CHAPTER2
Action Phasesand Mind-Sets
P E T EM
R . GOLLWITZER
Forschung
f ür psychologische
Max-Planck-/nstitut
The focus of this chapter is on the course of action, which is understood to be a
temporal, horizontal path starting with a person's desires and ending with the
evaluation of the achieved action outcome. The phenomena of choosing an action
goal, initiating the appropriate actions, and executing these actions are assumed
to be situated in between. This comprehensive perspective conceivesof the course
of action as a number of consecutive, distinct segments or phases. It raises
questions concerning how people choose action goals, plan and enact their
"mind-set"
is employed to
execution, and eaaluate thek efforts. The concept of
find answers to these questions in terms of the cognitive processesor orientations
that allow for easy completion of the different action phases.
A PHASEMODELOF ACTION
Goal Settingand Goal Striving
"Being
motivated" implies a number of different phenomena. But how many
distinct aspectsof being motivated to pursue a desired goal are there? Kurt Lewin
(Lewin, Dembo, Festinger, & Sears, 1944) made a major distinction between goal
"Goal
striving" is behavior directed toward existing
striving and goal setting.
"Goal
goals, and thus addressesquestions of moving toward the chosen goal.
setting," on the other hand, addressesthe question of what goals a person will
choose, and thus considers the expected value of the available choice options.
Noticing the unique nature of both of these problems, Lewin adopted a distinct
theoretical perspective for each of them. He referred to an expectancy X value
model when goal setting was at issue-for instance, when he and his colleagues
were attempting to explain people's changes in aspiration level (Lewin et al.,
1944).Issues of goal striving, however, were explained in terms of his theory of
tension systems (Lewin, 1926), through which he tried to discover the forces that
move a person toward a chosen goal. Lewin considered the strength of these
forces to be related not only to the valence of the chosen goal, but also to the
individual's perceived distance from the goal. By introducing the variable of
t3
54
and Self-Control
Part I. Selt'-Regulation
porency, Lewin (1936) tried to explain which of the many goals people entertain
at a given time actually guide their behavior in specific situations.
German researchersstudying goal-oriented beha'rior before Lewin devoted
themselves solely to the issue of goal striving by studying the initiation and
execurion of actions serving chosen goals or instructions given by others. This re"wii1
psychology"; its rnost
search tradition has become known as the German
prominent figure was Narziss Äch (1905, 1910, 1915). Researchersin this
tradition fiercely disagreed on the key questions of will psychology, such as
whether intentions specifying an appropriate opportunity to act favor smooth
action initiation, regardlessof the importance of the respective superordinate goal
(Ach, 1935; Selz, 1910). Nevertheless, they unanimously considered goal striving
"battle
of motives."
to be different from goal setting, which they referred to as the
For German will psychologists, it was clear that people's goal setting depends
primarily on their desires, needs, and interests, whereas a host of additional
variables determines whether and how people act on their chosen goals.It was the
latter that they tried to identify and that they explored in their theories.
Researchersstudying goal-oriented behavior after Lewin shifted their attention from goal striving to goal setting. Stimulated by Lewin's as well as Festinger's
work on shifts in the level of aspiration (Festinger,1942; Lewin et al., 1944),
Atkinson (19t7) presented a formal model of risk taking that made it possible to
compute the motivational tendency to choose tasks representing various difficulty
levels. Like Lewin, Atkinson considered goal setting and goal striving to be the
two major problems requiring solution in any psychology of motivation:
The first problem is to accountfor an individual'sselectionof one path of action
among a set of possiblealternatives.The secondproblem is to accountfor the
amplitude of the action tendencyonce it is initiated, and for its tendencyto
persistfor a time in a given situation.(1957,p. 359)
However, Atkinson hoped that the two distinct problems could be reduced to
one and thereby could be accounted for by a single theoretical model. In his view,
the theory best suited to accomplish this purpose was the risk-taking model
(Ätkinson, 1957) or a modified version of it (Atkinson & Reitman, 1956). This
model implies that the motivational tendency that makes a person choose a
certain task will also account for the effort the person exerts when working on the
chosen task and for the quality of the achieved performance.
Empirical studies investigating this implication sometimes observed the
predicted performance-enhancing effects of motivation, but more often failed to
"now
you see it,
do so (see Atkinson,1974, for a review). Atkinson attributed the
"overmotivation"
(and its
now you don't" character of this effect to what he called
"undermotivation").
(Yerkes
&
Referring to the Yerkes-Dodson law
opposite,
tendency
Dodson, 1908), he postulated that there should be an ideal motivational
for each individual task. Tasks differ with respect to the amount of motivation
that leads ro most efficient performance: Low levels of motivation are more
Chapter 2. Action Pbasesand Mind-Sets
55
appropriatefor some tasks,whereasmedium or high levelsare more appropriate
for others.Sincewhich tasksbelong to which categorywas unknown, Ätkinson
suggestedestablishingthis classificationempirically. Once it was known what
level of motivation is ideal for a given task, researcherswould finally be able to
developvalid predictionsof task performance,basedsolelyon their knowledgeof
the individual'smotivation.
Äs comparedto this empirical Sisyphus-likework, a more theoreticalsolution to the troublesomemotivation-perfcrmanceissuecertainly exists.That is,
one can re-establishthe old distinction betweengoal setting and goal striving.
Eric Kling er (7977) recognizedthis possibilitywhen he introducedhis conceptof
"current
concerns."He pointed out (1977, pp. 22-24, 329-)30) that expectancy X value theories have been only very modestly successfulin predicting vital
aspectsof goal striving, such as work effort and quality of performance.Consequently, his theory of current concernsfocusessolely on issuesof goal striving.
This theory has no difficulties in accountingfor the commonly observedinvigoration of acrivityin the faceof obstaclesen route to a chosengoal-a phenomenon
that cannot be explained by expectancy-valuetheories,becausethe setbackmust
be assumedto reducethe expectancyof achievingthe goal and thus the individual's motivation to work for it.
Kuhl (1983) also re-establishedthe classicdistinction by introducing the
"control
"choice
motivation." In his opinion, models
motivation" and
conceptsof
of choice motivation relate to goal setting, and he saw Atkinson's risk-taking
model and its many reformulationsand extensions(e.g.,Feather, 1967; Heckhausen,I9l7;Raynor, L969;\7einer,1974)as more or lessvalid examplesof such
models.Kuhl noted alackof theorieson goal striving and offered his own model,
"control
theory" (Kuhl, 1984). Stimulated by Atkinson and
which he labeled
Birch's (I970) assumptionthat a person is alwaysaffectedby numerousmotivational tendencies,all in constant flux, Kuhl saw effective goal striving as dependent on people'sefforts to shield it from competing action tendencies.Accordingly, whether people make progresswith respectto a chosengoal is no longer
seen as dependentoniy on the motivation that originally made them choosethis
people shield (control) the
goal. Rather, it is also a questionof how successfully
actions that lead to goal achievement.
The Rubicon Model of Action Phases
"Rubicon model" of action phases (Heckhausen,1987b; Heckhausen&
The
Gollwitzer, 7986, 1987) goes beyond the useful conceptualdistinction between
goal setting and goal striving. Although the model keeps these two problems of
both within a single theoretical
goal-orientedbehaviorseparate,it encompasses
model, thus permitting them to be analyzedin relation to each other. Furthermore, it provides a temporal perspectivethat begins with the awakeningof a
person's wishes prior to goal setting and continues through the evaluative
thoughts entertained after goal striving has ended.
56
and'Selt''Control
Part l. Self-Regulati.on
Separating the sequenceof events occurring within this comprehensive time
frame into discrete phenomena, the model posits four distinct phases: first, the
predecisional phase; second, the postdecisional but still Preactional phase; third,
ihe uctional phase; and iast, the postactional phase. These phases are seParatedby
three clear boundaries or transition points: the makin gof a decision,the initiation
of respective actions, and the conclusion of these actions. But what distinct
phenomena ^re associatedwith each phase?
PredecisionalAction Phase
People'smoby wishingand deliberating.
The first phaseis characterized
tives (McClelland,1930)producecertainwishes:For instance,a personwith a
srrong power motive and a weak affiliation motive is expectedto experiencemore
wishes related to power than to affiliation. However, people cannot act on all of
their wishes but must chooseamong them, becausesome wishes may contradict
each other, others are roo difficult to implement, and life is simply too short to
follow all of one's wishes. People have to deliberate over which of their many
wishesthey prefer to Pursue.
They may employ the criteria of
How can peopleestablishsuchpreferences?
feasibility and desirability. S7ith respect to feasibility, people may contemplate
whether they can obtain the outcomeimplied by a given wish through their own
activity and whether the situational context they face is facilitating or impeding.
Accordingly, they should also becomeconcernedwith questionssuch as whether
they will find enough time to strive for the desired outcomes and whether the
necessarymeans or oPPortunitieswill be available.
The desirability of the wanted outcome is determined by reflecting on its
expectedvalue. The expectedvalue is derived by estimating the pleasantnessorrpl.uruntness of potential short-term and long-term consequencesand by assessing the probability that achieving the desired outcome will lead to these
consequences.Such consequencesinclude the following: a positive or negative
self-evaluarion,a positive or negative evaluation by significant others, Progress
toward some important life goal, or some pleasant or unpleasant side effects
unrelated to the specific wish that initially started the person's striving (Heckwith the processof achievingthe
hausen,1977).In addition,incentivesassociated
desired ourcome (e.g., joy experienced while trying to establish the desired
outcome) should also be relevantwhen the desirability of a given wish is deliberated.
of the feasibilityand desirabilityof a given wish, however,
Proper assessment
with
requiresthat this wish be seenin relation to other wishes.A wish associated
many atractive consequencesmay suddenly appear less desirable when scrutinized in the light of a superordinatewish (e.g.,the wish to dine in fine restaurants
becomeslessdesirablewhen it conflictswith the wish to buy a house).Or it might
becomemore feasible when contemplatedin connection with the realization of
other wishes (e.g.,u busyperson'swish to learn to play tennis may aPPearmore
feasible when it is contemplated together with the wish to take an extended
vacation).
Chapter2. Action Phasesand Mind'-Sets
57
Makinga Decision and the PreactionalPhase
high desirabiiityand feasibilityand thus is
Evenwhen a wish is accorded
given highest preference,the model of action phasesassumesthat wish fulfillmenr further demandstransformingthe wish into an intention. Phenomenologically,thistransformationis characteizedas a resolutionresulting in a feeling of
that one will act
determinationto fulfill the wish (or at leasta feelingof assurance
on rhe wish at hand; Michotte & Prüm, 1910).The goal stateor desiredoutcome
specifiedby the wish thus becomesan end state that the individual feels committed to achieve.The model describesthis senseof obligation in stating that the
"goai
intention." To catch the flavor of this transition
individual has acquireda
from the fluid state of deliberation to a firm senseof commitment, Heckhausen
"crossing
the Rubicon."
(1987b)employedthe metaphor of
After forming a goal intention, people move to the preactional phase.The
phenomenon associatedwith this action phase is planning. Planning is often
necessarybecausenewly formed goal intentions cannot be implemented immediately if the individual is engagedin alternative activities that first need to be
completed or if relevant opportunities to act are not yet available.In addition,
most goal intentions specify goal states (e.g., to graduate from college) that
cannorbe achievedin a single step. Consequently,the individual is interrupted (or
must pause) repeatedly and is forced to await future opportunities to work
towardsthis goal.
The model of action phasesassumesthat people do not usethesetime breaks
of goal achievement;
or pausesto weigh the positive or negative consequences
rather, the feeling of obligation associatedwith the goal intention makespeople
cöncernedwith the issueof how to promote achievingthe chosen goal. Accordingly, they should addressquestionsof.wben anduhere to start acting,bow to act,
and bout long to act. STheneverpeople anticipate difficulties with respectto any
of these implementational issues,they should commit themselvesto one of the
many possibleways of initiating, executing,and terminating a relevant courseof
action.
Committing oneself to a particular implementational course constitutes
forming behavioralintentions. These behavioralintentions (i.e., initiation intentions, execution intentions, and termination intentions) focus on a person's
behaviorin pursuing the chosengoal. The model distinguishesbehavioralintentions from goal intentions, since the latter focus on desired goal states.In line
with the ideas of German will psychology (Ach, 1935), it is assumed that
behavioralintentions promote the smooth initiation, execution,and termination
of activitiesin pursuing a person'sgoal intentions.
Action lnitiation and the Actional Phase
When does a goal intention lead to initiating relevant actions?It primarily
depends on the goal intention's volitional strength-that is, how stron gly a
person is committed to implementing the chosengoal. The genuine amount of
volitional strength is consideredto be a positive function of the goal's desirability
and feasibility as perceivedprior to choosing this goal. However, this volitional
5B
Part I. Self-Regulationand Self-Control
strength may vary, depending on a person's experienceswith attempting to
initiate relevant actions.If a person repeatedlyignores good opportunities to
initiate relevantactions,volitional strength may decreaseover time. On the other
hand,volitional strength may spontaneouslyand momentarily increasewhen the
individual encountersobstacles.
More importantiy, goal intentions and their effects on the initiation of
without consideringthat many different goal
relevantactionscannotbe discussed
intentions may compete for implementation at any given point in time. One
the intention with the comparatively
would expectthat under thesecircumstances
highest volitional strength would prevail. However, the situation at hand may not
be equally conduciveto implementing all of these competing intentions; it may
favor implementing some of theseintentions more than others. In addition,for
someintentions the situationat hand may be better suitedfor smooth implementation than any future situation for which the individual hopes.Consequently,the
individual may be very eagerto take the opportunity at hand and to postpone the
implementationof competingintentions,even if these intentions are associated
with comparatively higher volitional strength.
To summarize: SThether a given goal intention leads to the initiation of
relevant actions depends on its volitional strength (as compared with that of
other competing goal intentions) and on how favorablethe situation is for readily
initiating the particular goal intention (as compared with initiating competing
goal intentions and as comparedwith relevant furure opportunities one hopes to
encounter).Finally, a goal intention that has been furnished with initiation
intentions during the postdecisional(preactional)phaseshould havean additional
advantageover competing goal intentions, given that the opportunity specifiedby
the initiation intention is present. In this case,the opportunity to act should be
more easily recognizedand, once recognized,should elicit a specialimpulse to
start acting on it.
Action initiation is the demarcationline signaling the transition to the
actionalphase.The phenomenoncharacteristic
of this phaseis actingtoward goal
achievement.A person'sefforts to pursuea goalintention arc againassumedto be
related to the goal intention's volitional strength. The amount of volitional
strength servesas a kind of threshold value for the individual's effort exertion.
This threshold, however, may be spontaneouslymoved upward if hindrancesare
encountered, thus allowing f.or a reactive, momentary increase in volitional
strength. Spontaneousnonconsciousincreasesin effort exertion were originaliy
reportedby German will psychologists(Ach, 1935;Hillgruber, \9I2), who interpreted these reactiveresponsesof the individual as attempts to hold on to one's
goal commitment. These ideas should not be confusedwith considerationsexpressedby models of effort calculation.For example, Brehm, \7right, Solomon,
Silka,and Greenberg(1983),Kukla (1972),and Meyer (1973) specifiedhow the
reflective appraisal of perceivedability, perceiveddifficulty, and subjectivevalue
of goal attainment determine a person's effort exertion.
Heckhausen(1987a) assumedthat the courseof action is directedby the
mental representation of the goal, and that determination to achieve a goal
Chaater2. Action Phasesand Mind-Sets
59
originates from the mental goal representation even when the goal itself is
outside of conscious awareness.The goal may be defined at various levels of
abstraction (i.e., at the lowest level to the intricacies of the actions to be executed,
at an intermediate level to the intended outcome, and at the highest level to the
conseqlrencesthat this outcome is expected to mediate), depending on the difficulties the person encounters when acting on it. In line with Vallacher and
\Tegner's (1987) action identification theory, goals are assumedto be defined on
low levels of abstraction (i.e., necessaryimplementational steps as compared to
the intended outcome and its desired consequences)when smooth goal pursuit is
thwarted.
Coal Achievementand the PostactionalPhase
The phenomenon associatedwith the final action phase is evaluating the
question of whether one's goal striving has succeeded.\X/hat criteria govern this
evaluation? Two successiveevaluative questions must be answered by the individual. First is the question of whether the intended outcome has been achieved,so
that the individual may stop acting and await the desired consequences. This
question is easily answered whenever the outcome is a discrete performance (e.g.,
to send a birthday gift to a friend). It becomes a problem difficult to solve and full
of uncertainty whenever the intended outcome can be continuously improved or
extended (e.g., to prepare well for a mathematics test). In the latter case, the
"I
will work through the
individual may resort to termination intentions (e.g.,
"I
practice examples twice" or
will stop when I succeed in solving every other
practice problem"), thus defining clear standards regarding when the intended
outcome is achieved.
Second, the individual must address the question of whether the actual value
of the goal striving matches its expected value. This implies that the individual
must wait for the desired consequences of the achieved outcome before this
question can be answered. Only then will the individual be in a position to
compare the actual value with the desired value, regardless of whether the desired
consequencesare a positive self-evaluation, positive evaluation by others, progress
toward some superordinate goal, or some pleasant side effects. In reality, the
actual value may not measure up to the expected value as assessed during
predecisional deliberation. The desirability of the goal may have been overestimated because certain negative consequenceswere neglected or underestimated,
whereas positive consequenceswere overestimated. Future predecisional deliberation should benefit from such evaluations; that is, the estimation of expected
values should become more accurate. In this sense, postactional individuals look
not only back into the past, but also to the future.
Postactional evaluation may not only benefit future deliberation, but may also
help a person's future planning. \Whenever the individual recognizes that the
achieved outcome does not meet the intended standards or that the achieved
outcome is not good enough to lead to the desired consequences,the individual
may furnish the goal intention with new initiation and execution intentions, thus
improving the chances of successful implementation. Or the person may lower
60
Part l. Self-Regalationand'Self'Control
the standards related to the quality of the outcome or the attractiveness of its
consequences.If such measuresare not taken or if all of these efforts fail, the goal
intention may linger on without successfulimplementation. \Thenever a situation
is encountered that could be perceived as conducive to implementing the goal
intention, it still should become activated, although chances to implement it are
rather slim. Since this acivation occupies cognitive capacity, it may even hinder
t|e implementarion of competing goal intentions, thus turning the individual into
a procrastinator who keeps failing to act on his or her intentions.
Summaryand Discussion
temporal(horiThe Rubiconmodelof actionphasestakesa comprehensive
zontal) perspectiveon the courseof action, and thus differs from most current
modelsof action.The latter are of a strict vertical,hierarchicalnature (e.g.,Carver
& Scheier,1981;Gallistel,1980,1985;FIacker,1985;Semmer& Frese,1985);they
assumethat the individual, when executing a courseof action, advancesfrom a
concern with abstract,superordinate,higher-level goals to concrete,subordinate,
lower-level goals. The temporal dimension of action is addressedsolely with
respect ,o ,h. organizationof single acts or action units within the course of
action (von Cranach,1982).
by the model of actionphaseshasso
The horizontalperspectiveassuggested
far not been very popular in psychology.The German philosopher and psychologist Christoph Sigwart (1889) introduced this perspectiveprior to the heydayof
will psychology.Although his work did not stimulate any systematicresearch,it at
le"si prevented German will psychologistsfrom confounding problems of goal
setring (which they referred to as problems of motivation) with goal striving
(which they referred to as problems of willing or volition). Recently,Heckhausen
and Kuhl (1985) employed a horizontal perspectivewhen they reflected on the
long way from a person's wishes to the execution of relevant actions.Although
their primary focuswas on the mental examinationsthat wishes must passbefore
winning accessto a person's behavior, they made a strong distinction between
pr.- unä postinrentional processes,which they also referred to as motivational
respectively.
and volitional processes,
The Rubicon model of action phasesincorporatesthis distinction; however,
insteadof focusingon a person'smental efforts (or blocks) in turning a wish into
relevant action, it artempts to delineate distinct phenomena of goal-oriented
behavior whose functioning obeys distinct principles. In temporal order, these
phenomena are deliberating, planning, acting, and evaluating. The Rubicon
model may lead ro a number of misconceptions if taken too literally. These
misconceptionsare as follows:
1. The model doesnot imply that every single initiation of action is directly
precededby deliberation of the desirability and feasibility of the underlying goal
and the forming of a goal intention. Many initiations of action are simply
resumprionsof activitiesthat were started some time before; forming the underlying goal intention anew is therefore unnecessary.The same is true for action
initiations postponed becauseof a lack of opportunities to act. Finally, people
Chapter2. Action Pbasesand,Mind-Sets
61
entertaingoal intentions that imply superordinate,identity-relatedgoals,suchas
becoming a psychologist.These identity intentions (Gollwitzer, 1987) lead to
initiating relevantactionswithout prior reflectionon the desirabilityand feasibility of the underlying goal; the individual needsonly to checkwhether a given
opportunity is conduciveto pursuing this goal.
2. The model doesnot imply that forming a goal intention is necessarily
followed by intense planning concerningwhere, when, how, and how long to
implement the chosengoal.It is rather assumedthat suchconcernsoriginateonly
when smooth implementationof the goal intention is threatened.Initiation may
or means are required that
be cumbersome(a) wheneverspecialcircumstances
still needto be developedor created;(b) wheneverthe critical opportunity may be
missedbecauseit is difficult to recognize,happensinfrequently, or presentsitself
only for a short moment; and (c) when competing goal intentions continue to
block implementing the critical goal intention. Execution is hampered when the
courseof action runs into difficulties becausethe individual does not possessthe
necessarycompetenciesor fails to focus attention on the goal pursuit when
consciouscontrol of the activity is needed.Finally, termination of the implementational activities becomesproblematic whenever it is unclear exactly what suffices as the intended outcome.In all of these casespertaining to the initiation,
execution,and termination of implementational actions,planning that results in
the formation of the respectivebehavioralintentions is to be expected.
3. The model of action phases doesnot exclude the possibility of overlap
betweenactionphases.In the predecisionalphase,deliberationof wishesconcerning a goal can easilybe interrupted so that actions in the serviceof other already
chosen goals may be planned, initiated, completed, or evaluated.Also, in the
postdecisional(preactional) phase,the individual may deliberatevarious wishes
and evaluatesome completed goal pursuit while waiting for the opportunity to
act on a chosen goal; the individual may even act on some other goal when
theseactionsdo not demandmuch cognitive capacity(i.e.,when they are automatized). Similarly, during the execution of goal-related actions, individuals may
deliberatewishes, ready themselvesfor implementing other goals, or evaluate
some terminated goal pursuit as long as executing the critical actions is largely
automatized.
4. The model of action phases doesnot ignore the fact that goal striving is
hierarchicallyorganized.This is most evident in the model's distinction between
goal intentions and behavioralintentions. Behavioralintentions are supplements
to goal intentions and serve to promote the implementation of goal intentions.
Accordingly, the formation of a goal intention precedesthe formation of behavioral intentions, and the latter are justified by the former. But not all of the
intentions formed subordinatelyto some goal intention must be behavioralintentions. People frequently form goal intentions in the serviceof other (superordinate) goal intentions (e.g.,when a personwho has decidedto becomea psychologist makesup his or her mind to go to schoolabroad).In this case,the formation
of the subordinate goal (i.e., going to school abroad) should be precededby a
concernnot only for the feasibility of this goal, but also for its desirability.
6Z
Partl. Self'Regulationand'Self'Control
describeforming
5. The modelusesrhe metaphorof crossingthe Rubiconto
point of no
a goalintention.The allusionis not so much to havinggone beyonda
that
..,orr. as it is to putting incessantdeliberationto a rest' The model assumes
"babble of competing inner voices" $ones &
making a goaldecisionsrops the
to the initiation
Gerurd, 1967,p.18i). After the decisionhas beenmade,but prior
goal is
of actions,.ro d"liberation of the pros and cons relative to the chosen
implemenexpectedto occur;rather,the individualis assumedto exploreefficient
,uiio' of the chosengoal (Beckmann& Gollwitzer, L987).
Still, the model *ro*", that making a goaldecisioncfeatesa rather durabie
do not
commitmenr to pufsue this goal, so that hindrancesto one'sgoal pursuit
to conquer
leadto immed.iateretfear.Rather, the individual is expectedto attemPt
increasingeffort, employingdifferent means'taking
hindrancesby spontaneously
taking
more time to overcomethese hindrances,or trying to get around them by
alternativefoutes to goal achievement(Gollwitzer & \Ticklund, 1985)'Obviously,
phasesis
the concept of co*m-itment employed by the Rubicon model of action
of
dissimilar to commirment notions that link commitment to the execution
action, as conceivedby dissonanceresearchers(Brehm & Cohen, 1962; \Ticklund
and Rusbult
& Brehm, 7976),u.ri ulso by Brockner and Rubin (1985), Farrell
(1981),Kiesler (1971),andSalancik(1977).since behavioris lessrevocablethan
thoughts $ones & Gerard, 1961),the latter concePtualizationfurnishes committhe action
.rr.n, with a point-of-no-fetufn quality. Contrary to this approach,
a goal' as
phasesmodel iorr..p,rrulizes commitment in terms of an obligation to
798ö;
portruy.d in researchon maintaining relationships(Kanter'1972; Kelley,
(Buchanan,
Lond.,1985;Rosenblatt, \977),on identificationwith an organization
on self1974;Mowday, Porter, & Steers,1982;O'Reilly & Chatman,1986),and
(Emmons'
defining goals (Wicklund & Gollwitzer, 1982) or personal strivings
1989).
the
At the core of the Rubicon model of action phasesis the assumPtionthat
(deliberating'
realm of goal-orientedbehavior comprises various phenomena
is it
planning, äctirrg,evaluating) that are ruled by different principles' But how
In
possibleto ,p..iiy theseprinciples so that one may test postulateddifferences?
an
the next section, I show that employing the concept of mind-set provides
interesting solution to this problem.
THECONCEPTOF MIND-SET
the Rubicon
If we assume that the phenomena associated with each phase of
the tasks to be
model are efforts at solving distinct tasks, we may try to specify
phase, the
solved at each of the four lhases of the model. In the predecisional
action goals,
pefson's task is to make the best possible choice between potential
promote the
irh.r.u, in the postdecisional (preactional) phase the task is to
the actional
initiation of actions that imply moving toward the chosen goal' In
the
phase the person faces the task of efficiently executing such actions, whereas
determine
to
task in the postactional phase may best be described as trying
Cbapter2. Action Pltase:and Mind-Sets
63
whether the intended outcome and its desiredconsequences
actually accrued.\7e
may further assumethat involvementin these taskscrearesa congruent"mindset"-that is, a phase-typicalcognitiveorientation that promorestask completion
(Gollwitzer,7990).This implies that analyzingthe task demandsof each action
phaseshouldleadto hypothesesaboutthe uniquequalitiesof the respectivemindset. Beforethis analysisis attempted,however,a historical review of the concepr
of mind-setis presented.
Historical Background
In 1904 Oskar Külpe, the founder of the \würzburg school (see Boring, 19i0,
pp. 401-406;Gibson, 1947;Humphrey,1951, pp. 30-131), reported his experiments on what he called"abstraction."Subjectsviewedpicturesof four nonsense
syllables,eachwritten in a differentcolor.The letterscomposingthe syllables,the
positioning of the colors,and syllablesthemselveswere varied over rrials. Most
importantly, Külpe also varied instructionsprior to each picture presentation.
Subjectshad to attend to a particular aspect of the stimulus display (e.g., the
frequencyof a certain letter, the positioning of the colors, the figure represented
by the positioningof the syllables,or the kind of letters composingthe syllables).
Immediatelyafter eachstimulus presentation,lasting 0.125second,he requested
the subjectsto report the solutionsto the tasks;in addition, he askedthem to
recall the other aspectsof the stimulus display, of which they had not been
instructedto take notice.The results showeddrastic effectsof instruction: $7henever the experimenter'squestionswere related to the instructions prior to viewing the display(e.g.,subjectswere askedto attend to the positioningof the colors
and then asked to recall it), subjectswere highly accurarein their answers;
however,wheneverthere was a mismatch (e.g.,subjectswere askedto attend to
the positioning of the colors but had to report on the different letters composing
the four different syllables),subjectswere exrremely inaccurate.
In a very similar experiment, Chapman (L932) observedcomparativelymore
accuratereports when the instructions given prior to srimulus presentation
matched the inquiry posed after stimulus presentation.ITatt (1905), another
representativeof the Sfürzburg school of thought, used particular words to talk
about sucheffects,speakingof the instructions prior to stimulus presentation as
constituting an Aufgabe (task), which createsin the individual who acceptsit a
corresPondingEinstellung (mind-set). This mind-set in turn should "prepare"
the individual so that the stimulus material presented should be analyzedefficiently, resulting in proper task completion.
Mind-Sets Related to Action Phases
In earlier papers stimulated by the Rubicon model of action phases,two distinct
mind-sets or states of mind were postulated (Gollwitzer, !987; Heckhausen,
1987b; Heckhausen& Gollwitzer, 1986, 1987). The predecisionaland postactional phaseswere seenas being similar, becausein both phasesthe desirability
64
and Self'Control
PartI. Self-Regulation
and feasibilityof a goal^re atissue.Becausethis issueis the mastertheme of the
modern psychologyof motivation (Ätkinson, 1964),we referred to theseaction
"motivational." Likewise,the postdeciphasesani the ^iiociuted mind-setsas
sio.rul(preactional)phaseand the actionalphasewere seenasbeing similar, since
in both phasesimplementing the chosengoal is at issue.Becausethis was the
mindsets
masrertheme of *ill psychology,theseactionphasesand the associated
"volitional." Moreover,the motivationalmind-setwas said to
were referredto as
"reality orien121i6n"-1hat is, an orientation
be characterizedby a so-called
unbiasedmanner. On
toward processingavailableinformation in a nonselective,
by a so-called
the othJr hand,the volitional mind-setwas saidto be characterized
"realization orientation"-that is, an orientation toward processingavailable
information in a selectivemanner biasedin favor of attaining the chosengoal'
This original conceptualization,however,createdconfusion.For instance,it
can be urgo.J (Kornadt, 1988) that not only predecisionalbut also postactional
individuais try to achieve(realize) something-namely, to make Pfoper decisions
or to developcorrect evaluations,respectively.Therefore, realization orientation
shouldalsobe presentin the latter actionphases.Similarly,people in the process
of planning or executing an intricate courseof action have to scrutinize available
situational information rather realistically,and thus should also evidencea strong
reality orientation.
ihis original conceptualizationalsofailed to exploit the theoreticalpower of
the concepro1mind-set. Applying the mind-set concePtto the action phases,first
of all, requirescritically aialyzingthe tasks individuals set for themselvesin the
various uition phases.Seco.rd,having discoveredthe characteristictask demands,
one is finally in a position to form hypotheses about the unique cognitive
orientation of the reipective mind-sets. Our original conceptualizationdiscouraged this approach beca.rseit lumped the predecisional and the postactional
piur., ,og.,h.r, banning the idea that different tasks are solved in these action
-phur.r.
Tte sume was done with the postdecisional(preactional) phase and the
-actional
phase.Moreover, the characterizationof the cognitive characteristicsof
the delineatedmind-setswas either extremely general (reality orientation) or did
not even relate to a cognitive orientation, but rather to the task expectedto elicit
the mind-set (realizationorientation).In the following discussion,our original
conceptualizationis abandoned.Instead, an analysisof the tasks to be solved at
each lction phase is presented. As a result, four distinct mind-sets (i.e., the
deliberativ., implemenlal, actional,and evaluativemind-sets) afe posnrlated,and
their distinctive cognitive orientations are spelled out'
Deli ber ative M i nd-Sets
\When reflecting on rhe task to be solvedin the predecisionalphase,one has
to keep in mind that predecisionalindividuals deliberatein order to determine
which of their wishes ur. ,ro, only most desirablebut also feasible' Solving this
task requiresthat an individual be primarily concernedwith information relevant
of a given wish's outto the positivity-negativity of the expectedconsequences
Chapter2. Action Pbasesand.Mind.-Sets
65
comein order to estimateits desirability.In addition,information that aliows the
individualto assessthe chancesof achievingthis outcomeseemscrucialin determining its feasibility. Reliable estimates should be favored when all of the
relevant information for assessingdesirabilityand feasibility is discoveredand
processed.Since it is unciear at the outset which pieces of information or
knowledge may be relevant to assessingdesirability and feasibility, a general
open-mindedness
toward processingincomingor storedinformation seemsbeneficiaL Finally, being concernedwith information that addresses(or potentially
addresses)
the desirabilityof the wish under scrutiny will not do much good if
deliberation is not conductedin an impartial manner. Ignoring negative consequencesor overemphasizing positive consequencesmay make the deliberated
wish appear more desirablethan is actuallyjustified. Similarly, if information
pertaining to the feasibility of the wish under scrutiny is not analyzedina manner
that favors accurate assessments,the individual may overestimate his or her
capabilitiesto implement the desiredwish, and thus may judge its feasibility to be
higher than it actually is.
Accordingly, the mind-set that clearly facilitatesthe task of the predecisional
phase (i.e.,to choosethe most desirablewish that is also feasible)should evidence
the following characteristics:First, there should be cognitive tuning toward
information relevant to the issuesof feasibility and desirability. Second,there
should be an orientation toward accurateand impartial processingof such information. And finally, there should be an open-mindednessor heightenedreceptiviry to information in general. This deliberativemind-set should originate whenever people becomeintensely involved with deliberating their wishes.
lmplemental Mind-Sets
The task to be solved by the postdecisional(preactional) individual is planning when, where, and how to act in order to promore action initiation. Solving
this task effectively requires the individual to be primarily concernedwith information related to these questions.Moreover, task solution is facilitated whenever
the individual commits himself or herself to a certain favorable opporrunity to
act-that is, forms an initiation intention. In this way attention is focused on a
specifiedopportunity to act, and the probability that the individual will forego
this opportunity is reduced. Flowever, all.of these concerns will fail to benefit
action initiation if the individual starts to questionthe desirability or feasibiliry of
the chosengoal. Accordingly, any suchdoubts should be counteredby the individual's boosting the desirability and feasibility of the chosengoal, thus maintaining
persistencein initiating actions ro reach the chosengoal.
Therefore, the mind-set that facilitatessolving the task of the posrdecisional
(preactional) phase should evidence the following characteristics:First, there
should be cognitive runing toward information relevant to when, where, and how
to act-Second,there should be closed-mindedness
in the senseof concentraringon
information that helps to promote the chosengoal. And finally, there should be a
partial and optimistic analysisof information related to the chosengoal's desira-
66
and.Self-Co.ntrol
Partl. Self-Regulation
bility and feasibility,respectively.This implemental mind-set should originate
wheneverpeoplebecomeintenselyinvolvedwith planning the implementationof
their goal intentions.
ActionalMind-Sets
The task of the actionalphasemay be describedas acting toward the goal so
that goal achievement is promoted. Solving this task requires one to avoid
disruftions, becauseany halting of the flow of action postponesgoal achievementof what
The mind-setrhat facilitatesthis shouldthereforeevidencecharacteristics
"flow experience"and Wicklund (1986) labeled
Csikszentmihalyi(L975) called
"dynamic orientation." The individual no longer reflectson the qualitiesof the
goul ,o be achieved,on his or her capacitiesto achievethe goal, or on alternative
itrategies on goal achievement;nor does the individual form behavioral intentions iegarding when, where, and how to act.Rather, the individual is completely
caught up in the actionscurrently being executed.Accordingly, only those aspects
of Ä. self and the environmentthat sustainthe courseof action are attendedto,
whereasany potentially disruptive aspects(e.g.,self-reflectivethoughts, comPeting goal intentions,disiractiveenvironmentalstimuli) are ignored.Therefore,the
mind-set that facilirates rhe promotion of goal achievement is one of closedmindednessto informarion that could trigger a re-evaluationof the goal that is
pursued,a re-evaluationof the chosenroute toward goal attainment, or any self"Can I be proud of my performance?Am I suited for this
evaluation (e.g.,
activity?"). Raiher, the actionalmind-set should evidencecognitive tuning toward
internal and external cuesthat guide the courseof action toward goal attainment.
It should originate whenever people move effectively toward goal attainment.
Evaluative Mind-Sets
The task to be solved in the postactionalphase is evaluating outcomesand
of goal striving in order to discoverwhether the intended outcome
consequences
have been obtained. Solving this
has been reacheJand its desired consequences
task, therefore, requiresthe individual to be primarily concernedwith the quality
Task
of the outcome (siandards)and the actual desirability of its consequences.
solution should be facilitated when the individual simply compares what is
achieved(outcomes) and obtained (consequences)with what was intended and
desired when the goal intention was formed. This comparison should benefit
from a correcr assessmentof the quality of the outcome and an objective,impartial view of the desirability of its consequences.
Accordingly, the mind-set that facilitatesthe task of evaluationin the postactional phase should evidencethe following characteristics:First, there should be
cognitive tuning toward information relevant to assessingthe quality of the
Second,there shouldbe
achievedourcomeand the desirabilityof its consequences.
an orientation toward accurateand impartial processingof this information- And
finaily, there should be a comparativeorientation; that is, the intended outcome
should be comparedwith the actualoutcomeand its
and the desiredconsequences
consequences.This evaluative mind-set should originate whenever people be-
Cbapter2. Action Phasesand M.ind-Sets
67
come intensely involved with evaluatingourcomesand consequences
of goal
striving.
S u mma ry
The various action phasesof the Rubicon model differentiatefour distinct
phenomenaof goal-orientedbehavior:deliberating,plannin g, acting,and evaluating. Sinceeachof thesephenomenaimpiies solving a distinct task,it is inferred
that different mind-setsevolvewheneverone is involved in theseparticular tasks.
For each of these distinct mind-sets (i.e., deliberative,implemental, actional,
evaluative),the associatedcognitive orientation is specifiedby analyzingconcrere
task demands.
MIND-SETS
AND COGNITIVE
FUNCTIONING:
RECENT
FINDINGS
So far, we have concentrated our empirical efforts on testing the cognitive
orientations postulated for the deliberativeand implemental mind-sets. Our
experimentshavefocusedon three key issues:(1) the postulatedcognitive-tuning
effects; (2) the distinct way of processinginformation related to feasibility and
desirability;and (3) the postulateddifferencesin open-mindedness.
Mind-Setsand CognitiveTuning
S7e hypothesized that both the deliberative and the implemental mind-sets
achievecognitive tuning toward task-congruousinformation. The deliberative
mind-set should lead to cognitive tuning toward information related to the
feasibility of the intended outcome (action-outcome expectancy) and to the
desirability of the expectedconsequences(expectedvalue), whereas the implemental mind-set should evoke cognitive tuning toward information related to
actioninitiation (when, where,and how to ger started).How doesone test these
hypotheses?In principle, there are two possible approaches.The first approach
focuseson the subjects'thoughts while they are in a deliberativeor implemental
mind-set; it is expectedthat congruousthoughts will be experiencedmore frequently than incongruousthoughts.The secondapproachfocuseson the subjects'
readinessto encodeor retrieve task-congruousinformation; here it is expected
that congruous information will be encodedand retrieved comparatively more
effectively.
Thought Sampling
The first approachled us to conductthree different but related experiments.
In the first experiment (Heckhausen& Gollwitzer, L987,Study l), experimental
subjectschose between two different sets of test material designed ro measure
creativity,one set consistingof black-and-whitepictures and the orher set consisting of color pictures.Subjectswere told that they would later write a creativestory
68
and,Self-Control
Partl. Self-Regulation
about whichever set of pictures they chose.The reasongiven for offering a choice
was rhat somepeoplemore easilyreachtheir full creativepotential when working
with color pictures,whereasothersperform better with black-and-whitepictures.
Subjectswere encouragedto avoid snap choicesbetween test materials,but to
engageinsteadin full-fledgeddeliberation.For this Purpose,we had subjectsview
a seriesof samplepictures (six black-and-whiteand six color pictures).
S7einterrupted one group of subjects90 secondsafter they had viewed the
samplepictures(the predecisionalgroup), while still undecidedas to which set of
test material they wanted to choose.In a thought-sampling task, they were
requestedro reporr their thoughts experiencedduring the 90-secondtime period,
starring with the last sample picture up to the point of interruption. To increase
the validity of this self-report,we proceededas follows: Subjectsfirst were to
write down their most recent thought, and then the secondmost recent thought.
Next, the first thought was to be reported, and then everything they had thought
of in between.According to Ericssonand Simon (1980), this procedureshould
produce valid self-reports becausesubjectscan retrieve the most recent thought
from short-term memory. \Triting down the most recent thought should then
facilitate recall of the thoughts experienced immediately before, since these
should be connectedthrough episodicassociationsin long-term memory.
Another group of subjectswas not interrupted until a decision had been
made and subjectshad spent 90 secondswaiting for the test material they had
chosen (the postdecisionalgroup). These postdecisionalsubiectsreported their
thoughts experienced during the 90-second time period starting with having
made a decisionup to the point of interruption.
Finally, we employed two groups of control subjects.Both groups were
simply assigneda set of test material (yoked with the pre- and postdecisional
subjects'choices).One of these groups was interrupted before this assignment
(the preassignmentgroup) and the other group afterwards (the postassignment
group). The time periods on which control subjectshad to focus when reporting
their thoughts were closely matched to those of the respective experimental
subjects.
Content analysisof subjects'reported thoughts revealedthat thoughts re"Being
a cfeative Person is important") and
lated to expected values (e.g.,
"I
thoughts related ro action-outcomeexpectancies(e.g., would do better with the
color material") were much more frequent in predecisionalsubjectsthan in any of
the other three groups of subjects.The frequency of implementation-related
thoughts (i.e., considerationsand expressedintentions about what kind of story
would be told and how this should be done) was elevatedin both the postdecisional and postassignmentgroups. However, hardly any implemental thoughts
were reported by predecisionaland preassignmentsubjects.
If it is assumedthat predecisionalsubjectsengagedin intensive deliberation
and thus developeda deliberativemind-set, the observedpredominanceof taskcongruousthoughts in predecisionalsubjectssupports our cognitive-tuning hypothesis.Similarly, becauseboth postdecisionaland postassignmentsubjectswere
planning to write a crearivestory on the test material, both of these groups of
Cbapter2. Action Phasesand Mind-Sets
69
subjectsshould have developedan implemental mind-set. Observingcomparativeiy more implemental thoughts in these two groups again suPPortsour hycognitivetuning.
pothesisof task-congruous
Distance from Making a ChangeDecision
One could argue that the findings describedaboveare rather trivial, because
subjectssimply entertainedthose thoughtsthat they were told to entertain.One
has to remember,however,that we did not tell predecisionalsubjectsto think of
issuesrelatedto action-outcomeexpectanciesand expectedvalues;nor did we tell
postdecisionalsubjects to stop thinking about such issues and to turn their
attention to implementationalissuesinstead.A still more convincingtest of the
cognitive-tuning hypothesismay be performed if the independentand dependent
variables are exchanged-that is, if one asks people who harbor a Personal,
unresolvedproblem pending a changedecisionto engagein exactly those mental
activities hypothesized to be associatedwith a deliberative mind-set. If our
hypothesisof mind-set-congruouscognitive tuning is correct,one should exPect
thesesubjectsto becomelost in deliberation and therefore to feel predecisionalthat is, far from making a changedecision.But if suchpeople are askedto engage
in mental activities that we believe to be associatedwith an implemental mindset, they should becomeintensely involved in the postdecisionaltask of planning
the implemenration of the change decision not yet made. Consequently,they
should feel postdecisionaland thus closer to making a changedecision.
To test these ideas, the following experiment was conducted (Gollwitzer,
Heckhausen,& Ratalczak,1990). Subjectsfirst named an unresolved personal
problem that was pending a change decision. They named problems such as
"should move away from home?" "should I switch my major?" or "Should I
I
break up with my boyfriend?" In order to measuresubjects'perceiveddistance
from making a changedecision,we askedsubjectsat the outsetof the experiment
how determined they felt at that very moment, how much resolution it would still
take them to arrive at a changedecision,and how far awaythey felt from the act of
making a change decision.Next, one group of subjects(the deliberation group)
was askedto estimare the expectedvalue of making the change decision.They
and the delayedpositive and negative
listed the potential immediateconsequences
and they estimated the probability of these consequences'occurconsequences,
the chancesof achievingthe resPectiveoutcomes.
ring. In addition, they assessed
Another group of subjects (the implementation group) was asked to plan the
implemenration of the not-yet-made change decision by listing a number of
difierent activities that could servethe purpose of implementing it. Subjectsthen
had to decideon a courseof action and to imagine themselvesexecutingthis planS7hen subjects were asked again about their perceived distance from a
changedecision (employing the three items listed above), deliberation subjects
continued to describe themselvesas undetermined and irresolute-that is, far
from making a changedecision.Implementation subjects,however, indicated an
increasein determination and resolution-that is, perceivedthemselvesas closer
to making a changedecision.An analysisof the potential mediators of this effect
70
Part I. Self-Regulationand Self-Control
but instead pointed to having
ruled our an increase in desirability or feasibility,
plan as the critical variable' In
committed oneself to a certain implementational
strongly contributed to apother words, forming behavioral intentions most
proaching the change decision'
? r
1r r:.r-^-^
far, I would like to point out
In summa rizirJgthe two studies reported thus
more thoughts rethe following: Pred-ecisionalsubjects showed comParatively
values, and when (undecided)
lated to action-outcome exPectanciesand expected
they felt strongly predecisional'
people were made to entertain such thoogÄts,
comParatively more implePostdecisional subjects, on the other hand, showed
were made to entertain such
mental thoughts, and when (undecided) people
to the act of decision' These
thoughts tn.! i.tt less predecisional-thatls, closer
produces cognitive tuning toward
findings suggest that a deliberative mind-set
and expected values' whereas the
thoughts related to action-outcome expectancies
how to achieve a chosen goal'
implemental mind-set tunes o.r. ,o*ird issuesof
Writing FairYlales
and implemental
The most conoi.rcing demonstration that delibefative
by the findings of the
mind-setsentail congfuous"thoughtproduction is provided
L989,Study 1)' First'
foliowing e"perimeit (Gollwitzer, Heckhausen'& Steller,
implemental mind-set by
subjects1"r" placed either into a deliberativeor an
problem pending a change
being asked to deliber ate an unresolved personal
resPectively'Again,
decisionor ro plan a personalproject pending realization,
projects of their choice;
subjectswere ilo*ed to *ork o., p.rrorral problems or
with apIifestyle-related,unä interpersonalissueswefe named
career-related,
grouP receivedthe same
proximately equal fäqo.r.y. The deliberativemind-set
experiinstructions as describedabove in the distance-from-a-change-decision
instructed to list five
ment. subjectsin the implemental mind-set grouP were
proiect they had
implementational stePs required to complete the intended
themselvesas to when'
named.For each of thesestePs,subjectshud to commit
did not receiveany
where, and how they plann.d ,o executeit. The control grouP
mind-set maniPulation.
presented with the
In the secondpart of the experiment, subjectswere
askedto continue these tales
beginnings of threeiiff.r.rr, fairy tales and were
point where the main
with three sentenceseach.All oi these tales ended at a
story' for example' a
character of the story faced a goal decision. In the first
at home to Protect his
widowed king faced tlre choiceof going to war or staying
of whether deliberative
beloveddaughter.The senter.., *-.r. scotedon the basis
verbs.relating to the
or implementational efforts were imputed to the king: Äny
in the predeciking were classifiedaccordingto whethef the king was engaged
task of implementing
goalsor the postdecisional
sionaltask of choosingbetween
"The king ,u.k"d Äis brain wondering what to do"' for
a chosengoal. The phrase
king, whereas the
instance,*u, ,.orä as imputittg u a.[berative effort to the
'iTh. king ordered a trusted officer to stay at home at the castle and
phrase
^p.o,..,
effort'
his daugliter" was scoredas imputing an implementational
Chafrter2. Action Phasesand Mind-Sets
71
Äs expected,subjects'mind-sets affected their flow of creative thought when
completing their fairy tales.Deliberative mind-set subjectsimputed more deliberative efforts to the king than implemental mind-set subjects, with control subjects' imputations faliing between those of the other two groups. An analogous
congruency effect was observed with imputing implementational efforts. The
implemental mind-set group scored higher than the deliberative mind-set group,
and the control group again scored in between these rwo groups.
Telling fairy tales follows a certain story grammar (Rabkin t979; Rumelhart,
1975, l9l7): Only when a solution is found to the problem introduced at the
beginning may the story come to an end. Since these solutions come about more
easily if the main character takes action, ascribing implementational efforts to the
king should have been the more common response. This was actually the case in
the controi group: Subjects imputed about 10 times as many implementational
efforts as deliberative efforts to the king. However, although telling a f.airy tale
strongly favors producing implementational thoughts, the deliberative mind-set
weakened this preorientation and the implemental mind-set strengthened it. It
seems important to note that these mind-sets had been elicited when subjects
meditated on quite different issues (i.e., unresolved personal problems or intended projects related to career,lifestyle, or interpersonal issues),and that some
minutes had passed before subjects proceeded with the fairy tales.
Cued Reca/l of Mind-Set-Congruous lnformation
Deliberative and implemental mind-sets not only should make congnrous
thoughts more readily available,but also should allow for more effective processing of congruous information. That is, people operating within a deliberative
mind-set should be particularly effective in processing information related to
outcome expectancy and expected vaiue, whereas people operating within an
implemental mind-set should be more adept at processinginformation related to
when, where, and how to act on a chosengoal. Demonstratingpotent mind-sets
implies, in addition, that this prediction should hold true not only for information
relevant to the meditated unresolvedproblem that has led to the deliberative
mind-set, or for information relevant to the planned project that has led to the
implemental mind-set. Rather, these mind-sets' cognitive-tuning effects should
transfer to unrelated deliberativeor implementational information.
To explore this supposition, we asked subjects to view a series of slides
(Gollwitzer et al., 1989,Study 2). Each of the eight slides depicteda different
person said to be experiencinga personalconflict of the following kind: Should I
do X or not? The accompanyingslide specifiedthis conflict (e.g.,"should I sell my
apartment or not?") and presented thoughts presumably entertained by the
person depicted.Two of thesethoughts were related to the expectedvaluesof the
changedecision("It would be good because. ."; "Ir would be bad because. ."),
and two were related to the questionof how to initiate relevant actions,given that
the changedecision had been made. One of these latter two was related to the
timing of relevant actions ( "If I should decide to do it, then I won't . . . before
72
PartI. Self-Regulationand Self-Control
. . ."), and the other to rhe sequencingof reievantactions("If I should decideto
d o i t , t h e nI w i l l f i r s t . . . a n d t h e n . - . " ) .
S7ecreateddeliberativeand implementai mind-setsby employing the choice
paradigmintroducedabove(i.e.,the first studyreported).The choiceofferedwas
b.,*..r, materials for a creativity test that required constructingcollagesfrom
material cur out of newspapers.Two sets of collagesegmentswere said to be
available,one set consistingof black-and-whiteelementsand the other of color
elements.Subjectswere told that people could reachtheir full creativepotential
only if they chose rhe type of material (black-and-white or color) they found
personally most appealing.
One group of subjectsviewed the slides and had to recall the information
depictedon the sidesprior to making a decision.Another group receivedand
recalledthe information af.tera decision had been made. If mind-set-congruous
information is processedmore effectively, predecisional subjects should have
recalledinformation relatedto expectedvaluesbetter than implemental information, whereas rhe reverse should have been true for postdecisional subjects.
Finally, we employed a group of control subjectswho receivedand recalled the
same information without either expecting to make a decision or having made
one.
performance(the beginningsof the sentencesas
Control subjects'cued-recall
listed abovewere provided as recall cues)was the samefor expected-value-related
information (positive and negative consequences)and the implementationrelated information (timing and sequencingof relevant actions).Predecisional
information than with
subjects,however, did better with expected-value-related
implementation-related information, whereas the reversewas true for postdecisional subjects.This pattern of recall performance strongly suggeststhat mindser-congruousinformarion is processedmore effectively than incongruousinformation. The present study should not be confusedwith experiments designedto
explore whether there is differential recall of information that is consistentwith
or contradictoryto rhe decisionmade (Dellarosa& Bourne,7984).In the present
study, the informarion provided was not even relevant to the choice to be made,
neither supporting nor undermining subjects'decisions.
S7hat kind of memory processesaccountfor the Present finding that mindser-congruousinformation is recalled comparatively more effectively? If we assume that subjects' retrieval attempts necessitateconstructing descriptions of
what they are trying to retrieve (Bobrow & Norman,l97l; Norman & Bobrow,
I976, 1979), it seemspossible that mind-sets provide perspectives(Bobrow &
\Tinograd,1977) that allow for the easyconstructionof specificdescriptions.The
deliberativemind-set, for insrance,should favor descriptionsphrased as pros and
cons,benefits and cosrs,hopes and fears, all tied to the specific conflicts of the
depictedcharactersgiven as retrieval cues.In other words, the deliberativemindset makes for the ready construction of descriptionsthat specify expected-valuereiated information, whereasthe implemental mind-set helps constructingdescriptions that specify implementation-relatedinformation. As Norman and
Bobrow (1979) point our, quick consrructionof specificdescriptionsat the time of
Chapter2. Action Phatesand Mind'-Sets
73
retrieval further successfulretrieval. Norman and Bobrow also assume that
wheneverthe descriptionof the information sought matchesthe elaborationof
this information at the time of encoding,recall performance is particularly
enhanced.It seemspossible,then, that deliberativeand implemental mind-sets
favor congruousrecall via congruouselaboration at the time of encoding and via
the ready construction of congruousdescriptionsat the time of retrieval.
Summary
The results of the four experiments reported show that deliberative and
implemental mind-sets tune people's cognitive functioning so that congruous
and congruousinformation is processedeffecthoughts becomereadily accessible
tively. Most interestingly,both mind-setspossesssome stability over time and
generalizeacrosssituations (the third and fourth studies reported).
Mind-Sets and Biased Inferences
The studiespresentedaboveprimarily addressthe questionsof what types of
thoughts or information are congruouswith the deliberative and implemental
mind-sets,whether congruousthoughts are more pervasive,and whether congruous information is processedmore effectively.However, both mind-sets can also
be assumedto differentially affect the way in which congruous or incongruous
information is handled. S7ehypothesizedthat information related to feasibility
and to desirability arc analyzedin a distinct manner. Whereas in a deliberative
impartially, assessment
mind-set information related to desirabiiityis assessed
mind-set.
Älso, feasibility
partialto the chosengoal is expectedin an implemental
is expectedto be assessedrather accuratelyin a deliberative mind-set, whereas
optimistic assessmentsthat overestimate the actual feasibility of the intended
outcome are expectedin an implemental mind-set. !7e conductedtwo studieson
eachof these issues.The first two studieswere thought-sampling studies related
to predecisionalindividuals' assessmentof desirability. The second two studies
used the illusion-of-control paradigm (Alloy & Abramson, 1982) and related to
assessingfeasibility when a deliberativeor implemental mind-set has been
createdexperimentally.
The Cou nterplea Heuristic
Is there an impartial analysisof expectedvalue in individuals with a deliberative mind-set? To answer this question,we (Gollwitzer & Heckhausen,1987,
Study 2) askedfemale university studentsto name an unresolvedpersonal problem for which they wished resolution but that for some reason they had not
"
" "Should
"Should
I switch my major?
resolvedyet (e.g.,
I move from home?
"Should
I studyabroad?").Then we askedsubjectsto achieveclarity with respect
to whether they wanted to make a changedecision;we expectedthis instruction
to trigger intensive deliberation and to create a deliberative mind-set. \7e also
askedsubjectsto report back to the experimenterwhen they felt that further
mentation would not achievegreater clarity.
74
and Self-Control
Partl. Self-Regulation
Subjectswere rhen given the thought-samplingquestionnaire(described
above),becausesubjects'answersto this questionnaireallowed us to study the
temporal order of the flow of consciousthought. $7e scoredsubjects'thoughts
accordingto a coding schemethat differentiatedbetweenpositive and negative
of having achievedthe desiredoutcome as implied by a change
consequences
were pondered
decision.First, v/e noticedthat positiveand negativeconsequences
with equalfrequency.More interestingly,when we put the thoughtsreportedinto
the correcr temporal order, we discoveredthat deliberationfoliowed a certain
partern. Deliberation started with reflection on the positive incentivesof wish
fulfillment (e.g.,having moved from home). However, subjectsdid not indulge in
but immediately turned toward reflecting on negathese positive consequences,
tive incentives.It seemedthat thesedeliberatingpeople playedtheir own devil's
advocate:Their initial enthusiasmaboutthe positiveaspectswas temperedwith a
counterpleapointing to negativeconsequences.
S7etried to replicatethis observedpattern of meditating on positive and
with a different sampleof subjects-that is, male students
negativeconsequences
at a military academy(Gollwitzer &. Heckhausen, 1987, Study 2). Älthough these
"Should
I acquire a
students named rarher different unresolvedproblems (e.g.,
" "Should
" "Should
I learn how to
I buy new skiing equipment?
flying license?
hang-glid€?"), the same temporal pattern of thinking about the positive and
negarive consequencesemerged. Obviously, people who become intensely involved with deliberatingan unresolvedpersonalproblem attemPt an impartial
are
Even though the positive consequences
analysisof potential consequences.
most salient at the beginning, negativeprospectsare quickly compiled in order to
with potential negativeconsequences.
contrast the desiredpositive consequences
are pitted against those that are feared; therefore,
Thus, desiredconsequences
there is no partial analysisfocusing exclusivelyor primarily on desiredconsequences.
The issueof impartial information processingprior to making a decisionis
nor new. Festinger (1964) reported a number of studies that addressedthe
cognitive functioning of predecisionalsubjects.All of these studiesassumedthat
an impartial analysiswould not affect the initial divergence of attractiveness
between choice alternatives(Davidson & Kiesler, 1964;Jecker, 7964; Walster &
Festinger,1964). Other studiesdefined impartiality as an equal amount of attention paid to eachof the choicealternatives,as measuredby looking time (Gerard,
1967),listeningtime (Brockg Balloun,1967),or more recentlyin terms of recall
performance (Beckmann& Gollwitzer, 1987).
Flowever, these different approachesto predecisional impartiality do not
capture the individual's analysisof decision-relevantinformation. An impartial
analysisdoesnot necessarilyleavethe divergenceof attractivenessbetweenchoice
alternativesuntouched,and a partial analysismay be conductedeven when the
amenrionpaid to choicealternatives(as measuredin terms of encodingtime or
recall performance) is about equal. Not surprisingly, then, conflicting findings
have been reported by studiesthat adhereto these operationalizationsof impartiality (e.g.,Janis& Mann, 1968;Mann,Janis,& Chaplin,1969). Therefore,the
Chapter2. Action Phasesand Mind-Sets
75
somewhat old-fashioned thought-sampling technique we employed may actually
provide the most valid data on the question of whether predecisional deliberation
is impartial, becauseit captures subjects' actual thought processes.
\Wehave not yet used our thought-sampling technique to test the hypothesis
that the implemental mind-set leads to a partial analysis of expected-value-related
information. But rather unambiguous support for such partial information processing after a decision is available elsewhere (Wicklund & Brehm, 1976). Researchers in the tradition of dissonance theory observed that postdecisional
subjects increased the attractivenessof the chosen alternative and decreasedthe
attractivenessof the nonchosen alternative (Brehm, 1956).In addition, postdecisional individuals were found to selectively seek information that potentially
supported their choice and to actively avoid nonsupportive information (Frey,
1986). These findings have been interpreted either in terms of justifying one's
decision in order to fulfiil a need for consistency (Insko, ITorchel, Folger, &
Kutkus, 1975) or in terms of justifying one's decision for the purpose of arriving
"unequivocal
action orientation" that precludes further deliberation of the
at an
choice alternatives $ones & Gerard, 1967; l7icklund & Frey, 198i).
This latter interpretation of postdecisional partiality is similar to our perspecrive that an implemental mind-set yields a pafüal analysis of expected-valuerelated information in order to promote immediate and persistent implementation of the chosen goal. The minor difference between the two perspectives may
be the following: The mind-set perspective suggests that doubts about the actual
desirability of the chosen goal arc first of all avoided by concentrating on the
implementation of the chosen goal. Only when this fails are postdecisional
individuals assumed to resort to justifying their choices by increasing their
expected desirability.
Illusionof Control
How do the deliberative and implemental mind-sets affect the analysisof
information related to the issue of feasibility?Our hypothesis was that the
deliberative mind-set should spawn ail acclrrateassessmentof the probability
of achieving a certain outcome, whereas the implemental mind-set should
\7e tested these hypothesesin two
lead to inaccurate,optimistic assessments.
experiments employing an illusion-of-control paradigm (Gollwitzer & Kinney,
1989).
In the contingency learning task designedby Alloy and Abramson (1919),
subjectsperform numerous trials on a single-stimulus apparatus. In this task,
subjectsare askedto determine to what degreethey can influence the onset of a
target light (outcome) by choosing to press or not to press a button (alternative
actions).By observingwhether or not the target light turns on, subjectsestimate
how much influence or control they have over the target light onset. The
experimenter can vary the actualdegreeof control by manipulating the frequency
of light onset associatedwith eachof the two action alternatives (pressingor not
pressing). The smaller the difference between these two frequencies,the less
objective control subjectshave over target light onset.
76
and Selt''Control
PartI. Selt''Regulati'on
review, see Alloy &
Än extensivebody of research(for a comPfehensive
Ä b r a m s o n , 1 9 8 8 ) h a s r e v e a l e d t h a t n o n d e p r e s s e d i n d i v i d uactions'
a l s c l a iwhenmtoPossess
subiects'
on
,ron.ontingent
ur"
that
control over desiredoutcomes
(e.g.,in a75-1.5 problem, where the target
ever theseourcomesoccurfrequently
responses)as
light comeson rn 75% of fressing ;d 15!.of .nonpressing
problem)' \7e hypothesizedthat the
comparedto infrequently (e.g.,in uiS-ZS
(action-outcomeexpectancy)found
inaccurate,optimistic iudgmentsof control
problems should be attenuated in
with noncontingent but frequent outcome
in implemental
be aggravated
deliberativemind-setsubjects,and that they sirould
mind-set subiects.
1989, Study 1), we modified the
In our first study (Gollwitzer & Kinney,
a secondaPParatusand by asking
Älloy and Äbramron ir9z9) paradigm by adding
."u-tt a deliberativemind-set'we told
subjectsto work on five ,.., of 20 tiials.'To
of the experimentwas to decide(after
subjectsthat their obiectivein the first part
two availableapparatusesthey wanted
completing the firre ,.ts; on which ol irr.
To allow for an informed
to work during the second part of the experiment'
between the two aPparatuses'\7e
decision,subjectswere "r,.oo-rug.dto alternate
by askingsubjectsto decideon the
attemptedro createan implemental mind-set
*ith 'espect to all five setsof trials
sequenceof alternation between aPParatuses
tfy to produce as many light onsets as
before starting the first set, and then to
possible.
T w o p r o b l e m c o n d i t i o n s w e r e e s t a b l i s h e d ' , a 7 5 _ T 5 p r o b l efrequent
manda2'.2'
noncontingent
pfesented-either
aPparatuses
both
problem. Äccordingly,
light' \ühen target light onset
or noncontingent infrequent onset of tü. target
tittd-tet subiectsreportedinacwas frequent (the 7t-75 problem), implemenäl
whereas deliberative mind-set
curately high, illusionary iudgments äf .on,.ol,
Apparently, the deliberativemind-set
subjectsshowedmodest.orr,rol iudgments.
optimism; that is' they recogprevents people from being lured into illusionary
is not necessarilya valid indicator of
nize that high frequencyof a desiredoutcome
\(ith respectto the 2t-25 problem'
one's degreeof lnfloer,ceover the outcome'
judgments' This indicates that
both mind-set groups showed modest control
of reality' They do not
implemen.* *äa-r.t subiectsdo adapt to the constraints
over target outcomes' but rather
blindly perpetuare an impression of control
many failures are encountered'
acceprthat they have little control when too
grounds' First' one could
This experiment may be criticized on two different
may havebeen reluctant to tell the
afgue thar the implemental mind-set subiects
alternation between aPPafatuses;
experiment., thu-t they made a bad choice of
judgments simply to keep up a
consequently,.i.y *uy huo. inflated their control
reported an illusion-of,BarrySchwartz (1983) "tt*ly
good impression-'secood
on a button-pushing task that
control experiment where he pretrained subiects
ta'k, he either establisheda rule
offered only two buttons. In ihe pretrainin!
which sequeocesof button
discovery orientation (subjects nua to determine
problems) or a reward
pressing led to reinforcement in three different sequence
in any order' on eachof the
orientation (subjectswere reinforced for four pushes,
illusion of control than reward
two buttons). Rule discoverysubiectsshowed-less
Chapter2. Action Pbasesand Mind-Sets
77
subjectsin a subsequentÄlloy and Abramson type of contingencylearning task.
One could argue that our deliberativemind-set manipulation was acruallya rule
discoverymanipulation,sincesubjectswere instructedto determine on which of
the two apparatusesthey would do better, whereasthe implemental mind-set
manipulation was a reward orientation manipulation, since subjectswere instructedto produceas many rarget light onsetsas possible.
Both problems can easilybe avoided,however,when a mind-set manipulation is employed,as describedin the fairy tale experiment above.Accordingly, we
ran a secondstudy (Gollwitzer & Kinney,1989, Study2) where one-third of the
subjectsmeditatedon an unresolvedpersonalproblem requiring a changedecision by carefully deliberating the expected value of making a change decision
(deliberativemind-set).Another third made specificplans for implementing an
intended project by delineatingexactly when, where, and how they wanted to
initiate relevant actions (implemental mind-set). Once both groups of subjects
had finishedthesementations,they were askedto work on a contingencyproblem
that presentedfrequent noncontingenttarget outcomes(a 75-75 problem). The
instructionsfor completing this task were identicalfor both groups of subjects;
that is, they had to discoverhow to producetarget light onset.A set of 40 trials
was offered.I7e also addeda control group (the last third of the subjects)that did
not receive any mind-set manipulation, but worked only on the contingency
problem.
Deliberative mind-set subjectsagainshowedthe most accuratejudgments of
control; that is, their control judgmentswere lower than those of control and
implemental subjects.Implemental mind-set subjectsevidencedcontrol j"dgments that were even (albeit not significantly) more illusionary than those of
control subjects.Supportingour mind-set interpretationof thesefindings, deliberative mind-set subjects'judgments of control correlated negatively with the
"personal
reported
importance" of the problem pondered during the predecisional mentation. Apparently, the more involved subjects'deliberation was, the
more realistic their subsequentjudgments of control were. Ä parallel finding was
observedfor implemental subjects.F{ere, judgments of control were positively
related to subjects' anticipated frustration if the project should (for whatever
reason)fail to be implemented.
The present findings not only support our hypothesis that a deliberative
mind-set spawnsaccuratecontrol judgmentswhereasthe implemental mind-set
favorsillusionary optimism, but also offer a new perspectiveon the phenomenon
of illusion of control. It seemspossiblethat illusion of control is generally based
on an implemental mind-set. That is, whenever subjectscommit themselvesto
achievinga desiredoutcome,they are likely to experienceillusion of control. Data
from Martin, Abramson, and Alloy (1984) and Vazquez(1987) support this line
of thought: \Thenever it was made difficult for subjectsto commit themselvesto
achievingthe target outcome,subjectsfailed to evidenceillusion of control. Also,
when Langer (1975) introducedthe conceptof illusion of control, she discovered
that variousfactors making a luck task (i.e.,random outcomes)appear to be a skill
task manageto produceillusion of control. Viewed in the context of our theoreti-
7B
Part I. Self'Regulationand Self-Control
cal framework, the presence of skill-related aspectsin a task may induce subjects
to commit themselves ro the goal of achieving the desired outcome, even though
its appearanceis solely determined by chance.In other words, a goal commitment
emerges that allows for the development of an implemental mind-set.
Finally, our mind.-setconceptualization also provides a new framework from
"depressive
realism" (i.e., the observation that depressedpeople do
which to view
not experience illusion of control when noncontingent outcomes appear freqoently;. Conceivabiy, depressives find it particularly difficult to set for themselvesthe goal at hand (e.g.,to maximize target light onset) becauseof pervasive
negative beliefs about themselves-that is, their abilities, past performances,
intelligence, and strengths (Beck, 1961, 1976). Such beliefs should generate
doubts concerning the attainability of the given goal and thus should hinder goal
commitmenr. Consequently,no implemental mind-set may evolve, and depressed
people thereby remain insusceptible to feelings of illusionary optimism.
Summary
The findings of the last four experimentsreportedsuggeststrongly that
deliberativemind-set subjectsdo analyzethe positive and negativeconsequences
of a prospectivegoal impartially.In addition,they make rather accurateestimates
of action-ourcomeprobabilities.The latter is definitely not true of implemental
mind-setsubjects:They overestimatetheseprobabilities,thus showing illusionary
optimism.
One wonders whether the deliberative mind-set favors accurateprobability
judgments in general, and not only when action-outcome expectanciesare at
stake.\When wishes are deliberated,the probability (certainty) that the desired
outcomemay be reachedis an important concern.Clearly, part of the answer to
this questionis relatedto action-outcomeexpectancy(i.e.,how certainPeopleare
that they can conrrol the desiredoutcome).However, there are other possible
answers.The more general answer relatesto people'sbeliefs that they possess
of relevantcompetencies);the
relevantaction potentials (i.e.,their self-concepts
more specific answer relates to beliefs associatedwith specific coursesof action
(i.e.,how cerainis it that doing X will leadto the desiredoutcome).Certaintyis of
relevance again when an individual ponders the probability that the achieved
To answer this question,the individourcomewill lead to a desiredconsequence.
ual must estimatethe probability that certain eventswill occur.As we know from
decisionresearch(Baron, 1988; Slovic,Lichtenstein,& Fischhoff,1988),people
havedifficulty with deriving accurateestimates.They employ numerousheuristics
to easethis task (Nisbett & Ross,1980;Tversky& Kahneman,1973,1974),b:ut
more often than not go astray(e.g.,the gambler'sfallacy,the conjunctionfallacy).
In particular, they are overconfident when estimating the probability of desired
evenrs(Hoch, 1985)or frequentevents(Fischhoff,Slovic,& Lichtenstein,1977).
In add,ition,they cling ro an initial estimateevenwhen evidenceaccruesthat urges
correction;in other words, people generallyfail to aPplyBayes'stheorem (Birnbaum & Mellers,1983).
Chapter 2. Action Phasesand Mind-Sets
79
Should deliberativemind-set subjectsdo better in all of these probability
judgments? Or do improvements in probability judgments generally require
instructionin statistics(Nisbett,Fong,Lehman,& Cheng,1987)? For someof the
above-reportedfailures in assessingthe probability of certain events correctly,
intensiveinstruction in statisticsseemsnecessaryto achievenoticeableimprovements (e.g.,when the applicationof the Bayesiantheoremis calledfor); however,
for other failures,such as the overconfidencephenomenon,benefit may result
from a deliberativemind-set.\fith respectto probabilityjudgmentsrelatedto the
feasibility of the desired outcome, however, one would expect that all relevant
judgments should becomemore accuratein a deliberativemind-set.
Moreover, one wonders whether the deliberativemind-set may also reduce
shortcomings people evidencewhen analyzing the desirability of a choice. Äs
decisionresearchershave repeatedlyobserved(Baron, 1988;Slovic et al., 1988),
people ignore minor differencesbetweenoptions; they employ simplified strategies (e.g.,elimination by aspects);they avoidtradeoffsbetweenequailyimportant
they weight negativeconsequences
differently from positive conseconsequences;
quences,thus falling prey to framing effects; and they overweight consequences
that are absolutelycertain (the certaintyeffect).Moreover,many potential consequencesare not even considered,and relevant values,attitudes,and goals that
are ignored (in particuallow one to estimate the attractivenessof consequences
lar, those that are contradictory).Again, one may raisethe questionof whether
people need explicit training to avoid theseshortcomings.So far, no researchhas
been conductedto explore whether a deliberative mind-set reducesthese shortcomings.
Researchconductedin the realm of person memory suggeststhat this may be
possible.During impressionformation, people give greater attention and processingto relatively infrequent behaviorsthat do not match the overall picture of
the target person (i.e., are inconsistentor contradictory).However, once an
impressionhasbeenformed,it is the consistentinformation that is preferentially
encodedand better remembered(for reviews,seeHiggins & Bargh, L987;Ruble &
Stangor, 1986). Apparently, as soon as one is ready to act on an impression
formed (implemental mind-set), the thorough analysisof the information at hand
(associatedwith the deliberative mind-set during impression formation) vanishes.
Mind-Sets and Open-Mindedness
S7ehypothesizedthat the deliberativemind-set increasesa person's receptiveness
to a broad range of incoming information, whereas the implemental mind-set
decreases
suchreceptiveness.$7ethought that two different structural qualities of
a person'sinformation processingmight accountfor this. First, the more heeded
information (i.e., information to which individuals are attending) people can
encodeinto short-term memory, the higher their receptivity should be. Second,
people should show higher receptivity if their allocation of attention is very
B0
Part l. Self-Regalationand Self'Control
mobile and flexible, becausethat should allow them to pick up peripheral information. Accordingly, we conductedexperiments that measured amount of heeded
information encoded and allocation of attention in deliberative and implemental
mind-sets.
Noun Span
As pointed out by Dempster (1981) and others (Case,Kurland, & Goldberg,
7982; Chi, L976), the width of a person'snoun span is a good indicator of the
amounr of heeded information a person successfullyencodesinto short-term
memory. In a typical noun span experiment, subjectsare read a list of words
presentedlessthan 1 secondapart.\7hen the experimenterhas pronouncedthe
last word of the list, subjectsare requestedto immediately rePeatall of the words
in the order presented.
For our experiment,we (Heckhausen& Gollwitzer,1987,Study2) prepared
a whole set of word lists, including lists consistingof five, six, and sevenonesyllable nouns. S7eread a first block of word lists to subjectsto obtain baseiine
scoresof their noun span.Then subjectswere given the cover story of the first
experiment reported above,which offered subjectsa choicebetweentest materials
for a creativitytest (Heckhausen& Gollwitzer,1987,Study 1). The subjectswere
interrupted either prior to making a choice of test material (the deliberative
mind-setgroup) or shortly thereafter(the implemental mind-setgroup). Immediately after the interruption, another block of word lists was read to the subjects.
The words on the lists for the baselinemeasureand the critical secondtest were
irrelevant to making a choicebetweentest materials.\7e also employed a control
group that was never offered a choice between test materials, but still took the
baselineand critical noun span test.
\7hen computing the noun spansvia a classicproceduredescribedby \7oodworth and Schlosberg(1954,p. 696),we found that deliberativemind-setsubjects
showed a broader span in the second,critical assessment,as compared to their
baseline scores. Their noun span increased after they had moved from the
ordinary test situation (testing sessionfor baselinedata) into a deliberativemindset associatedwith contemplating the choiceof test material. This increasednoun
span was also significantly elevatedas compared to both the first and second
assessmentsof the control group. For the implemental mind-set group, no
significant decreasein noun span was observedbetween the first and second
possibly,the task of planning one's performanceon the chosen
assessments;
creativity rest was just not complex and involving enough to createa Pronounced
implemental mind-set.
It might be suggestedthat the superior performance of the deliberative
mind-set subjectson the noun span test may have been rooted in an increasein
nonspecificactivation causedby disrupting deliberation.Flowever,the resultsof a
further study (Heckhausen & Gollwitzer, 1986, Study 3) do not support this
explanation. This study employed exactly the same paradigm as the noun span
experimenr, except that subjectsworked on simple arithmetic tasks instead of
recalling lists of nouns.I7e thought that performing highly routinized and nearly
Chabter2. Acti.onPhasesand Mind.-Sets
81
informationless mental tasks such as these should profit from an increasedlevel
of activation. Accordingly, if performance on such tasks was not facilitated when
these tasks were solved within a deliberative mind-set, nonspecific activation was
unlikely to have produced the increase in noun span. As it turned out, no
differences in performance were observed between the groups.
Mobile Allocation of Attention
The last study to be presentedhere suggeststhat a deliberativemind-set
involves a greater receptivity to incoming information, becausecomparatively
more heededinformation is encoded.But there should be a secondsourceof
greaterreceptivity-that is, an increasedreadinessto encodeperipheral,incidental information. Flexible, mobile allocation of attention should enhance the
individual'schancesof encodingsuch information, whereasallocatingattention
solely to information that is alreadyheededshould reducethem.
\Weconductedthe following experiment (Gollwitzer &. Heckhausen,1987,
Study 1) to explore whether the deliberativemind-set increasesflexible allocation
of attention and the implemental mind-set reducesit. The idea underlying this
study drew on researchdone by developmentalpsychologistsinterested in the
developmentof attentional processesin children (Hagen & Hale, 1973).Much of
"centralthis research(e.g.,Miller & S7eiss,1981) has involved the so-called
incidental learning task," in which children are askedto remember the location of
certain objects designatedas central while other objects designatedas incidental
serveas potential distractors.At the end of this task, children are unexpectedly
asked to recali the incidental objects in addition to the central objects; this
determines whether attention was also allocatedto the incidental information.
$7emodified this task so that adults would also find it involving. In our task,
the central information was a short story consistingof factualstatementsabouta
small university in northern Germany. This story was presentedsentenceby
sentence,and subjectswere told to memorize it for later recall. The incidental
information took the form of single, unrelated two-syllable nouns placed next to
each sentence.These nouns also were irrelevant to the decisionalproblem that
was used to create deliberative and implemental mind-sets (see below). After
presentation of the stimulus material, a rccognition test for the incidental information was applied. Recognition of the incidental material seemed a more
appropriate measure than recall, becauserecall speaks to the accessibilityof
information in memory, whereas the availability of the incidental information
was at issue here. The latter is more sensitively captured by a recognition
procedure(seeBargh & Thein, 198); Srull, 1981, 1984>.
Subjectsviewed the information and had to recall it either prior to making a
decision (deliberativemind-set group) or shortly thereafter (implemental mindset group). In the control group, subjectsworked on the central-incidental task,
neither having made nor expecting to make a decision.The decisionthat subjects
had to make was embeddedin the following cover story: Subjectswere to play the
part of a personnel manager. The problem at hand was to hire one of two
applicants for the position of a product manager.
Part l. Self-Regalationand Selt''Control
82
Deliberative mind-set subjectsrecognized the incidental information significanrly better than implemental mind-set subjects,with control subjectsshowing a
recognition performance that tended to be better than that of the implemental
minä-set subjects and worse than that of the deliberative mind-set subjects. In
addition, there was no difference among these three grouPs with respect to how
certain they felt about the correctness of their judgments. Äccordingly, the
hypothesis that the motivational mind-set is characterizedby more mobile allocation of attention than the implemental mind-set finds suPport in this pattern of
data.
Summary
The last two experimentspresentedsuggestthat deliberativemind-set subjectsare more recepriveto availableinformation than are implemental mind-set
subjects.This is becausedeliberativemind-setsubjects(1) encodecomParatively
more heededinforrnation into short-term memory, and (2) show more mobile
allocation of attention. The conductedexperiments suggestthat the deliberative
mind-set enhancesreceptivity via encoding of heeded information and that it
tends to do so via mobile allocation.\7ith respectto the implemental mind-set's
postuiated reduction in receptivity, our findings are much weaker. There is no
reduction in terms of encoding,and only a tendencywith respectto allocation of
arrention. Possiblywe failed to create strong implemental mind-sets in both
experimenrsreported. Subjectsmay not haveexperienceda full-blown implemental mind-set, and this may have been the reason why we did not observe any
significant reductions in receptivity.
AND MECHANISMS
THEIRCORRELATES
MIND.SETEFFECTS:
Not ail mind-sets influence cognitive processingto an equal degree;some are
more potenr than others. \ü7hat are the correlates (variables) that accompany
these differences?These variablesshould be different for each of the actionrelated mind-sets outlined at the beginning of this chapter (the deliberative,
implemental, actional.,or evaluativemind-set). But all of these variables are
related to the individual's involvement with trying to solve the task associated
with the respecriveaction phase (the predecisional,preactional,actional, or
postactionalphase).
For instance,in the predecisionalphase the task is to determine which of
one's wishes is most desirablebut still feasible.\Whether intensive task involvement (i.e.,intense deliberation)will actuallyoccur dependson the wish that is
scrutinized,the surrounding environment, and various personal attributes.Inrensedeiiberation should be hindered when the individual's freedom of choice is
restrictedby others (e.g.,superiorswho make the final decision);when habit or
need intervenes(i.e., there is no questionwhich wish will be chosen);or when
aheadychosen superordinate goals determine which subordinate wish will be
implemented (i.e., goal-closeddecisions;Toda, I976). It should be stimulated
Chapter2. Action Phasesand Mind-Sets
83
wheneverconflict existsbetweenwishesthat appearequallydesirableor between
feasibilityand desirabilityof a given wish. Other variablesthat should stimulate
intense deliberationare the accountabilityand irreversibility of the individual's
choiceand the amount of information available.A number of potentially relevant
personalattributes,suchascertaintyorientation (Sorrentino& Short, 1986),state
orientation (Kuhl, 1984), sensitizationcoping style (Olson & Zanna, 1979),
failure threat and depression(Pietromonaco& Rook, 1987), and the state of
(Wicklund & Ickes,1972),may all favor intense deliberaprivate self-awareness
tion. Finally,it alsoseemspossibleto intensify deliberationthrough self-instructions and through instructionsgiven by the experimenter (seeMischel's[1983]
work on self-regulation).
All of these variablesdeterminehow involved a person may becomewith
deliberatingwishes,and thus have to be consideredas potential correlateswith
the potency of the deliberative mind-set. Similarly, becoming intensely involved
with the task of promoting the initiation of actions(implemental mind-set),the
effectiveexecutionof goal-directedactivities(actionalmind-set),and the proper
evaluationof the effects of one's goal striving (evaluativemind-set) should all
dependon the specificqualitiesof the problem at hand, the situationalcontext,
and variouSpersonality attributes.I havetried to specifythesevariableselsewhere
(Gollwitzer, 1990);as it turns out, there are different correlateswith the potency
of eachof thesemind-sets.I do not repeatthis analysishere,but insteadraisethe
question of what types of mechanismsproduce mind-set effects.
The classicdefinition of mind-set (Einstellung) advancedby the Wörzburg
schoolsuggeststhat the mechanismsmediating mind-set effectsare locatedin the
cognitiveprocessesadvancingthe solution of the task that stimulatedthe mindset. In the presentedresearchon deliberativeand implemental mind-sets,we
observedthat mind-sets affectedsubjects'thought production, the recall of congruous information, the analysis of desirability-relatedinformation, the inferencesmade on the basis of feasibility-relatedinformation, and finally the attentional processeswhen irrelevant information had to be encoded.It appears,then,
that deliberativeand implemental mind-sets make any knowledge that helps to
Part of this knowledge is categoricalor
solve the respectivetask more accessible.
episodicand relatesto the specificproblem at hand (i.e.,the decisionto be made
or the project to be planned). The other part is procedural and relates to how
wishesare deiiberated(deliberativemind-set) or how projectsare planned (implementalmind-set) in general.It is this latter part that we found to transfer to
subsequent,unrelated tasks.
In this sense,the observedmind-seteffectsare most similar to the cognitivetuning effects originally analyzedby Zajonc (1960) and extended by Brock and
Fromkin (1968),Cohen (1961),Leventhal(1962),and most recentlyby Higgins,
McCann, and Fondacaro (7982). This researchemploys a paradigm in which
subjectsare assigneddifferent tasks.Half of the subjectsare told to transmit their
impressionof a target personto others,whereasthe other half arc told to receive
others' impressions of the target person. Subsequently,how subjects organize
information on the target personand what kind of information is suppressedare
84
and Self'Control
Partl. Selt''Regulation
analyzed.Clearly,thesestudiesalso demonstratethat different task assignments
may act as steering mechanismsfor organizingpresentedinformation. The researchreported here expandson this ideaby stating that the taskspeople faceat
the various action phasescreatedistinct mind-setsthat tune people'scognitive
functioning.
The observedmind-seteffectsshouldnot be confusedwith researchfindings
"New
Look" in perception (Bruner, \957;
reported under the heading of the
"category
Bruner & Goodman, L947). This researchintroduced the notion of
accessibility."It is assumedthat the easewith which a given stimulus input is
codedin terms of a given categorydependsnot only on the match between the
featuresof the stimulus and of the category;other factors,such as expectancyand
needstates,can alsoincreasethe likelihood that a particularcategoryrather than
an alternative will be applied to the input. More recently,it has been demonstrated that simply activating (priming) a concept in one task is capableof
increasingthe accessibilityof that construct in an unrelated subsequenttask
where subjects are asked to categorizea target person's behavior (Higgins,
Rholes,&Jones, 1977;Srull * $(yer, 1979).In general,this researchfocuseson
so that it influencesthe interpretahow a certain categoryis mademore accessible
tion of availableinformation in its own terms. Our mind-set research,on the
other hand, exploresthe effectsthat becomingintensely involved with the task of
deliberating or planning has on the accessibilityof appropriate cognitive procedures.In this sense,our researchis more similar to recent attempts to delineate
effectsof goals or roles on the elaboration and organizationof availableinformation, as reported by Srull and \Wyer(1986) and Zukier (1986), respectively.
Also, there is the question of whether mind-set effects are basedon active,
consciousprocessesinvolving deliberate strategies and control, or on passive,
unconsciousprocessesthat occur automaticallyand are uncontrolled (Higgins &
King, 1981; Posner& S7arren, 7972) . As Bargh ( 1989) pointed out, these aspects
of being an intentional (active) processcharacterizedby awarenessand control
may nor always come in the two configurations described (i.e., all three are
presentor all three are absent),as seemsto be the casewith our subjects.First,
our subjecrswere not awareof the mind-set effectswe observedin the illusion-ofstudy (i.e., subjects'cercontrol study or in the mobile-allocation-of-attention
tainty ratings were high and did not differ betweengroups). Second,however,if
we had made subjectsawareof them (and this is also true for the fairy tale study),
subiects could have easily halted (controlled) them. Finally, deliberative and
implemental mind-set effects would not occur in the absenceof an explicit
intention to deliberate an unresolved problem or to plan a chosen project,
respecrively.So it appearsthat mind-setscarry more of the qualities of activesets
than of passivesets (Higgins & King, 1981).
Finally, how do we accountfor the observedtransfer of the cognitive processesstimulated by deliberatingor planning to subsequentunrelated tasks?This
effect reminds of Luchins's (1942) problem-solvingexperiments,in which he
demonstrated that when subjectsrepeatedly solved a given type of arithmetic
Chapter2. Action Phasesand Mind-Sets
85
problem suggestinga certain strategy,they then applied this strategyto subsequent arithmetic problems even when other ways of solving the problem were
possibleor necessary.
Obviously,practicedmental operationsmay transfer frorn
the training context to subsequentcontexts to which they do not immediately
aPply. The questionsof what featuresof the subsequentcontext enhancethis
effect, how much time may pass before such effects vanish, and what exactly
happens to mental proceduresduring practice remain open. Recent researchby
J. R. Anderson on the acquisitionof cognitive skill has begun addressingthese
issues(Anderson,1982,1,987;in the realm of socialcognition,seeSmith, Branscombe,& Borman, 1988;Smith & Lerner, 1936).
GENERAT
SUMMARYAND CONCLUSION
The starting point of this chapter is the questionof whether the courseof events
associatedwith goal-oriented behavior is homogeneous. Although early researchers(e.g.,Narziss Ach, Kurt Lewin) studying goal-oriented behavior vehemently argued against such a view by suggesting that goal setting and goal
striving are ruled by different principles, this insight was widely ignored in later
researchon motivation (e.g.,in researchstimulated by Atkinson's risk-taking
model). S7eintroduceda model of action phasesin a renewedattempt to delineate
the various distinctive phenomena of goal-orientedbehavior.These phenomena
are consideredto be deliberatingwishes (potential goals),planning the implementation of chosen goals, acting on these goals, and evaluating one's goal
striving (i.e.,the outcomeand its consequences).
Under the assumptionthat thesephenomenapresent themselvesto individuals engaged in goal-oriented endeavors as tasks that need to be solved in
succession,the conceptof mind-set has been introduced.It has been argued that
being involved with these tasks leads to characteristic cognitive orientarions
(mind-sets) that are beneficial for solving these tasks efficiently. The cognitive
orientations related to eachof these tasks (or phenomena) have been spelledout.
Finally, a number of experiments have been reported that empirically tested the
cognitive orientations postulatedfor the deliberativeand implemenral mind-sets.
This researchhas shown that the deliberativemind-set is characterizedby cognitive tuning toward outcomeexpectancyand expected-value-related
thoughts and
information, by an accurateanalysisof outcome-expecrancy-related
information
and an impartial analvsisof expected-value-related
information, and by a heightened general receptivity to availableinformation. The implemental mind-ser, on
the other hand, is characterizedby cognitive tuning roward the implementational
thoughts and information, and by an optimistic analysisof expecrancy-related
information.
These findings strongly suggestthat researchersof motivation should question the still-common view that goal-orientedbehavior is a homogeneousphenomenon. It seemsmore appropriate to conceiveof goal-oriented behavior as a
86
and Self-Control
Partl. Self-Regulation
of distinctivephenomenathat are ruled by their own principles.The
succession
very recentrevival of interest in goal concepts(Pervin, i989) and in the issueof
commitment (Brickman, 1987) seemsbasedon this vievrBut the presented findings are also important in their own right. They
suggestthat the individual's cognitive apparatusreadily adjuststo the various
demandsof goal-orientedbehavior,and they thus stimulatea new persPectiveon
common topics in the psychologyof motivation (e.g.,illusion of control), clinicai
psychology(e.g.,depressiverealism),decisionmaking (e.g.,certaintyjudgments),
and socialcognition (e.g.,impressiontesting vs. impressionformation). Finally,
they imply interestinganswersto the questionof how peoplecan more effectively
rurn their wishes into action.
Acknowledgments
I arn gratefulto John
This chapteris dedicatedto the memoryof Heinz Heckhausen.
Bargh,Tory Higgins,ChristineLiu, RichardSorrentino,and RobertWicklundfor their
comments on an earlier draft.
References
Ach, N. (190r). Über d,ie \Y'illenstätigkeitund das Denken [\Tilling and thinking]. Göttingen:
Vandenhoeck* Rupprecht.
Ach, N. (1910).Über den \Y/illenvkt and.dat Temperamenr[Temperamentand the act of willing].
Leipzig:Quelle& Meyer.
Ach, N. (1935).Analyte des \Y'illens[The analysisof willing]. Berlin: Urban & Schwarzenberg.
Alloy, L.8., & Äbramson,L. Y. (1979).Judgmentof contingencyin depressedand nondepressed
students:Sadderbut wiser?Journalof ExperimentalPsychology:General,108,44I-485.
depressionand the illusion of control.
Alloy, L. B., & Abramson,L. Y. (1982).Learnedhelplessness,
Journal of Personality and'Social Psycb ologit, 42, 1114-1126.
In L. B'
Alloy, L. B., & Abramson,L. Y. (1988).Depressiverealism:Four theoreticalperspectives.
Alloy (Ed.), Cognitiaeprocertesin depretsioo(pp.22i-265). New York: Guilford Press.
Reuieu, 89,369-406.
Anderson,J. R. (1982).Acquisitionof cognitive skill.Psychological
Anderson, J. R. (1987). Skill acquisition: Compilation of weak-minded problem solutions.
Psych ological Reaieu.,,94, I92-210.
Atkinson,J.\7. (1957).Motivationaldeterminantsof risk-takingbehavior.PrychologicalReuiett,64,
4\g-17)
Atkinson,J. If. (1964). An introduction to motiaation. Princeton, NJ: Van Nostrand.
Atkinson,J. W. (1974). Strength of motivation and efficiency of performance. InJ. \7. Atkinson and
V. H. STinJ. O. Raynor (Eds.), Motiaation and.achieaentent (pp.193-218). \Tashington, DC:
ston.
Atkinson, J. S7.,& Birch, D. (1970). Tbe d.Tnatnicsof action. New York: ITiley.
Atkinson, J. W., & Reitman, !7. R. (i956). Performance as a function of motive strength and
expectancyof goal attainment. Journal of Abnormal and Social Psychologlt, 53,36I-366.
Bargh, J. A. (1989) Conditional automaticity: Varieties of automatic influence in social perception and
cognition. InJ. S. Uleman &J.A Bargh (Eds.), Unintended thought (pp.3-5t). New York:
Guilford Press.
Bargh, J. A., & Thein, R. D. (1985). Individual consrruct accessibility,person memory, and the recalljudgment link: The case of information overload. Journal of Pertonality and. Social Prychology,
49, 1L29-1146.
Baron, J. (1988). Thinking and.decid.ing.New York: Cambridge University Press.
Chaöter 2. Action Pbases and Mind-Sets
87
Beck, A. T. (1967). Depression: Clini.cal,experimental, and theoretical aspectr. New York: Hoeber.
Beck, A. T. Q976). Cognitioe therapy and the emotional dhorders. New York: International Universities Press.
Beckmann,J.,& Gollwitzer, P. M. (1987).Deliberativeversusimplemental statesof mind: The issueof
impartiality in pre- and postdecisional information processing. Social Cogniti.on, 5,2t9-279.
Birnbaum, M. H., & Mellers, B. A. (i983). Bayesianinference:Combining baserates with opinions of
sourceswho vary in credibility. Journal of Personality and. Soc.ialPsycbologjt,45,792-804.
Bobrow, D. G., & Norman, D. A. (197t). Some principies of memory schemata.In D. G. Bobrow &
A. Collins (Eds.), Representationand.understanding (pp. 131-149). New York: Academic Press.
Bobrow, D. G., & \7inograd,T. (1977). An overview of KRL, a knowledge representation language.
Cognitiue Science,1, 3-46.
Boring, E. G. (i950). A history ol experimental psychologT.New York: Appleton-Cenrury-Crofts.
Brehm,J. \7. (1956). Postdecisionchangesin the desirability of alternatives.Journal of Abnormal and.
Social P ry cbolo gy, 52, 384-389.
Brehm, J. W., & Cohen, A. R. (1962). Explorations in cognitiue d.issonance.New York: I7iley.
Brehm, J. W., \Wright, R. 4., Solomon, S., Silka, L., & Greenberg, J. (1983). Perceived difficulty,
energization, and the magnitude of goal valence.Joarnal ot' Experimental Social Psycbologlt, 19,
2r-48.
Brickman,P. (1987).Comnzitrzent,cont'lict,and.caring.EnglewoodCliffs, NJ: Prentice-Hall.
Brock, T. C., & Balloun,I. L. (1967).Behavioralreceptivity to dissonantinformation. Journal of
Abnormal and SocialPsychology,6, 413-428.
Brock,T. C., & Fromkin, H. L. (i968). Cognitivetuning set and behavioralreceptivityto discrepant
information.Journal ot' Personality,36, 108-125.
Brockner,J.,* Rubin,J. Z. (L985).Entrapmentin escalatingconflicts:A socialpsychologicalanalysis.
New York: Springer.
Bruner,J. S. (1957).On perceptualreadiness.
Psychological
Reuiett,64,I23-1r2.
Bruner,J. S.,& Goodman,C. C. (1947).Valueand needasorganizingfactorsin perception.Journalof
Abnormal and,SocialPsychology,
42,13-44.
Buchanan,B. (I974). Building organizationalcommitment:The socializationof managersin work
organizations. Ad.minbtratia e ScienceQuarterllt, 19, 133-546
Carver,C. S., & Scheier,M. F. (1981). Attention and selt'-regulation:
A control-theoryapproacbto
ltuman behaaior.New York: Springer.
Case,R., Kurland, D. M., & Goldberg,J. Q982).Operationalefficiencyand the growth of short-term
memory span.Journalof ExperimentalChild.Psychologlr,
33,386-404.
Chapman,D. W. (1932). Relative effectsof determinateand indeterminateAufgaben.American
Journal of Prychologt, 44, 763-174.
Chi, M. T. H. (1976).Short-term memory limitations in children: Capacttyor processingdeficits?
Memoryand Cognition,4,559-572.
Cohen,A. R. (1961).Cognitivetuning asa factoraffectingimpressionformation.Journalof PersonalitJl,29,235-24r.
Csikszentmihalyi,
M. (T97t). Beyondboredomand.anxiety.San Francisco:
Jossey-Bass.
Davidson,J. R., & Kiesler,S. B. (T964).Cognitivebehaviorbeforeand after decisions.In L. A. Festinger
(Ed.), Conflict, d.ecision,
and dissonance(pp. 45-6I). Stanford,CA: StanfordUniversiry Press.
Dellarosa,D., & Bourne,L. E. (1984).Decisionand memory:Differential retrievabilityof consistent
and contradictoryevidence.
Journalof VerbalLearningand.VerbalBehaaior,23,669-682.
Dempster,F. N. (1985). Short-term memory developmentin childhood and adolescence.
In C. J.
Brainerd (Ed.), Basicprocesses
in memory deuelopment(pp.209-248). New York: Springer.
Emmons,R. A. (1989).The personalstriving approachto personality.In L. A. Pervin (Ed.), Goal
conceptsin personalityand.socialpsychology(pp. 87-126).Hillsdale,NJ: Erlbaum.
Ericsson,
K. A., & Simon,H. A. (1980).Verbalreporrsas data.Psycbological
Reuieut,87,2l5-25I.
Farrell,D., & Rusbult,C. E. (1981).Exchangevariablesaspredictorsof job satisfaction,job commitment,and turnover:The impactof rewards,costs,alternatives,and investments.Organizational
Bebaaiorand Human Performance,
28, 78-95.
BB
Part l. Self-Regulationand Self-Connol
Feather, N. T. (1967). Valence of outcome and expectation of successin relation to task difficulty and
perceived iocus of control. Journal of Personality and.Social Psychologjt,7,372-186.
Festinger, L. A. (1944. A theoreticai interpretation of shifts in level of aspiration. Psychological
R e u i e t u ,4 9 , 2 3 > - 2 t 0 .
Festinger, L. A. (Ed.). (1964). Conflict, d.ecirion,and dissonance.Snnford, CÄ: Stanford University
Press.
Fischhoff, B., Slovic, P., & Lichtenstein, S. (1917). Knowing with certainty: The appropriateness of
extreme confidence. Journal of Experimental Psychology: Hurnan Perception and. Performance,
3 ,t 5 2 - 5 6 4 .
Frey, D. (1936). Recent researchon selectiveexposure to information. In L. Berkowitz (Ed.),Adaances
in experimental social psychology (Vol. 19, pp. 4i-80). Orlando: Academic Press.
Gallistel, C. R. (1980). The organization of action: A new rynthesb. Hillsdale, NJ: Erlbaum.
Gallistel, C. R. (198t). Motivation, intention and emotion: Goal-directed behavior from a cognitiveneuro-ethological perspective. In M. Frese * J. Sabini (Eds.), Goal-directed behaaior: The
concept ot' action in psycbologlt (pp. 48-65). Hillsdale, NJ: Erlbaum.
Gerard, H. B. (1967). Choice difficulty, dissonance,and the decision sequence.lournal ot' Personality,
35.9r-r08.
Gibson, I. I. Q94T). Ä critical review of the concept of set in contemporary experimental social
psychology. Psy ch olo gical Ball et in, 3 8, 7 8I -817 .
Gollwitzer, P. M. (1987). The implementation of identity intentions: A motivational-volitional
perspective on symbolic self-completion. In F. Halisch * J. Kuhl (Eds.), Motiaation, intention
and a olition (pp. 349-169). Heidelberg: Springer-Verlag.
Gollwitzer, P. M. (1990). Abuägen und. Planen ab BewuJ3tseinslagenfDeliberating and planning as
mind-sets]. Göttingen: Hogrefe.
Gollwitzer, P. M., & Heckhausen, H. (1937). Bread,th of attention and the coanrerplea bearistic:
Further eaid.enceon tbe motitational and aolitional nzind'-setdistinction Unpublished manuscript, Max-Planck-Institut für psychologischeForschung, Munich.
Gollwitzer, P. M., & Heckhausen,H., & Ratajczak,H. (1990). From weighing to willing: Äpproaching
a change decision through pre- or postdecisional mentation. Organizational Behaaior and
Haman De cision Procesle t, 45, 41 -65.
Gollwitzer, P. M., & Heckhausen, H., & Steller, B. (1989). Deliberatioe and implemental mind-tets:
Cognitiae tuning toward congruou thoughtt and. information. Unpublished manuscript, MaxPlanck-Institut für psychologischeForschung, Munich.
Gollwitzer, P. M., & Kinney, R. F. (1989). Effects of deliberative and implemental mind-sets on
illusion of control. Journal of Personality and' SocialPsitchologt, 56,53I-542.
Goliwitzer, P. M., & N7icklund, R. A. (1985). The pursuit of self-defining goals. In J. Kuhl &
J. Beckmann (Eds.), Action control: From cognition to behaaior (pp. 61-85). Heidelberg:
Springer-Verlag.
Hacker,Iü. (1985). Activity: A fruitful concept in industrial psychology. In M. Frese &J. Sabini (Eds.),
Goal-directed behauior: Tbe concept of action in psychology @p.262-283). Hillsdale, NJ:
Erlbaum.
l{agen,J.\7., & Hale, G. A. (1973). The development of attention in children. In A. D. Pick (Ed.),
Minnetota Symposiam on Child. Psychology (Vol. 7, pp.Il7-1,40). Minneapolis: University of
Minnesota Press.
Heckhausen,H. (T977). Achievement motivation and its constmcts: A cognitive model. Motiaation
and.Emotion, 1, 283-329.
Heckhausen, H. (1987a). Intentionsgeleitetes Handeln und seine Fehler [Action slips]. In H. Heckhausen, P. M. Gollwitzer, & F. E. I7einert (Eds.), Jeueits d.es Rubikon: Der V'ille in den
Hunt anuit senscbat'ten (pp. 143-\7 5 ). Heidelberg: Springer-Verlag.
In H. HeckHeckhausen, H. (1987b). \7ünschen-\üählen-S7olien fWishing-weighing-willingl.
hausen, P. M. Gollwitzer, & F. E. I7einert (Eds.), Jenseits des Rubikon: Der lY/ille in d'en
Hunzantaissenscb aften (pp. 3-9) Heidelberg: Springer-Verlag.
Heckhausen,H., & Gollwitzer, P. M. (1986). Information processing before and after the formation of
Cbapter 2. Action Phasesand Mind-Sets
89
an intent. in F, Klix & H. Hagendorf (Eds.), In m.ertoriam Hermann Ebbinghaus: Symposium
on tbe stractareand function of buman ntemorl (pp. 1071-i082). Ämsterdam: Elsevier/NorthHolland.
Heckhausen, H., & Gollwitzer, P. M. (1987). Thought conrenrs and cognitive functioning in motivational versus volitional states of mind. Motiuation and.Emotion, l1, r0r-r20.
Heckhausen,H., & Kuhl,J. (1985). From wishes to action: The dead ends and short-cutson the long
way to action. In M. Frese & J. Sabini (Eds.), Goal-d.irected,
behaaior: The concept of action in
psychology (pp. i34-159). Hillsdale, NJ: Erlbaum.
Higgins, E. T., & Bargh, J. A. (1987). Social cognition and sociai perception. Annual Reuieu of
Prych olo gy, 3 8, 369-425.
Higgins, E. T., & King, G. (1981). Accessibiiity of social constructs: Information-processing consequences of individual and contextual variability. In N. Cantor & J F. Kihlstrom (Eds.),
Personality, cognition, and social interaction (pp.69-T2r). Hiilsdale, NJ: Erlbaum.
Higgins, E. T., McCann, C. D., & Fondacaro, R. (1982), The communication game: Goal-directed
encoding and cognitive consequences.
SocialCognition, 1,2L47.
Higgins, E. T., Rholes, N7.S., &Jones, C. R. (1977). Category accessibiliryand impression formation.
Journal of Experimental Social Psycbologjt, 13,l4l-I54.
Hillgruber, A. (1912). Fortlaufende Arbeit und N7illensbetätigung [Continuous work and willing].
Untersucbungen zur Psychologie und. Philosopbie, 1(6).
Hoch, S. J. (1985). Counterfactual reasoning and accuracyin predicting personal events. Journal of
Exp eriment al Psych olo gy : Learnin g, Menzory, and. Cognition, 1 l, I Ig -7 jI.
Humphrey, G. (19t1). Thinking. London: Methuen.
Insko, C. A., \7orchel, S., Folger, R., & Kutkus, A. (1975). A balance theory interpretation of
dissonance.Prychological Reaieut, 82, 169-I83.
Janis, I. L., & Mann, L. (1968). A conflict theory approach to attitude change and decision making. In
A. Greenwald, T. Brock, & T. Ostrom (Eds.), Psychologicalfoand.ations of attitudes (pp.327360). New York: Academic Press.
Jecker,J. D. (1964). The cognitive effects of conflict and dissonance.In L. A. Festinger (Ed.), Conflict,
d.ecision,and.d.issonance(pp.2l-10). Stanford, CA: Stanford University Press.
Jones, E. E., & Gerard, H. B. (1967). Found.ationsof social psychology. New York: STiley.
Kanter, R. M. (1972). Commitment and contmunity: Cornmunes and.utopias in sociologicalperspectiae. Cambridge, MA: Harvard University Press.
Kelley, H. H. (1983). Love and commitment. In H. H. Kelley, E. Berscheid, A. Christensen, H.
J.
Harvey, T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, & D. R. Peterson (Eds.), Close
relationship: (pp.26)-314). San Francisco: \7. H. Freeman.
Kiesler, L. A. (1971).Tbe psychology of commitrzen. New York: Academic press.
Klinger, E. (L977). Meaning and. aoid: lnner experience and. tlte incentiues in people's liues. Minneapolis: University of Minnesota Press.
Kornadt, H. J. (1988). Motivation und Volition: Anmerkungen und Fragen zur wiederbelebten
$TillenspsychologiefMotivation and volition: Comments on the revived psychology of willing].
Archia für Psychologie, 140, 209-222.
Kuhl, J. (1983). Motiuation, KonfliAt und. Hand.lungskontrolle. fMotivation, conflict, and action
controll. Heidelberg: Springer-Verlag.
Kuhl, J. (1984). Volitional aspects of achievement motivation and learned helplessness: Toward a
comprehensive theory of action control. in B. A. Maher (Bd.), Progrets in experimental
personalitT researcb (Vol. 13, pp. 99-U1). New York: Academic Press.
Kukla, A. (1972). Foundations of an attributional theory of performance. Psychological Reuieu,,,79,
454-470.
Langer, E. J. (1971). The illusion of control. Joarnal of Personalityand SocialPqtcbolog4, 32,31L-328.
Leventhal, H. (1962). The effects of set and discrepanry on impression change. lournal of Personality,
30,r-tl.
Lewin,K. (1926).Vorsatz,S7illeund Bedürfnisflntention, will, and need].Prychologische
Forscltung,
7, 330-385.
90
Part l. Self-Regalation and Self-Control
New York: McGraw-Hili'
Lewin, K. (19i6). Principlet ot' topologicalpsYcholog\'
InJ. M' Hunt (Ed'),
i. 4., & Sears,P. S. (1944).Levelof aspirationLewin,K., Dembo,T., Festinger,
(pp. 3i3-178). New York: Ronald Press'
personalit\and.the bebaaiord.isorders
Monographs,54,1-95'
solving.PsTchological
problem
in
Luchins,A. S. (7942).Mechanization
for predictingcontinuityof
scales
Lund,M. (198t). The developmentof investmentand commitment
2,3-23'
s. Journalof Socialand PersonalRelationships,
personalreiationship
informationon
forthcoming
of
anticipation
of
Mann, L., Janis,I. L., & Ciaplin, R (1969).Effects
11,10-16'
Journal of Personalityand SocialPtychology,
predecisionaiprocesses.
Y.,
& Älloy,L. B. (1984).illusionof controlfor selfandothersin depressed
Martin,D. J.,Abramson,L.
r25-l'36'
collegestudents.Journalof Personalityand SocialPsycbologL,46,
and nondepressed
social
and'
personality
ot'
(Ed.),
Reoietu
McClelland,D. C. (1980).Moiiu. dispositions.In L. Irheeier
psychology(Vol. 1, pp. i0-41). BeverlyHills, CA: Sage'
uon Erfolg and'MilSert'olglAchievement
U.y"f \7. U. (1973). Leisiungsrnotiuund' [Jrsachenerklärung
and failure]. Stuttgart:E. Klett.
motive and attributionsfor success
volontaireet sesant6c6dents
Michotte, A. E., & Prüm, E. (1910).f,tude e*p6rimentalesur le choix
10,lI9of ,h. act of choice).Archiaesde PsYchologie,
imm6diats.[The immediateantecede.r,,
)99
allocation, understanding
of attention'
and
H., & Süeiss, M. G. (1981). Children's attention
'perf6r.rru.r..
on the central incidental learning task.Child Deaelopment,52, 7183-1190'
variable in development'In
Mischel, W. (1933). Delay of gratification as processand as person
perspectiue
D. Magnusson& v. P. Allen (Eds.), Huntan deuelopnent: An interactional
(pp.149-16>).New York: AcademicPress'
linkages:Tbe psychologT
Mowday,R. T., Porter,L. W.,& steers,R. M. (1982). Emplo4ee-organization
and.tarnoaer.New York: AcademicPress.
ot' commitment,absenteeism,
reasoning.science,238,
Nisbett, R. E., Fong,G. T., Lehman,D. R., & cheng, P. W. (1987).Teaching
Miil.;;
625-61r.
shortcomings of social judgment'
Nisbett, R., & Ross, L. (i980). Human inference: Strategies and
Englewood cliffs' NJ: Prentice-Hatltr,
pfocessesln Plerception and
n
,no role of active
a*ive memory
Norman, D. A., & Bobrow, D. G. (1976). On the
(pp.
1i4-132)' san Francisco'
cognition. In c. N. cofer (Ed.), The structure of hurnan fTremory
\7. H. Freeman.
stage in memory retrievai'
Norman, D. A., & Bobrow, D. G. (1979). Descriptions: Än intermediate
Cognititte Psychology, I 1, 701-123'
p. (1919). A new look at selectiveexPosure'Journal of Experimental Social
Olson,J. M., &,Zanna, U.
Psycbologit,15,l-1r.
and psychological attachment:
o'Reilly, c. o., ]|I, & chatman, J. (1986). Organizational commitment
prosocial behavior' Journal of
The effects of compliance, identification, and internalization on
Applied P sycbology, 71, 492-499.
psycholog'Hillsdale, NJ: Erlbaum'
Pervin, L. A. (Ed.). (198t1. Goal conceptsin personality and social
contribution of perceived risks
Pietromonaco, P., & Rook, K. (1987). Decision style in depression:The
vs. benefits. Journal of Pertonalit\ and Social Ps|cbology, 52,399-408.
In A' \7' Melton &
Posner,M. I., & $Varren,R. E. (1972). Traces,concePtsand consciousconstructions'
\Tashington, Dc: v' H'
E. Martin (Eds.), cod.ing processesin human 7ltefnory @p.25-44).
\7inston.
Oxford University Press'
Rabkin, E. S. (1979). Fantastic tt,orld.s:Myths, tales and stories' Oxford:
An elaboration of the
O. (1969). Future orientation and motivation of immediate activity:
Raynor,J.
'
606-610.
th.o.y of achievement motivation. Ps\chological Reuiew, 76,
In G' Levinger g H' L' Rausch
Rosenblatt, P.C. (1977).Needed researchoo commitment in marriage'
Ämherst:
(Eds.), close relatiouhips: Perspectiaes on tbe meaning of intimacY Gp' 73-86)'
University of MassachusettsPress.
from developmental and
Ruble, D. N., & Stangor, c. (1986). Stalking the elusive schema: Insights
-26I'
social psychological analysesof gender schemas.social cognition, 4,221
Chapter 2. Action Phasesand Mind-Sets
91
Rumelhart,D. E. (197i). Notes on a schemafor stories.In D. G. Bobrow & A. M. Collins (Eds.),
and understanding
Repre:entation
Gp.2LI-236), New York: AcademicPress.
and summarizingbrief stories.In D. LaBerge&J. Samuels
Rumelhart,D.E. (1977).Understanding
(Eds.),Basicprocestesin readi.ngand.comprehension(pp.265-303). Hillsdale,NJ: Erlbaum.
Salancik,G. R. (i977). Commitment and'the control of orgaoizationalbehaviorand belief.In B. M.
in organizationalbehaaior(pp. i-54). Chicago:
Staw& G. R. Salancik(Eds.),Netu d.i.rections
St. Clair Press.
Schwartz,B. (1988).The experimentalsynthesisof behavior:Reinforcement,behavioralstereotypy,
and problem solving. In G. H. Bower (Ed.), The psychologyot' learning and motiaation:
Adaancesin researcltand theory (YoL.22, pp.93-li5). New York: AcademicPress.
Selz,O. (1910).Die experimentellelJntersuchungdesWillensaktes.Zeitschrift für Ptychologie,j7,
24r-270.
In M. Frese&J. Sabini(Eds.),
Semmer,N., & Frese,M. (1985).Action theoryin clinicalpsychology.
psychology
(pp. 296-310). Hillsdale, NJ:
behaaior: The concept ot' action in
Goal-directed.
Erlbaum.
Sigwart,C. (1889).Der Begrit'fdes\Y/ollentund tein Verbähnitzum Begrift'der UrsachelThe concept
of willing vs. the conceptof motivel. Freiburg:J. C. B. Mohr.
Slovic,P.,Lichtenstein,S.,& Fischhoff,B. (1988).Decisionmaking.In R. C. Atkinson, R. J. Hernstein,
G. Lindzey,& R. D. Luce (Eds.),Steuent'handbookof experimentalpsycbologjr:Learningand
cognition (Vol. 2, pp. 673-738).New York: S7iley.
Smith, E. R., Branscombe,
N. R., & Borman,C. (1988).Generalityof the effectsof practiceon social
judgment tasks.Journalof Personalityand SocialPsychologjt,
54,385-395.
Smith, E. R., & Lerner, M. (1986). Developmentof automatismof social judgments.Journal of
Personality and Soci.alPsych ology, 50, 246-259.
Sorrentino,R. M., & Short,J. (1986).Uncertaintyorientation,motivation, and cognition.In R. M.
of social
Sorrentino& E. T. Higgins (Eds.),Handbookof motitation and cognition:Found.ations
bebaaior(Vol. 1, pp.179-403). New York: Guilford Press.
Srull,T. K. (1981).Personmemory:Sometestsof associative
storageand retrievalmodels.Joarnalot'
ExperinzentalPsychology: Human Learning and.Memory, 7, 440-463.
Srull,T. K. (1984).Methodologicaltechniquesfor the studyof personmemoryand socialcognition.In
R. S. S7yer* T. K. Srull (Eds.),Handbookof socialcogni.tion(Vol. 2, pp.I-72). Hillsdale,NJ:
Erlbaum.
Srull, T. K., & \7yer, R. S. (1979). The role of categoryaccessibilityin the interpretation of
information about persons:Some determinantsand implications.Journal of Perconalityand
gy, 37, 1660-1672.
SocialPsycholo
Srull, T. K., & \7yer, R. S. (1986).The role of chronic and temporary goals in socialinformation
processing.
in R. M. Sorrentino& E. T. Higgins (Eds.),Hand.bookof motiuationand cognition:
Foundationsof socialbebaaior(Vol. 1, pp.503-549).New York: Guilford Press.
Toda,M. (1976).The decisionprocess.Ä perspective.lntemationalJournalof GeneralSystenu,3,19-88.
Tversky,4., & Kahnemann,D. (I97j). Ävailability:A heuristicfor judgingfrequencyand probability.
CognitiuePrychologt, 5, 201-232.
Tversky,4., & Kahnemann,D. (1974).Judgmentunder uncertainty:Heuristicsand biases.Science,
185,tt24-tt3r.
Vallacher,R. R., &'$7egner,D. M. (1987).Sfhat do peoplethink they'redoing?Psyc.hological
Reaiew,
94.3-r5.
Yazquez, C. V. (1987). Judgments of contingency: Cognitive biases in depressed and nondepressed
subjects. Journal of Personality and. Social Psycholo g1t,5 2, 419-431.
von Cranach, M. (1982). The psychological study of goal-directed action: Basic issues. In M. von
Cranach & R. Harr6 (Eds.), Tbe analysis of action (pp. 35-13). Cambridge, England: Cambridge
Universiry Press.
\Walster,E., & Festinger, L. (T964). Decisions among imperfect alternatives. In L. A. Festinger (Ed.),
Cont'lict, d.ecision,and,d.issonance(pp. I3l-l4t).
Stanford, CA: Stanford University Press.
92
Part L Self-Regalation and, Self-Control
Beiträgezu einerTheoriedesDenkensfExperimentson a theory
\üatt, H. J (190t). Experimentelle
of thinkingl.Archiu t'ilr d.iegesanttePsychologie,4,289-436.
rnotiuation
and.attributiontheory.Morristown,NJ: GeneralLearning
\7einer,B. (I974).Achieaement
Press.
with humanpotential.
Wickiund,R. Ä. (1986).Orientarionto the environmentversuspreoccupation
In R. M. Sorrentino& E. T. Higgins (Eds.),Hand.bookof rnotiuationand cogni.tion:Found.ations of socialbehaaior(YoI. I, pp. 64-95).New York: Guilford Press.
Hillsdale,NJ: Erlbaum.
on cognitiued.issonance.
\Ticklund,R. A., & Brehm,J.\7. (1976).Perspectiues
\Ticklund, R. 4., & Frey, D. (1981). Cognitive consistency:Motivational versusnon-motivational
on euerydayunderstand,ing
perspectives.
I. J.P. Forgas(Ed.), Socialcognition: Perspectiaes
(pp. 141-163).London:AcademicPress.
Sficklund,R. A., & Gollwitzer,P. M. (1982). Sym.bolicself-completion.Hlllsdale,NJ: Erlbaum.
exposure
on predecisional
\0ickiund,R. A., & Ickes,\7. J. Q972).The effectof objectiveself-awareness
8,378-i87.
SocialPsychologT,
to information.Journal ot' Exper'i.mental
H. (1954).Experimentalpsychology.New York: Holt, Rinehart &
lToodworth,R. S.,& Schlosberg,
\7inston.
Yerkes,R. M., & Dodson,J. D. (1908).The relation of strength of stimulus to rapidity of habit
Neurologyand PsycholoBI,18,459-482.
formation.Journal of Conaparatiae
(1960).
The processof cognitivetuning in communication.Journal of Abnormal and
Zajonc,R. B.
SocialPsychology,
61, L59-I61.
Zukier, H. (1986).The paradigmaticand narrativemodesin goal-guidedinference.In R. M. Sorrenot' social
tino & E. T. Higgins (Eds.),Hand.bookof motiration and cognition: Found,ations
behaaior(Vol. 1, pp. 465-502).New York: Guilford Press.
HANDBOOKOF
MOTIVATIONAND
COGNITION
of SocialBehavior
Foundations
VOLUME2
Edited by
E.TorYHiggins
Columb'ia IJnia er si'ty
RichardM. Sorrentino
[JniaersitYof \VesternOntario
THEGUILFORDPRESS
New York ' London