between wishes and
desires in the predecision phase versus the
typical demands
of
the task of getting started
on a chosen goal in the preaction phase, it
becomes possible to detect further cognitive
features
of
the deliberative
as
compared
to
the implemental mindset that can then be
tested
in
new experiments. The task
of
delib-
erating in the predecisional phase is to
choose, from among various wishes and
desires, those
few
that one wants
to
realize
(Gollwitzer, 1990). The criteria for selection
should be the feasibility and desirability
of
the wishes and desires at issue. The system-
atie analysis
of
the chances of realization
as
weH
as
the desirability
of
realization requires
that relevant information
be
preferentially
encoded and retrieved. But such cognitive
tuning
to
this information should not suffice,
as feasibility-related information needs to
be analyzed objectively (and not in a self-
serving manner), and desirability-related
information needs to be analyzed in an
impartial manner (and not in a biased
manner). Only
if
feasibility-related informa-
tion
is
analyzed realistically, and the pros and
cons are weighed impartially, can the indi-
vidual turn those desires into binding goals
that can potentially be realized and possess a
genuine attractiveness. Moreover, deliberat-
ing requires a general open-mindedness (as
was demonstrated in the Heckhausen and
Gollwitzer [1987] study described above)
with respect
to
any available information,
as
undecided individuals do not know yet in
whieh direction their decision will finally
take them.
Once a goal decision has been made, the
task
of
planning
is
to
promote the initiation
of
goal-directed behavior. This requires com-
mitting oneself to when, where, and how
to
get started. Accordingly, one needs to dis-
cover good opportunities and link them to
appropriate goal-directed actions, thus creat-
ing plans for action. For this purpose, cogni-
tive tuning toward implementation-related
issues should be beneficial. Feasibility-
related and desirability-related issues should
no
longer matter, and,
if
forced on the indi-
vidual, they are avoided by distorting the
relevant information in support
of
the goal
decision made: the person sees the feasibility
of the chosen goal in an overly optimistic
way,
and views the desirability
of
the chosen
goal in a partial manner (i.e., pros exceed
cons). Finally, processing all
of
the available
information in an open-minded manner
should be dysfunctional, as it might derail
the individual from the chosen course
of
action. Accordingly, a reduced open-minded-
ness (cIosed-mindedness) favouring the
selective processing
of
information in sup-
port
of
the chosen goal
is
to be expected.
Given these different features
of
the delib-
erative and implemental mindsets, one should
not forget that the two different
mi
nd sets also
possess many similar attributes. For instance,
the mindset theory
of
action phases assumes
that both deliberative and implemental mind-
sets become more pronounced
as
a person
gets more involved with deliberating between
potential goals and with planning chosen
530
HANDBOOK
OF
THEORIES
OF
SOClAL PSYCHOLOGY
goals, l'espectively. Mol'eovel', neither
mi
nd-
set should immediately vanish when the task
activity that produced it is ended; instead, the
mindset should show a moment
of
inertia.
This implies that the cognitive orientations
associated with the deliberative and imple-
mental mindsets can be detected in theil'
effects on performing temporally subsequent
tasks
of
a different nature. These ideas regard-
ing the similarity between the deliberative
and implemental mindset have been used to
develop a research program aimed at testing
the proposed different cognitive features
of
the deliberative and implemental mindsets.
In this research, the following method
01'
inducing the deliberative and implemental
mindsets turned out to be most efIective:
experimental participants are asked either to
extensively deliberate an unresolved personal
problem to be named by the participants
(who indicate problems such as, "Should
I move to another city or not?;' "Should
I change my major?," "Should I buy a new
car?,"
or
"Should I get involved with some-
body?") or to plan the implementation
of
a
chosen goal indicated by the participants
(projects such as,
"I
will move to another
city,"
"I
will change
my
major," etc., are
named). These requests create a deliberative
and an implemental mindset, respectively.
To
intensify these mindsets, deliberating partici-
pants are asked to list the shorHerm and
long-term pros and cons
of
making and not
making
adecision,
in order to get heavily
involved with deliberating. Planning partici-
pants, on the other hand, are asked to list the
five most important steps
01'
implementing
the chosen goal, and then to specify when,
where, and how they intend to execute each
step, all
01'
which serves the purpose
01'
creating an intensive involvement with
planning. Thereafter, both the deliberating
and the planning participants are asked to
perform presumably unrelated tasks (usually
presented by a different experimenter in a
different situational context), which
are designed to measure the very cognitive
features hypothesized to differ between
the deliberative and implemental mindsets.
This procedure
of
inducing the deliberative
and implemental mindsets in one situation al
context and assessing their cognitive and
behavioral consequences in a different
setting, has been referred to as procedural
priming or mindset priming (Bargh and
Chartrand, 2000), as research participants
commonly stay unaware
of
the mindset
effects they evidence.
Deliberative versus implemental
mindsets
and
cognitive tuning
The hypothesis that the deliberative mindset
creates cognitive tuning toward information
relevant to making goal decisions (informa-
tion on feasibility and desirability), whereas
the implemental mindset tunes a person's
cognitions to implementation-related infor-
mation (information on where, when, and
how to act), was tested most critically by
Gollwitzer et al. (1990). Participants were
placed into either a deliberative
01'
an imple-
mental mindset by having them deliberate on
unresolved personal problems or plan chosen
goal projects, respectively (the standard
procedure described above was used).
In
a
presumably unrelated second part
01'
the
experiment, participants were presented with
the first few lines
01'
a number
01'
novel fairy
tales and were instructed to complete each
tale. Even though participants were allowed
to continue the stories in any way they liked,
deliberating participants had the protagonists
01'
the tales reflect
on
reasons for choosing or
not choosing certain action goals to a greater
degree than planning participants did.
Thoughts about how to accomplish a chosen
goal, however, were more frequently attrib-
uted to the protagonists by planning partici-
pants than by deliberating participants.
Focusing on the processing
01'
mindset-
congruent information, Gollwitzer et al.
(1990, Study 2) conducted an experiment in
which pa1'ticipants had to recall the p1'esented
delibe1'ative and implemental thoughts
01'
others. Patiicipants were placed into either a
deliberative
01'
an implemental mindset by
MINDSET THEORY
OF
ACTION
PHASES
5311
having them reflect the choice
of
one of two
tests (i.e., decide between two different
creativity tests) or plan to perform a chosen
test. While participants were involved
in
deliberating 01' planning, slides were
presented that depicted different persons
mulling over personal decisions. For exam-
pIe, a depicted elderly lady
was
thinking
of
the pros (i.e.,
"It
would be good because ... ")
and cons (i.e., "It would be bad because ... ")
of
having her grandchildren spend their
summer vacation at her home. For each
of
these slides, next to the pros and cons
of making
adecision,
potential plans
of
implementation were also presented. These
specified how the person would get started
with the particular goal-directed actions (i.e.,
"If
I decide
to
do it, then I will first
...
and then
...
!"; "If I decide to do
it,
then
I won't ... before
...
!"). A cued-recall test
of
this information was given following a dis-
tractor task;
it
provided participants with the
pictm'es
of
the persons they had viewed
and the stems
of
the sentences (as above)
describing their thoughts. The deliberating
participants, who had to view the slides and
to
recall the information depicted on the
slides prior to making
adecision
about the
two types
of
creativity tests, recalled
pros and cons bettel' than they recalled infor-
mation on the when, where, and how
of
implementation. The recall performance
of
the planning participants,
who
had received
and recalled the information after adecision
on the creativity tests had been made, showed
the reverse pattern.
All
of
these findings corroborate the
cognitive-tuning hypothesis. But how do
these differential recall performances
observed in the last study (Gollwitzer et
a1.,
1990, Study 2) come about?
If
one assumes
that individuals' retrieval attempts necessi-
tate constructing descriptions
of
what they
are trying to retrieve (Norman and Bobrow,
1979),
it
seems possible that mindsets pro-
vide perspectives (Bobrow and Winograd,
1977) that allow the easy construction
of
specific descriptions. The deliberative mind-
set for instance should favour descriptions
phrased in terms of pros and cons, benefits
and costs, and so forth. In other words,
the deliberative mindset supports the ready
construction
of
descriptions that specify
desirability-related information, whereas the
implemental mindset supports the construc-
ti
on
of
descriptions that specify implementa-
tion-related information. As Norman and
Bobrow (1979) point out, quick construction
of
specific descriptions at the time
of
retrieval
facilitate further successful retrieval. Norman
and Bobrow also assume that whenever the
description
of
the information sought matches
the elaboration
of
the information at the time
of encoding, recall performance is particu-
larly enhanced.
It
seems possible, then, that
deliberative and implemental mindsets favor
cortgruent recall through both congruent
elaboration at the time
of
encoding and ready
construction
of
congruent descriptions at the
time
of
retrieva1.
Deliberative versus
implemental mindsets
and
biased
inferences
Deliberative and implemental mindsets are
also postulated to differentially affect the
way in which feasibility-related and desira-
bility-related information is handled. In a
deliberative
mi
nd set, information related
to
desirability should be analyzed impartially;
in an implemental mindset, an analysis
partial
to
the chosen goal
is
expected. Also,
feasibility-related information
is
expected
to
be analyzed rather accurately in a delibera-
tive mindset, whereas optimistic inferences
that overestimate the actual feasibility
of
the
chosen goal are expected in an implemental
mindset.
Desirability-related
information
With respect
to
testing the postulated impar-
ti
al versus partial analysis
of
desirability-
related information, a first study (reported
by
Taylor and Gollwitzer, 1995, Study
3)
was
conducted
by
asking participants
to
name
either potential goals 01' chosen goals and
532
HANDBOOK
OF
THEORIES
OF
SOClAL PSYCHOLOGY
subsequently attempt
to
achieve clarity
on
the question of whether they should make an
affirmative decision or had made the correct
decision, respectively. Whereas the predeci-
sional participants reported on positive and
negative consequences wirh the same fre-
quency, postdecisional participants failed to
do so. The latter reported about five times
more thoughts about pros than about cons,
indicating a strong partiality in favor
of
the
chosen goal in postdecisional participants.
Evidence for differences between the
deliberative and implemental mindset in
processing pros and cons
is
also provided by
Harmon-Jones and Harmon-Jones (2002,
Study 2). They tested the effects
of
mindsets
on the postdecisional spreading of alterna-
tives, a classic cognitive dissonance para-
digm (Brehm and Cohen, 1962). Using this
paradigm, dissonance researchers have found
that after making a choice between two
options, the chosen option becomes evalu-
ated more positively whereas the nonchosen
option becomes evaluated more negatively.
Harmon-Jones and Harmon-Jones found that
the implemental mindset increased postdeci-
sional spreading
of
alternatives, whereas the
deliberative mindset reduced it.
There
is
an important set
of
studies by
Gagne and Lydon (2001
a)
suggesting that
deliberation only then leads to an impartial
analysis
of
pros and cons when deliberation
is linked to the predecisional action phase.
Deliberation over goal decisions that have
already been made can initiate defensive
processing
of
information that leads to even
greater biasing. In one study, they asked par-
ticipants involved in romantic relationships
to deliberate a relationship or a nonrelation-
ship goal decision. They found that when
asked to rate how their partner compared
with the average, those individuals asked to
deliberate over a relationship goal decision
gave much higher ratings than those
who were asked to deli berate over a nonrela-
tionship goal decision.
Of
interest, these
ratings were also higher than those
of
imple-
mental participants who had been planning
the implementation
of
a relationship goal.
Gagne and Lydon (2001a) argue that the
deliberation
of
a relationship goal may have
been perceived
as
threatening, resulting in
greater enhancement
of
the partner's
attributes. In a second study, they measured
the commitment participants had to their
relationship and found that high-commitment
but not low-commitment partlclpants
defended against the threat
of
a deliberative
mindset by increasing their positive views
of
their partner. This pattern
of
findings indeed
supports the assumption that deliberation
may have threatened the participants' per-
ceived ability to attain the goal
of
maintain-
ing the relationship. In response, these
individuals reasserted their commitment to
the relationship
by
boosting the ratings
of
their partner.
Feasibility-related
information
The hypothesized accurate analysis
of
feasi-
bility-related information in the deliberative
mi
nd
set, and the expected overly optimistic
assessment in the implemental mindset, were
observed in experiments by Gollwitzer
and Kinney (1989) using the contingency-
learning task designed by Alloy andAbramson
(1979). In this task, participants are asked to
determine to what degree they can influence
the on set
of
a target light (outcome) by
choosing
to
press 01' not to press a button
(alternative actions). Participants commonly
go through
aseries
of trials (at least 40); the
start
of
each trial
is
indicated
by
a warning
light. By observing whether 01' not the target
light comes on after they have pressed 01' not
pressed the button, participants estimate how
much control they have over the target light
onset. The experimenter varies the actual
control by manipulating the frequency
of
the target light onset associated with each
of
the two action alternatives (pressing or not
pressing). The smaller the difference between
these two frequencies, the less objective
control participants have over the target light
onset.
Nondepressed individuals commonly
claim
to
possess control over target light
on set that
is
noncontingent on their actions,
f
e
MINDSET THEORY
OF
ACTION
PHASES
533
whenevcr thc target light onset occurs
fre-
quently in thc "75/75" problem, whcre
the target light comes on
in
75
percent
of
pressing and 75 percent
01'
nonpressing
responses: see Alloy
Clnd
Abramson. 1979).
Gollwitzer
and Kinney (1989, Study 2)
askcd deliberating, planning, and control
participants to work on a contingency
problem that presented frequent and noncon-
tingent target light onset (i.e., the 75/75
problem). Participants were given the instruc-
tion to discover how to produce the target
light onset. A set
of
40 trials was offered, and
participants were then asked to
judgc
how
much control they
cOLlld
exert over the target
light onset.
Deliberating participants showed the most
accurate
judgment
of
control; their judg-
ments
of
control were lower than those
of
eüher
the control group or the planning
group. The planning participants'
judgments
of
control tcnded to be even higher than those
of
the control participants. The mindsets
were created via thc standard proccdure
described above. A mindset interpretation