Polish Psychological Bulletin

Original Papers
2009, vol. 40 (2), 74-84
DOI - 10.2478/s10059-009-0008-0

Bogdan Wojciszke*,**
Wiesław Baryła*
Aleksandra Szymków-Sudziarska*
Michał Parzuchowski*
Katarzyna Kowalczyk*

Saying is experiencing: Affective consequences of complaining and affirmation

In four experiments mood was measured before and after complaining or affirmation. Participants complained or affirmed
either themselves or listened to such communications of another person. Mood decreased after complaining and increased
after affirmation – a “saying is experiencing” (SIE) effect. This effect was found also in the cognitive load condition
suggesting that automatic mood contagion underlies the SIE effect rather than mechanisms based on self-perception or
self-awareness. Appropriateness of a topic for complaining appeared a boundary condition of the SIE effect: When a topic
was considered by participants the most appropriate for complaining, the act of showing dissatisfaction with the topic led
to mood improvement.

Keywords: Affirmation, Complaining, Mood, Saying is experiencing effect

Saying is Experiencing: Affective Consequences gap in knowledge by presenting a line of four studies on
of Complaining and Affirmation affective consequences of complaining and affirmation.
Complaining is defined as expressing dissatisfaction
Complaining is more frequently heard than studied. independently of whether it is actually experienced or
Using a diary method with a sample of American students, not (Kowalski, 1996). Complaining is, then, an affect-
Alicke et al. (1992) found the average number of complaints expressive behavior, so it is logical to assume that it can
to exceed four per day per participant of their study. In other result in direct changes of affective states. However, it is
cultures complaining may be even more frequent. A recent not clear whether complaining leads to positive or negative
national survey conducted in Poland revealed that 40.5% changes in affective states. Theoretical arguments may
of that population believes Poles complain very often be developed for both improvement and deterioration of
and only less than one percent believes they do it never affective states after complaining.
or rarely (Wojciszke, Szymków-Sudziarska & Baryła, Alicke et al. (1992) and Kowalski (1996, 2003) believe
2008). At least in some contexts complaining may be also in cathartic function of complaining, that the expression of
important – Kelley (1979) found that it ranked third in a dissatisfaction leads to venting negative emotions, provides
list of 15 problems faced by romantic couples. Still, outside an emotional release from frustration and “gets it off one’s
specific areas of consumer complaints (e.g. East, 2000) and chest”. This assertion is based on what participants believe
hypochondria (e.g. Smith, Snyder & Perkins, 1983) only a to be the reason of their own complaining – the desire to
few empirical studies on complaining have been published vent frustration was the most frequently cited reason of
(Alicke et al., 1992; Kaiser & Miller, 2001; Kowalski & complaining by Alicke et al.’s participants of the diary
Cantrell, 2002). Despite a comprehensive theoretical model study (and by Wojciszke et al.’s respondents of a national
of antecedents, functions and consequences of complaining sample). Nevertheless, a subjective reason of a behavioral
developed by Kowalski (1996), empirical research on this act is evidently different from an objective effect of the act,
topic remains scarce. The present work attempts to fill this and the former cannot be considered a proof of the latter.

* Warsaw School of Social Sciences and Humanities, Sopot Campus, Poland

** email: [email protected]
74
Unauthenticated
Download Date | 4/22/17 3:50 PM
75 Bogdan Wojciszke et al.

Studies where affective states would be measured before conscious control. Because people are frequently motivated
and after complaining are needed to demonstrate the point, hedonistically (striving to remain in good mood or to
but to our knowledge such studies have not been published. achieve it if they are initially in bad mood, Larsen, 2000),
Experiments on effects of writing or talking about this suggests that the saying is experiencing effects would
traumatic emotional experiences seem to be the closest be in most situations restricted to affirmation or at least that
approximation of such studies. A number of experiments a mood improvement after affirmation should be stronger
showed that the mere act of disclosure has astonishingly than a mood deterioration after complaining.
large therapeutic effects – it improves not only affective The second mechanism is offered by the objective self-
states, but also betters an array of physiological and awareness theory (Wicklund, 1975) and its finding that
behavioral outcomes like the number of physician visits, the self-focused state increases the intensity of affects and
efficiency of the immunological system or reemployment emotions experienced during this state (Carver & Scheier,
following job loss (Davison, Pennebaker & Dickerson, 1981). As far as talking about one’s own negative or
2000; Pennebaker, 1997, Smyth, 1998). Interestingly, these positive opinions and emotional responses can be assumed
delayed and beneficial effects of emotional expression to be self-focusing, self-awareness may be responsible
are preceded by strong increases in distress immediately for mood decreases after complaining and its increases
after the expression. In all those experiments the topic after affirmation. This mechanism also presumes people’s
of disclosure involved deep emotional experiences awareness of what they are talking about (whether the
(usually traumatic) and emotional expression seemed tone is positive or negative), although it does not presume
necessary though not sufficient for the beneficial effects to the awareness of the fact that the tone of own utterances
materialize. What seems necessary is a transduction of the influences one’s own mood. However, the basic premises are
traumatic experience into linguistic structures that promote that (1) people actually feel bad when complaining or good
assimilation and understanding of the event. This requires when affirming and that (2) they focus on their own selves
a deep, thoughtful processing of the relevant memories or at least on own affects experienced during complaining
using what Pennebaker, Mayne and Francis (1997) or affirmation. Neither of these premises is necessarily
called causal words (because, reason) and insight words true in a society, where complaining is extremely frequent
(understand, realize). In this respect, most cases of (Poland) and in many situations expressing dissatisfaction
complaining are dissimilar to “emotional writing” as the seems to be a socially shared habit, requiring no preexisting
former involve mundane topics of low emotional intensity emotions nor conscious thoughts (Wojciszke et al., 2008).
and the typical act of complaining seems to be rather The extreme frequency and – presumably – the mindlessness
superficial than profound (Alicke et al., 1992; Wojciszke of complaining acts makes the self-focus an implausible
et al., 2008). explanation of the saying is experiencing effect.
Provided this superficial nature of complaining and The third mechanism is mood self-contagion based
the immediate affective aftermath of expressing negative on automatic links between the perception, action, and
emotions, we believe that complaining leads to immediate feelings (Chen & Bargh, 1999; Strack & Deutsch, 2004).
deterioration of mood, while an act of affirmation leads Numerous dual-process theories of social cognition assume
to mood improvement. So, we postulate a “saying is that information can be processed in two systems – one
experiencing” effect similar to changes in private opinions which requires mental capacity and motivation and bases
on a topic following own public utterances on the topic – on symbolically represented rules structured by language
the saying is believing effect (Higgins & Rholes, 1978). and logic, and another which functions automatically and
There are at least three mechanisms which may lead to the draws on associations that are structured by similarity and
saying is experiencing effect. contiguity (Smith & DeCoster, 2000). In their reflective
The first is offered by the self-perception theory (Bem, and impulsive model Strack and Deutsch (2004) extended
1972) which assumes that people infer their attitudes and this theorizing on behavior as well, postulating that the
preferences from their own overt behavior if the latter is reflective and impulsive systems regulate information
perceived as unconstrained by situational pressures. If processing and social behavior in parallel, although
people feel free to complain or affirm, they may infer from there is an asymmetry in the sense that the impulsive
their verbal behavior that they are in a bad or good mood system is always engaged, whereas the reflective system
and experience the mood accordingly. Inferring affective may be disengaged when cognitive capacity is lacking.
states from own behavior requires awareness of the The impulsive system can be oriented toward either
behavior in question, of its possible situational constraints approach or avoidance and numerous data shows that this
(or lack of them), and most probably also awareness of motivational orientation is elicited by the processing of
the relation between the two (Olson, 1992). In effect, if positive or negative information, experiencing of positive
induced by self-perception, any change in mood would be or negative affect, or executing of approach or avoidance
conscious and, therefore, at least potentially amenable to behaviors. The impulsive system functions according to a

Unauthenticated
Download Date | 4/22/17 3:50 PM
 Saying is experiencing: Affective consequences of complaining and affirmation
76
compatibility principle – the processing of information, us to discern between the mood contagion versus self-
the experience of affect and the behavior execution perception and self-awareness explanations of the basic
facilitate each other when they are compatible in valence. effect.
Accordingly, negative emotional expressions should Finally, in Experiment 4 we asked participants to
lead to negative affective states via activating automatic show their satisfaction or dissatisfaction with topics
associations, while positive expressions should lead they considered either right or wrong to complain about,
to positive affect even when people do not recognize according to what people in their immediate social milieu
consciously the sense of their expressions nor the thought and did. As elaborated later, some topics may be
expression-affective state links. This theorizing is supported seen as normative for complaining and it is possible, that
by numerous results showing that congruent affective complaining on such topics actually improves the affective
states follow expressive behavior even when this behavior state because it is a case of knowledge sharing and norm-
is not recognized as such. For example, Stepper and maintenance behavior. In other words, Experiment 4
Strack (1993) showed that people experience more intense looked for a boundary condition of the hypothesized saying
emotions after a success or failure when adopting an is experiencing effect.
upright or slumped posture under the pretext of studying
different working conditions. Similar effects were reported
for “subjectively non-emotional” facial expressions Experiment 1
(Strack, Martin & Stepper, 1988) and nodding or shaking
head movements (Förster & Strack, 1996). First two experiments were based on a similar design
The automatic mood contagion mechanism does not where mood was measured repeatedly before and after
discern between mood decreases after complaining and communication that was negative (complaining) or positive
mood increases after affirmation – both effects can be (affirmation) in tone. In Experiment 1 participants listened
expected to arise and equal in strength. This mechanism to communication of another person; in Experiment 2
allows also a prediction that it does not matter whether it is participants produced their own communication.
own or other person’s complaining or affirmation (because
affective codes are associated with perceptions and it is
irrelevant whether these are self-perceptions or perceptions Method
of others, cf. Dijksterhuis & Bargh, 2001). On the other
hand, self-perception and self-awareness accounts predict Participants and design. Sixty employees of a law firm
the saying is experiencing effect to be constrained to own (lawyers and paralegal workers, 33 men and 27 women,
complaining or affirmation. Finally, the mood contagion mean age = 27.88 years, SD = 3.76) participated in small
explanation predicts the saying is experiencing effect groups from three to five persons. The design was 2 (tone
to emerge also in conditions of cognitive load where of communication: complaining vs. affirmation) x 2 (time
participants are cognitively busy with a parallel task, of mood measurement: before vs. after communication)
while the two remaining explanations predict the effect to with repeated measurements on the second factor.
disappear in such conditions. Procedure and manipulation. Participants listened
to a seven-minute audiotaped story about a vacation in
Present Studies Canary Islands (a popular vacation place in Europe) with
Our first aim was to provide empirical support for the instruction to make an impression about the story teller.
the saying is experiencing effect, that is, to show that The story teller was a young woman (a paraprofessional
complaining decreases and affirmation increases the mood. actress) who presented a vivid and emotional account
To this effect we conducted two studies were mood was of either a very successful vacation (very good weather,
measured before and after affirmation or complaining. In excellent service, good company etc.) or a vacation which
Experiment 1 participants listened to audio-taped affirmation appeared unpleasant (harsh hot weather, poor service, and
or complaining of another person. In Experiment 2 they even poorer company etc.). Before and after listening to
affirmed or complained themselves by voicing their own the story, participants filled a short scale measuring their
opinions on topics which had prompted positive or negative mood. After listening to the communication, participants
statements in the majority of pilot participants. also showed their impressions of the story teller.
Experiment 3 was a conceptual replication of Measures. The “before” mood scale consisted of two
Experiment 2 with inclusion of cognitive load conditions positive and two negative general statements describing
where participants were asked to remember some content the participant’s present mood in general terms (e.g. I feel
while speaking. The purpose of this study was to check pretty good at the moment. I am in a bad mood now.).
whether the saying is experiencing effect emerges in Answers were given on scales ranging form 1 (doesn’t
conditions of scarce cognitive resources, thereby allowing describe my mood at all) to 7 (describes my mood very

Unauthenticated
Download Date | 4/22/17 3:50 PM
77 Bogdan Wojciszke et al.

well). Average agreement with the four statements (with 6

negatives inverted) was used as the mood index. The Complaining
“after” mood scale consisted of four similar statements that Affirmation
in previous research had been shown to be equivalent to the 5,8
first four set of statements in terms of means and standard
deviations. Both sets came from a general mood scale 5,6

Mood (1-7)
developed in Polish by Wojciszke and Baryla (2005). After
recoding negative statements, the mood measure appeared
internally consistent both in the before (Cronbach’s α = 5,4

.71) and after (α = .91) condition.

After listening to the communication participants 5,2
showed their impressions of the story teller on three –5
to 0 to 5 rating scales answering the question whether the
5
women was happy or unhappy (manipulation check) and
Before After
whether she was generally likeable and emotionally stable
Time of mood measurement
or not.
Figure 1. Mood before and after listening to other person’s complaining or affirmation
(Experiment 1).

Results vs. affirmation condition with the initial mood serving as

a covariate, and this analysis showed a highly significant
To check validity of the manipulation, the perceived effect of the condition, F (1, 57) = 10.65, p = .002, η2 =
happiness of the target woman was subjected to one-way .16.
ANOVA which yielded a strong effect of the communication
emotional tone, F (1, 58) = 575.71, p < .001, η2 = .91. Discussion
Whereas the woman describing a successful holiday (the
affirmation condition) was perceived as very happy (M = Clearly, listening to other person’s affirmation increased
4.43), the woman telling about the unpleasant vacation (the mood in the listeners while listening to complaining led to
complaining condition) was perceived as unhappy (M = mood deterioration. This is in line with previous findings
-3.63). This effect remained equally large after including that the vocal expression of emotions is able to induce
“before” and “after” mood indices as covariates, which corresponding feelings in listeners (Siegman & Boyle,
shows that the perception of complaining vs. affirmation 1993) and that other persons’ communication tone induces
is independent of the current mood of perceivers. The corresponding mood in listeners (Neumann & Strack, 2000).
affirming person was also perceived as much more likeable Although such “emotion contagion” effects are frequently
(M = 3.77) and emotionally stable (M = 2.23) than the explained in terms of perspective taking which is an
complaining one (M = -1.60 and M = -.07, respectively). attention-consuming activity (Wispé, 1986), Neumann and
The main analysis was performed on the mood Strack (2000) showed these effects also in a cognitive load
measure in a 2 (tone of communication: complaining condition suggesting the emotion sharing may be based on
vs. affirmation) x 2 (time of mood measurement: before automatic rather than controlled information processing.
vs. after communication) factorial design with repeated These results show also that complainers are perceived as
measurements on the second factor. This analysis revealed less emotionally stable and less likeable than persons who
the expected interaction between both factors, F (1, 58) affirm the same topics. This is in line with observations
= 11.94, p = .001, η2 = .17. No main effect appeared that complaining is an aversive behavior and complainers
significant. As can be seen in Figure 1, affirmation led are perceived in a negative way (Kaiser & Miller, 2001;
to significant increases in mood, and the before-after Kowalski, 2003; Kowalski & Cantrell, 1997).
difference was significant, t (29) = 2.62, p = .007 (one-tailed
test for dependent data). On the other hand, complaining
led to significant decreases in mood with the before-after Experiment 2
difference being significant, t (29) = 2.26, p = .016 (one-
tailed). In effect, the final mood was higher in the affirmation In this experiment mood was measured in a different way
than complaining condition, though there was not such a but also before and after a communication. This time,
difference in the initial mood. However, because the initial however, the negative or positive communication was
mood was slightly higher in the complaining condition, this produced by participants themselves.
difference could have contributed to the whole interaction.
Therefore, we performed also an analysis of covariance on
the final mood measure as a function of the complaining

Unauthenticated
Download Date | 4/22/17 3:50 PM
 Saying is experiencing: Affective consequences of complaining and affirmation
78
Method 0,6
Complaining
Participants and design. Fifty nine university students (8 0,4 Affirmation
men and 51 women, of mean age from 20 to 24) volunteered
to participate individually. The design was 2 (tone of

Mood (standardized)
0,2
communication: complaining vs. affirmation) x 2 (time of
mood measurement: before vs. after communication) with 0
repeated measurements on the second factor.
Procedure and manipulation. Participants were asked -0,2
to tell their opinions about four topics. For half participants
the topics were negative (prices, public transportation, -0,4
salaries, and the public health service), that is, they
instigated negative opinions in a majority of students as -0,6
found in several pilot studies. Another half of participants Before After
were randomly assigned to positive topics (holidays, pets, Time of mood measurement
Christmas, receiving gifts). A young male experimenter Figure 2. Mood before and after own complaining or affirmation (Experiment 2).
purportedly wrote down their opinions. Before and after
telling their opinions, participants filled a short scale
measuring explicit mood. In the final measurement a Results
measure of implicit mood was taken as well.
Measures. The “before” mood adjective check list The main analysis was performed on the mood
consisted of four positive (elated, peaceful, pleasant, index in a 2 (tone of communication: complaining vs.
easy-going) and four negative (tense, upset, worried, affirmation) x 2 (time of mood measurement: before vs.
disenchanted) mood descriptors and participants checked after communication) factorial design with repeated
the adjectives describing their current mood. The “after” measurements on the second factor. This analysis revealed
mood measure consisted of similar eight mood descriptors the expected interaction between the two factors, F (1,
(glad, optimistic, fine, relaxed, unpleasant, dejected, 57) = 7.32, p = .009, η2 = .11. As can be seen in Figure 2,
discontented, unsettled) which in previous research had been own affirmation led to significant increases in mood, and
shown to be equivalent to the first set in terms of proportion the before-after difference was significant, t (29) = 2.35,
of people checking them as describing their current mood in p = .013 (one-tailed test for dependent data). On the other
neutral (non-manipulated) situations (Wojciszke & Baryla, hand, complaining led to a marginally significant decrease
2005). The mood was scored as a number of positive minus in mood, t (28) = 1.66, p = .054 (one-tailed).
negative descriptors checked. Additionally, participants The analysis revealed also an unexpected main effect
rated their current mood on a graphic, continuous, 12-cm of the communication tone, F (1, 57) = 11.50, p = .001,
long scale anchored with statements “I feel bad” and “I feel η2 = .17, although it was constrained by the described
good” (the answer was scored in millimeters from the left interaction and the affirmation vs. complaining difference
end of the scale). Since the adjective measure correlated was not significant in the initial measurement, t (29) = 1.55,
with the mood ratings both before (r = .72) and after (r = p = .126 (two-tailed). Nevertheless, because participants
.56) communications, the two were standardized (to make assigned to the affirmation condition tended to be in a higher
them comparable) and averaged into a global mood index initial mood than participants assigned to complaining,
which served as the main dependent measure. we performed also an analysis of covariance on the final
A measure of implicit mood was devised after Rusting mood measure with the communication tone as a factor
and Larsen (1998) and consisted of five incomplete words, and the initial mood serving as a covariate. This analysis
each lacking one letter (e.g. JO_). Participant were asked to revealed a clear main effect of the communication tone, F
fill the gap with a letter which would make a sensible word. (1, 57) = 20.42, p = .0001, η2 = .27, despite the significant
In each case two fillings were possible – one which made contribution of the initial mood, F (1, 57) = 26.54, p =
a positive word (e.g. JOY) and one which made a neutral .0001, η2 = .32.
word (e.g. JOB). The number of positive fillings (varying Finally, the implicit mood measure (taken only after the
for 0 to 5) served as the measure of implicit mood. communication) was compared between the conditions,
revealing a significant difference, t (57) = 2.7, p = .022, d =
.54 with the ambiguous words being completed in positive
way more frequently in the affirmation (M = 1.72) than the
complaining (M = 1.21) condition. This moderately strong
effect remained virtually unchanged after inclusion of the
final (or initial) explicit mood measure as a covariate. This

Unauthenticated
Download Date | 4/22/17 3:50 PM
79 Bogdan Wojciszke et al.

suggests independence of the communication influences on an experimenter. For half participants the topics were
explicit and implicit measures of mood. Indeed, the two themes which invited predominantly negative opinions,
measures appeared unrelated, r (58) = .01, which is not an for another half – themes which invited positive or neutral
uncommon finding (Fazio & Olson, 2003). Also a mediation ones (as various pilot studies suggested). Cognitive load
analysis with the condition serving as an independent manipulation was crossed with the emotional tone of
variable, implicit mood as dependent, and explicit mood as the topics and introduced by asking half participants to
a mediator failed to show a significant mediation. remember the topics they spoke about. This manipulation
was shaped after a Gilbert, Pelham and Krull (1988)
who found that participants who had been to remember
Discussion themes on which a target person spoke (and became
this way “cognitively busy”) were less able to correct
Results of this study provide a clear support for the their impressions of the target for situational constraints
saying is experiencing effect. Whereas own affirmation and in effect fell prey to the correspondence bias. That
leads to mood improvement, own complaining leads to is, they attributed high anxiety to a target who spoke on
mood deterioration. Because these differences appear both embarrassing topics, ignoring the fact that at least part of
for explicit and implicit mood measures, it is an argument the target’s anxiety shown in nonverbal behavior (the basis
for the mood contagion rather than the self-perception or of participants’ perceptions) was due to the nature of the
self-awareness explanation of the saying is experiencing topics. The interesting part of this manipulation was that
effect. It seems that affirmation increases accessibility the to be remembered material consisted of the topics of
of positive thoughts, or complaining decreases this the target’s utterances (flashed on the screen on which the
accessibility, or both, although the present design is unable target’s nonverbal behavior was reproduced) – exactly the
to discern between these possibilities. Nevertheless, the information on the situational constraints which should
implicit mood measure points to mood contagion as the have been taken into account when inferring the target’s
more plausible explanation of the saying is experiencing anxiety. Still, what really counted was not the remembered
effect. content (which should reduce inferences of anxiety) but the
The same conclusion is suggested by the similarity of fact there was something to remember (i.e. a parallel task
the present results to the effects of listening to other person’s which disrupted the correction for situational influences on
complaining or affirmation found in Experiment 1. Only behavior). We used this subtle manipulation of taxing our
automatic mood contagion can parsimoniously explain the participants’ cognitive resources to see whether the “saying
similarity of affective consequences of the own and others’ is experiencing effect” will be found in such conditions.
complaining versus affirmation. It should be also noticed
that in both experiments the saying is experiencing effect
was equally strong for complaining and affirmation. This Method
also suggests that controlled processing strategies were not
involved in the mood changes because controlled mood Participants and design. Eighty persons (40 men and 40
processing tend to be hedonistic (i.e. mood improving – women, mean age = 44.61 years, SD = 13.03) agreed to
Larsen, 2000) and hedonistic self-regulation would have participate individually. The participants were passersby in
led to stronger mood improvement after affirmation than a public park (in summer) approached by a young female
mood deterioration after complaining. experimenter. The design was 2 (tone of communication:
complaining vs. affirmation) x 2 (cognitive load: load vs.
no load) x 2 (time of mood measurement: before vs. after
Experiment 3 communication) with repeated measurements on the last
factor.
In this experiment we introduced a cognitive load Procedure and manipulations. Participants were asked
manipulation to test whether the saying is experiencing to convey their opinions on nine topics. For half participants
effect will emerge in conditions where cognitive resources all topics were negative (salaries, prices, crime, politicians,
are scarce. This would be a strong argument for the health service, functioning of state administration, the
impulsive nature of the effect stalked here since both self- image of Poland in the world, environment pollution,
perception and self-awareness mechanisms require mental efficiency of the police – according to pilot studies they
capacity which is impaired in the double-task condition we instigated negative opinions in a majority of Poles). Another
used. half participants were randomly assigned to positive topics
To explore this way of thinking, we designed a study (holidays, gifts received, the best liked celebrities etc.)
where (under a pretext of recording spoken speech to be The experimenter purportedly recorded their opinions (“to
later analyzed for prosodic qualities) participants were analyze prosodic features of speech” which was introduced
asked to voice their opinion on nine topics showed by

Unauthenticated
Download Date | 4/22/17 3:50 PM
 Saying is experiencing: Affective consequences of complaining and affirmation
80
as a topic of the study) while showing slips of paper with 2,5
the topics that the participants were to talk about. Half Complaining
participants were randomly assigned to the cognitive load Affirmation
condition – they were asked to remember the topic they
talked about. The remainder did not receive this part of
2
instruction. At the end all participants were asked to recall

Mood (-4 to 4)
the topics they talked about, debriefed, and thanked for
their participation.
Measures. Before and after their communications
1,5
participants received short adjective lists to show their
present mood. The lists were identical to those used in
Experiment 2. The differences between the number of
positive and negative adjectives checked served as indices
1
of the initial and final mood.
No load Load
Cognitive load condition
Figure 3. Mood after own complaining or affirmation in cognitive load versus no
Results and Discussion load condition (Experiment 3).

Table 1
The main analysis was performed on the final mood Within-Cell Correlations (Pearson’s r) Between the Mood Measures
index in a 2 (tone of communication: complaining vs. and Topic Recall in Experiment 3.
affirmation) x 2 (load vs. no load) with the initial mood Recall and final mood Initial and final mood
serving as a covariate. This analysis revealed a significant
Complaining
influence of the initial on the final mood measure, F (1, 75)
No load .07 .60**
= 23.93, p < .001, η2 = .24, and a significant main effect of
the communication tone. The latter effect, however, was Cognitive load .02 .14
entirely constrained by the interaction between the tone Affirmation
and cognitive load, F (1, 75) = 4.71, p = .033, η2 = .06. No load -.10 .86***
As illustrated in Figure 3, the final mood (controlled for Cognitive load .08 .12
the initial mood) was significantly higher in the affirmation Total
than complaining condition, but this was true solely in the No load .00 .72***
cognitive load condition, F (1, 37) = 6.32, p = .016. In the
Cognitive load -.03 .04
no load condition the difference between complaining and
affirmation was not found, F < 1.
explanation assumes that: (1) the participants remembered
These results suggest that the communication tone
more topics under the memory instruction, i.e. in the load
influences own mood of the speaker also when the latter
vs. no load condition and that (2) the number of topics
is cognitively busy with other, parallel tasks. Clearly,
remembered correlated with the final mood – negatively in
unconstrained cognitive resources are not necessary for
the complaining condition and positively in the affirmation
one’s mood to follow the tone of one’s own utterances. This
condition, especially among the participants who were
speaks for the explanation of the “saying is experiencing”
burdened with the parallel task. Actually, none of these
effect in terms of mood self-contagion which can be
assumptions received support in the data.
automatic, rather than in terms of self-perception or self-
A 2 (load vs. no load) x 2 (complaining vs. affirmation)
focus which probably need conscious inferences and,
analysis of variance on the number of topics remembered
therefore, cognitive resources.
revealed no main effect of the load (i.e. memorizing
However, provided that the parallel task in this
instruction), F (1, 76) = 1.54, p = .22, nor any other
experiment involved memorizing the topics of own
significant effect. Participants remembered on the average
statements, an alternative explanation of our results is
5.39 and 4.90 topics in the no load and load condition,
plausible: May be the differences in the final mood are not an
respectively. Also the analysis of within cell correlations
effect of cognitive load, but rather, they are due to increased
between the number of topics remembered and the final
accessibility of the topics inducing negative or positive
mood measure yielded null results. As can be seen in
thoughts. If under the memory instruction people keep in
Table 1, the correlation did not even approach significance
mind a greater amount of negative topics (complaining
in any of the four cells.
condition) or positive ones (affirmation condition), this may
This does not mean that no interesting effects of the
make their final mood decreased or increased, accordingly
load manipulation were found. As also illustrated in
(compared to the no load condition). This alternative
Table 1, strong correlations were found between the initial

Unauthenticated
Download Date | 4/22/17 3:50 PM
81 Bogdan Wojciszke et al.

and final mood measurements in the no load conditions, et al., 1992). Although it may be the safest and easiest
showing a strong continuity of mood, the interjecting response on the side of a listening person, agreement can be
utterance of opinions notwithstanding. These correlations, also rewarding for the speaker because it conveys support
however, were absent in the cognitive load conditions – for the speaker’s opinions which is a case of a more general
attempts to memorize topics while voicing own opinions phenomenon of social reality sharing. Hardin and Higgins
on them erased totally the influence of the initial on the (1996) argue that socially shared reality serves not only the
final mood resulting in discountinuity of the affective state. epistemic function of establishing the reliable and valid
Mood represents one’s current life situation – whether representation of the world, it also fosters interpersonal
it is welcoming and poses no threat or is problematic trust and reliance on each other’s view of the world. If
(Robinson, 2000; Schwarz, 1990). Much data suggests met with agreement, complaining may be a rewarding
people automatically evaluate current stimuli (Bargh, experience despite the immediate and probably short-
Chaiken, Raymond & Hymes, 1996; Duckworth, Bargh, lived mood deterioration. Such agreement needs not be
Garcia & Chaiken, 2002) and even appraisal theorists of actually received because sometimes a strong expectation
emotion assume that appraisal processing tends to occur of agreement may be enough to experience social sharing.
automatically (Clore, 1994; Lazarus, 1995). The function This may be the case when people voice (negative) opinions
of subjective affect is to inform the conscious mind about consistent with a norm strongly shared by their audience.
the results of unconscious appraisal (Robinson, 2000). To test this line of reasoning we devised a study
Probably, our participants in the no load conditions were where we asked our participants to list up to five topics
aware of influence of the communication tone on their they considered either right or wrong to complain about,
mood (manipulation of the communication valence was according to what people in their immediate social milieu
quite blatant as it involved 8 negative or positive topics) thought and did. Then they were asked to choose the most
and they made an effort to neutralize this influence. This right or wrong topic and to write down their own opinions on
correction of affective influences irrelevant to the current the topic while being randomly assigned to the complaining
life situation assured mood continuity. Cognitively busy or affirmation condition (i.e. they were asked for clearly
participants had no mental resources to manage their mood, negative or positive opinions about the topic). This way,
so it remained a direct function of current affective states half participants were prompted to show satisfaction or
induced by the experimental manipulation. dissatisfaction with a topic which was strongly normative
for complaining, whereas other half made the same with
a topic which was counter-normative for complaining.
Experiment 4 Although numerous results suggest that complaining is
a general norm in Poland (especially when talking about
Having the saying is experiencing effect established, public and general matters, Wojciszke, 2004), normativity
we turned in the last study to its possible boundary was established individually for each participant. Several
conditions. Both American (Alicke et al., 1992) and Polish measures were taken, most importantly a measure of mood
(Wojciszke et al., 2008) participants widely believe that before and after affirmation or complaining. Our main
complaining has beneficial affective consequences, that it expectation was that voicing dissatisfaction with a topic
allows venting frustrations, leads to mood improvement most normative for complaining will actually lead to an
and that people engage in complaining to achieve these increase in mood, that is, to an inversion of the typical
effects. Three studies presented so far suggest that people saying is experiencing effect.
are wrong in their commonsense belief in beneficial effects
of complaining. Nevertheless, it cannot be excluded that
people are at least sometimes right, that in some conditions Method
complaining really leads to mood improvements and
may be people overgeneralize their experience from such Participants and design. Ninety-nine employees of
situations and believe that any venting of dissatisfaction has a telecommunication firm (35 men and 64 women, mean
beneficial effect. So, the question is in what kind of situations age = 27.20 years, SD = 3.74) participated individually
complaining can lead to actual mood improvements. or in small groups. The basic design was 2 (tone of
Complaining is by definition a social activity – it means communication: complaining vs. affirmation) x 2 (topic
expressing dissatisfaction to other people. So it involves normativeness for complaining: high vs. low) x 2 (time of
social context which was ignored in our studied reported mood measurement: before vs. after communication) with
so far. Responses of listeners are probably one important repeated measurements on the last factor.
contextual factor. Showing agreement with the complaining Procedure and manipulations. Participants were asked
person is the most frequent response of the audience to to jot down up to five topics which were either right or
everyday complaints (35% of responses as found by Alicke wrong to complain about according to what surrounding

Unauthenticated
Download Date | 4/22/17 3:50 PM
 Saying is experiencing: Affective consequences of complaining and affirmation
82
people typically thought and did. This way topics normative 6,5
or counter-normative for complaining were established Complaining
individually for each participant. Next, participants were Affirmation
asked to choose the topic which was the most appropriate 6
or inappropriate for complaining and to write down
their personal opinions about it. The opinions were to be

Mood (1 to 7)
either negative or positive. In this way participants were 5,5
randomly assigned the communication tone (complaining
vs. affirmation) condition. Before and after jotting down
their opinions, participants filled a short scale measuring 5
their mood.
Measures. The “before” and “after” measures of
mood were four-item, equivalent, scales (ranging from 1 4,5
to 7) identical to those used in Experiment 1. Both scales Counter-normative Normative
appeared reliable with Cronbach’s αs mounting to .92 and Topic normativeness for complaining
.88 in the before and after condition. Figure 4. Mood after own complaining or affirmation on topics normative or counter-
To the end of experiment, participants were asked to normative for complaining (Experiment 4).
rate the intensity of several emotions experienced at the
94) = 6.81, p = .011, η2 = .07. As can be seen in Figure 4,
moment. Six of them tapped affective positivity-negativity
when the topic was counter-normative, mood after
(self-content, slightly disheartened, sad, self-confident,
communication was marginally higher in the affirmation
optimistic, joyful) and four of them tapped tension emotions
than complaining condition, F (1, 45) = 2.89, p = .096, η2
(slightly embarrassed, slightly tensed, ashamed, peaceful).
= .06 (two-tailed) replicating the saying is believing effect.
All emotional states were described on scales form 1 (not
However, when the communication topic was normative
at all) to 7 (very much so), the averaged items served as
for complaining, the final mood (controlled for the initial
the indices of emotion positivity (α = .84) and emotional
mood) was significantly higher in the complaining than
tension (α = .80).
affirmation condition, F (1, 48) = 4.19, p = .046, η2 = .08.
This was the expected inversion of the saying is believing
effect.
Results
Emotions. Similar analyses were performed on the
emotion positivity index, revealing a tone by normativeness
Manipulation check. The content of opinions written
interaction, similar in shape to that depicted in Figure 4,
down by the participants was given to three independent
although weaker and barely significant, F (1, 94) = 3.98,
raters (senior students of psychology) who were blind
p = .049, η2 = .04. Simple effect analysis (with the initial
to conditions and rated the opinions for negativity-
mood serving as a covariate) showed that when the topic
positivity on a scale ranging from –5 (strongly negative)
was normative for complaining, emotional positivity was
to 0 (neutral) to 5 (strongly positive). The raters appeared
marginally higher in the complaining than affirmation
highly consistent (mean rho correlation was = .93) and
condition, F (1, 48) = 3.38, p = .072. However, these two
their ratings were averaged yielding a positivity index that
conditions did not differ when the topic was counter-
served as a manipulation check. The index was subjected to
normative for complaining (Ms were 4.60 and 4.81 for
a 2 (complaining vs. affirmation) x 2 (topic) ANOVA. This
complaining and affirmation respectively).
analysis revealed only one significant effect – the main
Finally, a similar two-factorial ANOVA performed on
effect of the communication tone, F (1, 95) = 170.03, p
the emotional tension index showed only a marginally
< .001, with participants instructed to complain presenting
significant tone by topic interaction, F (1, 94) = 2.97, p =
strongly negative opinions (M = -3.06) and participants
.088, η2 = .03. When participants complained on a topic
instructed to affirm presenting clearly positive ones (M =
that was counter-normative for complaining they tended to
2.48). This effect was equally strong in the normative and
report a marginally higher level of emotional tension (M =
counter-normative topic condition.
2.88) than in the remaining conditions, which varied from
Mood. The main analysis was performed on the final
2.34 to 2.45. This is in line with the idea that compared to
mood index in a 2 (tone: complaining vs. affirmation)
norm-consistent action, counter-normative behavior results
x 2 (topic: normative vs. counter-normative) with the
in an increased tension (Festinger, 1957, Matz & Wood,
initial mood serving as a covariate. This analysis revealed
2005; Rudman & Fairchild, 2004).
a significant influence of the initial on the final mood
measure, F (1, 94) = 152.80, p < .001, η2 = .62, and a cross-
over interaction between the tone and normativeness, F (1,

Unauthenticated
Download Date | 4/22/17 3:50 PM
83 Bogdan Wojciszke et al.

Discussion message on this person and change their own memories and
beliefs on this person accordingly. Participants convinced
The present results show a boundary condition for that the audience dislikes the target tend to produce a
the saying is experiencing effect. When the topic of negative message on this person and follow the message in
communication is seen as subjectively normative for their private opinion on this person.
complaining, showing dissatisfaction with the topic Despite obvious similarities between the two effects
actually increases mood of the communicator even when (both are biases resulting from own communications), there
the communication is negative in tone. This result helps are also important differences between them. Whereas the
to explain why people widely believe in beneficial effects saying is believing effect refers to evaluative judgments
of complaining – because sometimes complaining indeed and involves stable changes in opinion and memory, the
increases mood (when the topic is highly normative for saying is experiencing effect refers to mood and involves
dissatisfaction), people may overgeneralize this to situations only transitory changes. Most importantly, however, the
where complaining actually decreases mood. mechanisms of the two effects differ which shows their
divergent nature. While the saying is experiencing effect
stems from the impulsive mood contagion, the saying is
General Discussion believing effect draws on reflective reasoning and depends
on the communicator’s conviction of sharing reality with the
The presented line of studies is the first empirical audience. Echterhoff et al. (2005) showed that changes in
demonstration of the saying is experiencing effect – that beliefs following own communication tuned to an audience
negative or positive emotional communications lead to disappear when the audience consists of out-groups (who
corresponding changes in the mood of the speaker (and do not share reality with the communicator) and that these
a listener). Complaining leads to decreases in mood changes are mediated by the communicators’ trust in their
while affirmation leads to immediate increases in mood. audience’s ability to judge people.
Moreover, this effect does not require cognitive resources The present results suggest that affective consequences
(as Experiment 3 showed), it is observed with both explicit of own communications do not require the experience
an implicit (Experiment 2) measures of mood, it seems of sharing reality with the audience as a prerequisite.
to be equally strong for affirmation and complaining Nevertheless, such experience may play an important
(Experiments 1 and 2), and it emerges after listening to role, as suggested by the boundary condition of the saying
communications of others (Experiment 1) as well as is experiencing effect found in Experiment 4. When
own communications (Experiments 2 and 3). All these participants showed their dissatisfaction with a topic they
additional findings consistently suggest that automatic considered most appropriate for complaining in their social
mood contagion (Strack & Deutsch, 2004) lays at the milieu, their mood actually increased. This suggests that
crux of the saying is experiencing effect, rendering two the saying is experiencing effect may be actually inversed
other explanations of the effect improbable. These two in social contexts highly conducive to complaining.
alternative explanations delineated in the introduction – Exploration of the ways social context influences affective
self-perception and self-awareness – assume the saying is consequences of affirmation and complaining seems to be
experiencing effect to draw on highly conscious information a promising avenue for further research on the saying is
processing which should be impaired in the cognitive load experiencing effect.
condition. Neither the self-perception nor self-awareness
account predicts changes in mood after listening to
other person’s communications and both these accounts References
suggest (due to hedonistic mood regulation) that affective
consequences of the emotionally laden speaking should be Alicke, M. D., Braun, J. C., Glor, J. E., Klotz, M. L., Magee, J., Sederholm,
H., Siegel, R. (1992). Complaining behavior in social interaction.
greater for affirmation than complaining. These results do
Personality and Social Psychology Bulletin, 18, 286-295.
not exclude a role of conscious self-inferences in the saying Bargh, J. A., Chaiken, S., Raymond, P., & Hymes, C. (1996). The automatic
is experiencing effect, but they obviously show that such evaluation effect: Unconditional automatic attitude activation with
exercises are unnecessary for the effect to emerge. a pronunciation task. Journal of Experimental Social Psychology,
We likened the present saying is experiencing effect to 32, 104-128.
Bem, D. J. (1972). Self-perception theory. Advances in Experimental
the classical saying is believing effect showed by Higgins
Social Psychology, 6, 1-62.
and associates (Echterhoff, Higgins & Groll, 2005; Higgins Carver, C. S., Scheier, M. F. (1981). Attention and Self-regulation. New
& Rholes, 1978). Saying is believing effect refers to changes York: Springer-Verlag.
in opinion on a person resulting from own communication Chen, M., & Bargh, J. A. (1999). Consequences of automatic evaluation:
tuned to the audience. Participants convinced that the Immediate behavioral predispositions to approach and avoid the
stimulus. Personality and Social Psychology Bulletin, 25, 215-224.
audience likes the target person tend to produce a positive
Clore, G.L. (1994). Why emotions are never unconscious. In P. Ekman

Unauthenticated
Download Date | 4/22/17 3:50 PM
 Saying is experiencing: Affective consequences of complaining and affirmation
84
& R.J. Davidson (Eds.), The nature of emotion: Fundamental Olson, J. M. (1992). Self-perception of humor: Evidence for discounting
questions (pp. 285 - 290). New York: Oxford University Press. and augmentation effects. Journal of Personality and Social
Davison, K. P., Pennebaker, J. W., & Dickerson, S. S. (2000). Who talks: Psychology, 62, 369-377.
The social psychology of support groups. American Psychologist, Pennebaker, J. W. (1997). Writing about emotional experiences as a
55, 205-217. therapeutic process. Psychological Science, 8, 162-166.
Dijksterhuis, A., & Bargh, J. A. (2001). The perception-behavior Pennebaker, J. W., Mayne, T., & Francis, M. (1997). Linguistic predictors
expressway: Automatic effects of social perception on social of adaptive bereavement. Journal of Personality and Social
behavior. Advances in Experimental Social Psychology, 33, 1-40. Psychology, 72, 863-871.
Duckworth, K. L., Bargh, J. A., Garcia, M., & Chaiken S. (2002). The Robinson, M. D. (2000). The reactive and prospective functions of mood:
automatic evaluation of novel stimuli. Psychological Science, 13, Its role in linking daily experiences and cognitive well-being.
513-519. Cognition and Emotion, 14, 145-176.
East, R. (2000). Complaining as planned behavior. Psychology and Rudman, L. A. & Fairchild, K. (2004). Reactions to counterstereotypic
Marketing, 17, 1077-1095. behavior: The role of backlash in cultural stereotype maintenance.
Echterhoff, G., Higgins, E. T., & Groll, S. (2005). Audience-tuning effects Journal of Personality and Social Psychology, 87, 157-176.
on memory: The role of shared reality. Journal of Personality and Rusting, C. L. & Larsen, R. J. (1998). Personality and cognitive processing
Social Psychology, 89, 257-276. of affective information. Personality and Social Psychology
Fazio, R. H., & Olson, M. A. (2003). Implicit measures in social cognition Bulletin, 24, 200-213.
research: Their meaning and use. Annual Review of Psychology, 54, Schwarz, N. (1990). Feelings as information: Informational and
297-327. motivational functions of affective states. In E. T. Higgins &
Festinger, L. (1957). A Theory of cognitive dissonance. Evanston, IL: R. Sorrentino (Eds.) Handbook of motivation and cognition:
Row, Peterson, & Co Foundations of social behavior, Vol. 2 (pp. 527-561). New York:
Főrster, J., & Strack, F. (1996). The influence of overt head movements Guilford Press.
on memory for valenced words: A case of conceptual-motor Siegman, A. W., & Boyle, S. (1993). Voices of fear and anxiety and
compatibility. Journal of Personality and Social Psychology, 71, sadness: The effects of speech rate and loudness on fear and anxiety
421-430. and sadness and depression. Journal of Abnormal Psychology, 102,
Gilbert, D. T., Pelham, B. W., Krull, D. S. (1988). On cognitive busyness: 430-437.
When person perceivers meet persons perceived. Journal of Smith, E. R., & DeCoster (2000). Dual process models in social and
Personality and Social Psychology, 54, 733-740. cognitive psychology: Conceptual integration and links to
Hardin, C. D., & Higgins, E. T. (1996). Shared reality: How social underlying memory systems. Personality and Social Psychology
verification makes the subjective objective. In R. M. Sorrentino Review, 4, 108-131.
& E. T. Higgins (Eds.) Handbook of motivation and cognition: Smith, T. W., Snyder C. R., Perkins, S. C. (1983). The self-serving
The interpersonal context (Vol. 3, pp. 28-84). New York: Guilford functions of hypochondriacal complaints: Physical symptoms as
Press. self-handicapping strategies. Journal of Personality and Social
Higgins, E. T., & Rholes, W. S. (1978). “Saying is believing”: Effects of Psychology, 44, 787-797.
message modification on memory and liking for the person described. Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome
Journal of Experimental Social Psychology, 14, 363-378. types, and moderating variables. Journal of Clinical and Consulting
Kaiser, C. R., & Miller, C. T. (2001). Stop complaining! The social costs Psychology, 66, 174-184.
of making attributions to discrimination. Personality and Social Stepper, S., Strack, F. (1993). Proprioceptive determinants of affective
Psychology Bulletin, 27, 286-295. and nonaffective feelings. Journal of Personality and Social
Kelley, H. H. (1979). Personal relationships: Their structure and Psychology, 64, 211-220.
processes. New York: Wiley & Sons. Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants
Kowalski, R. M. (1996). Complaints and complaining: Functions, of social behavior. Personality and Social Psychology Review, 8,
antecedents, and consequences. Psychological Bulletin, 119, 220-247.
179-196. Strack, F., Martin, L. L., Stepper, S. (1988). Inhibiting and facilitating
Kowalski, R. M. (2003). Complaining, teasing and other annoying conditions of human smile: An unobtrusive test of the facial
behaviors. New Haven: Yale University Press. feedback hypothesis. Journal of Personality and Social Psychology,
Kowalski, R. M., & Cantrell, C. C. (2002). Intrapersonal and interpersonal 54, 768-777.
consequences of complaints. Representative Research in Social Wicklund, R. A. (1975). Objective self-awareness. Advances in
Psychology, 26, 26-33. Experimental Social Psychology, 9, 233-275.
Larsen, R. J. (2000). Toward a science of mood regulation. Psychological Wispé, L. (1986). The distinction between sympathy and empathy: To call
Inquiry, 11, 129-141. forth a concept, a word is needed. Journal of Personality and Social
Lazarus, R.S. (1995). Vexing research problems inherent in cognitive- Psychology, 30, 314-321.
mediational theories of emotion - and some solutions. Psychological Wojciszke, B. (2004). The negative social world: Polish culture of
Inquiry, 6, 183 - 196. complaining, International Journal of Sociology, 34, 38-59.
Matz, D.C., & Wood, W. (2005). Cognitive Dissonance in Groups: The Wojciszke, B., Baryła, W. (2005). Skale do pomiaru nastroju i szesciu
Consequences of Disagreement. Journal of Personality and Social emocji. [Scales for measuring mood and six emotions] Czasopismo
Psychology, 88, 22-37. Psychologiczne, 11, 31-47.
Neuman, R., & Strack, F. (2000). “Mood contagion”: The automatic Wojciszke, B., Szymkow-Sudziarska, A., & Baryla, W. (2008). Rules
transfer of mood between persons. Journal of Personality and and Norms of Conversation: The Case of Compatibility and
Social Psychology, 79, 211-223. Complaining. Unpublished manuscript.
Nyer, P. U. (2000). An investigation whether complaining can cause
increased consumer satisfaction. Journal of Consumer Marketing,
17, 9-19.

Unauthenticated
Download Date | 4/22/17 3:50 PM