Motivation and Emotion, Vol. 7, No.

1, 1983

Participant Descriptions of Guilt and Shame 1

Frank W. Wicker, 2 Glen C. Payne, and Randall D. Morgan
The University o f Texas at Austin

The purpose o f this research was to see if naive raters could distinguish
between guilt and shame in ways consistent with the descriptions o f emotion
theorists. In two studies, 152 participants recalled occasions on which they
had experienced guilt or shame and rated these experiences on a large
number o f scales that represented either basic dimensions o f emotion or
attributes previously postulated to differentiate between these two
emotions. Shame and guilt situations differed on a number o f attributes,
including felt powerfulness, self-control, self-consciousness and exposure,
activity, inferiority, surprise, alienation from others, facial sensation, self-
attribution of justice, and expectation o f punishment. Many commonalities
in the meaning o f the two concepts were also suggested, most importantly in
terms o f basic attributes such as pain, tension, and arousal. Results were
consistent with several previous accounts o f the essential differences
between guilt and shame, but not with all such descriptions.

In this research we were concerned, first, with attempts to establish criteria

for differentiating the emotions of guilt and shame. The issue is relevant for
mental health professionals. It has been claimed that a clear
differentiation of these two terms could greatly facilitate our understanding
of the dynamics of neurotic symptoms as well as of general emotional
development (Lewis, 1971; Piers & Singer, 1953). A second concern of this
research was with whether "naive" raters would distinguish between guilt
and shame in the same ways as the professionals who have written on this
distinction.

1The authors would like to thank Maria Emanuele, Sang Min Lee, Joe Kahler, and Frank
Lambert for their assistance with this study.
2Address all correspondence to Frank W. Wicker, Department of Educational Psychology,
University of Texas, Austin, Texas 78712.
25
0146-7239/83/0300-0025503,00/0 © 1983PlenumPublishingCorporation
26 Wicker, Payne, and Morgan

Achieving precise, agreed-upon definitions is one of the difficult on-

going tasks for emotion theorists. Over the years there has been consider-
able disagreement about the meaning of emotion terms, and guilt and
shame have not been exceptions. Some authors appear to have used the two
terms almost interchangeably, others have treated them as but different
aspects of the same underlying emotion, while others have taken pains to
pinpoint what they see as the essential differences between them (Izard,
1977).
Many differences have been proposed. There has been some overlap
of opinion among theorists, even though agreement is by no means
universal. In general, guilt is said to follow from acts that violate ethical
norms, principles of justice, feelings of responsibility, religious codes, or
moral values, whereas shame follows from acts that lack propriety,
appropriateness, or adequacy. Piers and Singer (1953) proposed that
violations of the limits of the superego produce guilt, whereas failure to live
up to the ego ideal produces shame. The two emotions are then opposite in
the sense that in one case (shame) a goal is not reached, while in the other
case (guilt) a boundary is exceeded.
Somewhat suprisingly, it has been argued that shame is the stronger
emotion, perhaps because ego ideals can be more basic or powerful than the
prohibitions of conscience (Lynd, 1958). Lynd asserts that shame is more
likely to be a diffuse feeling involving the whole self, even though the
occasion for shame may appear slight to the unemotional observer: "Men
blush for their crimes much less than for their weaknesses or vanity. "3
Shame is said to be a greater threat to self-esteem and personal identity
(e.g., Izard, 1977; Lynd, 1958). Pride is therefore "the shield against shame"
(Tomkins, 1963, p. 133).
Numerous proposals imply that shame is more overpowering or inca-
pacitating. It is incongruous, surprising, and bewildering (Lynd, 1958); it
causes one to lose control, to feel powerless and externally controlled
(Erikson, 1963); it destroys our ability to think logically and act efficiently
(Darwin, 1872; Lewis, 1979); it destroys our trust in ourselves and
everything else (Lynd, 1958). The "objective" character of guilt, on the
other hand, and the accompanying feeling of personal responsibility,
should lead to active, self-controlled attempts to restore or make amends.
Likewise, it is claimed that shame will be more often accompanied by
heightened self-consciousness, self-imaging, and a "sense of exposure"
(e.g., Erikson, 1963; Lewis, 1971). This self-consciousness is described as a
two-edged sword: It involves (1) greater focus on the self4 but also (2)

3Quoted by Lynd from Pere Goriot, by Honore de Balzac.

4Lewis (t971) argued that the direct object of shame is the self, whereas the direct object
of guilt is the thing done.
Ratings of Guilt and Shame 27

greater concern with "others" who are observing and evaluating. In support
of point 2, Lewis (1979) emphasized the dependency of shame on caring for
the evaluating other and on "the vicarious experiences of other's negative
evaluations" (p. 381). To Ausubel (1955) shame may be a component of
guilt involving external sanctions and evaluations rather then independent
self-reactions. Sartre (1956) proposed that the way one appears in the eyes
of others is more basic to shame than the way one appears in one's own
eyes. There is some disagreement, however, about whether such other-
orientation is more characteristic of shame than of guilt. Lynd (1958)
stated that it is the exposure of oneself to oneself that is crucial in shame;
Rawls (1963) argued that closeness to others intensifies guilt too; and De
Rivera (1977) proposed that guilt and shame are both reactions to an
implicit other. Piers and Singer (1953) maintained that shame may be
internalized and guilt may have an implicit audience.
If closeness to others is a more important antecedent of shame, it is
possible that alienation from others is a more likely consequence of the pain
that comes from exposure to their negative evaluation. Thus, shame may be
more of an isolating experience (e.g., Lynd).
Corresponding to these differences are differences in the predicted
consequences of the two emotions. With guilt one expects punishment and
attempts to get forgiveness, to restore prior relationships, to confess, to
atone. With shame one expects abandonment and attempts to change the
self, to hide, or to run away (Morris, 1971).
Several authors have suggested that guilt is more long-lasting, because
it does not subside until reconcilation, because it may be followed by
laborious attempts to sort out the ethical issues involved, and because it
may be sustained by unconscious gratification from the implicit self-
statement "I am still a good person because I feel guilty about what I did"
(e.g., Lewis, 1979).
Kemper (1978) discussed emotions in terms of a model emphasizing
felt power and felt status. One of his suggestions is that guilt tends to occur
when one senses that one has too much power, and shame occurs when one
senses excess status.
Differences related to bodily expression and bodily awareness have
been proposed. It is said that one blushes in shame and feels a hot face, but
the guilty face is heavy. There is a greater bodily awareness in shame and
less distinctive facial expression in guilt (Izard, 1977).

PRESENT STUDY

The present study was an attempt to get untrained observers to tell us

whether the terms guilt and shame mean different things to them and to
28 Wicker, Payne, and Morgan

compare their descriptions to those in the literature. So stated, this appears

to be asking too much of naive observers. The method employed in this
study, however, is based on the assumption that "people usually find it easy
to tell the difference between these two emotions in their personal
experiences [even though] they may not find it easy to describe the
differences" (Izard, 1977, p. 422). Participants were asked to recall a
previous experience in which they had felt each of these emotions and to
rate the experience on several scales. The scales were based on the
hypothesized differences reviewed in the above introduction. Thus, subjects
were required only to select and rate particular emotional experiences, but
average ratings of the two types of experiences could then be compared?
This approach is consistent with the prototype theory of concept forma-
tion, the view that real-life concepts often "have fuzzy boundaries but ideal
examples" (Wickelgren, 1979, p. 297). We assumed that even though our
subjects could not give formal definitions of emotion concepts in terms of
attributes and rules, they could recall representative exemplars or
prototypes that could then be rated, one attribute at a time. Recalled
prototypes should be ones that have much personal significance for them,
so their ratings may be helpful in the search for a clear differentiation of
guilt and shame.

EXPERIMENT1

Method

Subjects. Thirty-one students of introductory educational psychology

at the University o f Texas at Austin participated in this study. Such par-
ticipation was one o f their options for meeting a research requirement in the
introductory course.
Materials and Procedure. Subjects worked in groups o f eight or fewer
in a small laboratory room. Upon entering, they were given a rating
booklet, which they were not required to sign. At the top of each page o f the
booklet was a noun indicative of an emotion. Six such nouns were presented
in a randomly different order for each subject: Guilt, Shame, Fear,
Anxiety, Hope, and Serenity. For each noun, subjects were asked to recall
an event in their own experience and to write a brief description of that
event in a space provided under the noun. If they felt uncomfortable about
describing a particular event, they were allowed to write cryptic notes, write

5This approach is similar to use of the DifferentialEmotions Scaleby Izard and his colleagues.
Ratings of Guilt and Shame 29

descriptions on a separate blank sheet that they could take away with them,
or write nothing at all. Next, they rated the recalled experience on each of 34
9-point rating scales. Scales had been chosen in the attempt to represent
feeling, evaluation, motivational-, and situational-appraisal aspects of
emotion, and to reflect dimensions emphasized by the theoretical
differentiations of guilt and shame discussed earlier. Ratings were given by
checking one of nine spaces on bipolar scales. The scales, anchored by
contrasting expressions such as I felt pain versus I felt pleasure, can be seen
in Table I. It was emphasized that a neutral value on this underlying
dimension was to be represented by the fifth space on the scale. If subjects
could not recall an appropriate experience, they were asked to imagine one,
and to rate all experiences on all of the scales. They were encouraged to
work rapidly but accurately. Instructions also included an explanation of
some of the more difficult scales.
For approximately half the subjects, scales were presented in the order
given in Table I. The left-right orientation of anchoring points and the
order of presentation of the scales were reversed for the other half.

Results

Table I gives means and standard deviations for Guilt and Shame on
each scale. For descriptive purposes, a t test was computed to compare each
mean to the constant neutral value of 5 (McNemar, 1955, pp. 107-108), and
the resulting significance levels are indicated with each mean. F ratios for
comparisons between Guilt means and Shame means are presented only
where p < .10. Table I suggests more similarities than differences between
the images evoked by these two nouns, but there are suggestive differences
too. Both guilt and shame situations were judged as painful, tense, arousing
but inhibiting, and novel but not imaginary, with a known cause. People
felt they should be punished because they had done the wrong thing, even
an unjust thing, and were not living up to their ideal. Although their lives
had previously been going very well, they now felt conflicted, self-
conscious, and lacking in self-confidence. With heavy faces they anticipated
the frowns of others. Note, however, that even though both situations
produced a self-conscious or "exposed" feeling, the shame situation
produced it to a higher degree, along with a greater tendency to feel low
power, low status, lack of control, and passivity (although raters expressed
activity in the guilt situation more clearly than they expressed passivity in
the shame situation). Two marginal differences, indicating greater sub-
missiveness and less self-confidence with shame, are consistent with this
general picture of greater helplessness in the shame situation. The two
emotions were not different in terms of concern with other-evaluation
Table I. Mean Scores for Guilt and Shame on Each Scale, Significance of Differences from Neutral Score, F and P for Difference
Between Guilt and Shame Means, and Means for Serenity and H o p e - - E x p e r i m e n t 1
Guilt Shame
Serenity Hope
Scale Mean SD Mean SD F p mean mean
I felt pleasure (vs. pain) 2.68 b 1.81 2.84 b 2.18 8.52 6.16
The situation was familiar (vs. novel) 3.29 b 2.55 3.19 b 2.46 5.23 4.84
The emotion began suddenly (vs. gradually) 6.48 b 2.29 5.65 3.04 4.10 4.61
I felt extremely self-confident and capable 3.77 ~ 2.63 2.65 b 1.96 3.67 < 10 7,65 5.74
(vs. not at all)
I felt tense (vs, relaxed) 8.06 b 1.29 7.94 ~ .81 1.52 5.74
My face was not at a / / h o t and flushed (vs. extremely) 4.45 2.35 4.61 2.38 7.29 6,29
I was dominant (vs. submissive) 5.23 2.79 4.10 ~ 2.29 3.69 < 10 4.71 4.84
The emotion lasted a short time (vs, a long time) 4.55 2.78 3,32" 2.10 5.74 < 05 3.00 3.13
I was more concerned with how others viewedme (vs. I
viewedmyselJ) 5.90 2.96 5.90 2.83 3.90 5.16
Prior to this event, things in my life had been going
well (vs. poorly) 7,58 b 1.91 7.32 b 2.06 6.90 7.48
I felt I did the wrong thing (vs. right thing) 7.00 b 2.34 7.00 b 2.11 2.00 2.74
The emotion was caused more by my imagination (vs.
things that actually happened) 3.16 b 2.57 3.554 3.04 3.94 3.32
I was experiencing conflict (vs. no conflict) 7.81 b 1.80 7.94 h 1.84 2.68 5.45
I felt my status in the situation was too tow (vs.
too high) 4.90 2.15 5,74 2.08 3.89 = 05 4.26 4.68
My action was inappropriate to the situation (vs.
appropriate) 5.00 2.89 5.52 2.91 1.84 1.97
The cause of the emotion was unknown to me (vs. known) 2.06 b 1.55 2.23 b 2.04 2.32 1,77
I thought I would be punished (vs, abandoned) 4.97 2.40 4.90 2,71 4.71 4.77
I thought I should bepunished (vs. abandoned) 6.55 b 1.79 6.00 a 2.05 4.87 5.13
I wanted to hide (vs. make restitution) 5.26 3,17 5.87 3.41 4.19 4.42
I thought others would f r o w n (vs. laugh) 7.06" 2.29 7.52 b 2,05 4.71 5.00
I was surprised (vs. not surprised) 4.84 2.63 5.55 2.72 4.19 4,87
I felt extremely self-conscious or exposed (vs. not at all) 6.29" 2.75 7.61 a 1.65 6.14 < 05 3.13 4.48
t felt my power in the situation was too little (vs. too
much) 4.97 2.52 6.19" 2.48 4.93 < 05 4.74 6.23 m"
I felt not inhibited (vs. inhibited) 3.45 b 2,05 3.13 ~ 2.22 7.55 5.71
I wanted to reward others (vs. punish others) 4.61 1.67 3.74 b 1,63 4.66 < 05 6.81 6,71
I felt, extremely, I had done an unjust thing (vs. not
at all) 7.42 b 1.98 6.45 b 2.67 1.71 2.74
I was concerned with the future (vs, the present) 5.19 3.07 4.39 2.89 4.19 6.32
I was active (vs. passive) 6.42 b 2,54 4.74 2.35 9.96 <01 4.45 6.19
My face felt not at all heavy (vs. extremely) 3.68 b 1.96 3.71 b 2.08 7.68 5.58
The situation was ambiguous (uncertain) (vs. clear) 4.10 2.94 5.07 2.95 2.70 < 10 2.13 4.55
I felt I was not al all living up to my ideal self (vs.
extremely) 7.77 b 2.04 7.74 b 1.95 2,58 3.45
I was not in control (vs. in control) 5.32 2.75 6.94 b 1.97 7.62 <01 2.39 4.42
I felt calm (vs. aroused) 2.94 b 2.25 2.84 b 1.85 7.35 4.39
Others were approaching me (vs. avoiding me) 6,16 b 2.24 5.45 2.55 5.74 6,52
~p < .05.
~p < .Ol.

~aa
32 Wicker, Payne, and Morgan

versus self-evaluation of the self. The only indication of a difference in

interpersonal orientation is that shame produced a greater desire to punish
others. Finally, it was found that shame lasted longer than guilt.
Table I also contains the means for Serenity and Hope ratings, which
appear to support the conceptual validity of the instrument by showing an
expected pattern. Note that Serenity (and sometimes, to a lesser degree,
Hope) contrasts with both Guilt and Shame in that subjects felt calm and
relaxed pleasure and self-confidence. They felt they were living up to their
ideal self, not that they had done a wrong or unjust thing, and they did not
suffer from self-consciousness, conflict, or inhibition. Fear and Anxiety
ratings are not presented here, but can be found in Wicker and Payne
(submitted for publication).

EXPERIMENT2

Experiment 1 was essentially exploratory. The small number of

subjects employed, which was less than the number of scales, limited the
power of that study to detect real differences, and limited the appropriat-
ness of multivariate analysis. Several steps were taken to increase sensitivity
in Experiment 2, including a considerable increase in the number of obser-
vations per cell.

Method

Subjects. Participants were 121 students of introductory educational

psychology at the University of Texas at Austin.
Materials and Procedure. The settings, booklets, and procedure of
this study were similiar to those of the previous one. Only Shame and Guilt
were presented in the booklets this time, however, and subjects were asked
to recall situations that were relatively pure examples of each one,
independent of the other one, before rating each situation on 34 scales. This
procedure was repeated over three trials, so each subject rated three guilt
situations and three shame situations in all.
The set of scales for this study overlaps only partially with that of
experiment 1 (see Table II). A few scales were dropped from the first list, a
few were modified in the attempt to clarify their meaning, and a few new
scales were added in the attempt to capture distinctions proposed by
previous theorists that were not captured by the first list. For example, the
scale "I was unhappy with myself in general (vs. only with my behavior)"
was inspired by the claim that shame is about the self while guilt is about the
thing done. "A short written description of why I felt uncomfortable would
Ratings of Guilt and Shame 33

be easy (vs. difficult)" was suggested by the view that shame is more
"wordless' and guilt a more articulated experience (Lewis, 1971).
Instructions for this study were given in writing as well as orally. One
subject was randomly discarded for analysis, leaving 60 with each order of
presentation.

Results'

Again, much similarity between the two emotions is suggested in Table

II: Both involve negative self-evaluation and are painful, tense, agitating,
real, present, and depressing. 6 With the more sensitive procedures of this
study, however, differences between the two also became highly salient.
There were significant differences on more than half of the scales. A Wilks's
lambda of .53 was obtained with the MANOVA for the guilt versus shame
variable (F(34, 85) = 2.23, p < .01). Univariate F's for which p < .10 are
shown in Table II. It can be seen that all significant differences at p < . 10 in
Experiment 1 were replicated at p < .05, except one: In Experiment 2 the
two emotions did not differ in duration. In addition, several significant
differences were found in Experiment 2 that had not been obtained in
Experiment 1: Shame scored higher on desire to hide, hot and flushed face,
and surprise, while guik implicated more unjust behavior and expectation of
punishment. The scales used only in Experiment 2 also produced several dif-
ferences in expected directions: Expectation of laughter, feelings of
rejection, inferiority, and competition were all greater with shame than
with guilt. The MANOVA showed an effect of trials or repetitions, Wilks's
Lambda = .55 (F(68,406) = 2.10, p < .01), but trials did not interact with
treatment. The last column of Table II shows all univariate P s for trials
with p < .05. Trends of means over trials suggest that raters recalled
gradually less "potent" situations in later repetitions: In the first-recalled
situations they felt more inferior, tense, generally unhappy with themselves,
low in status, self-conscious, powerless, out-of-control, and agitated than in
those rated later. It is as if subjects became more reluctant over time to
recall traumatic events, or used their most memorable examples on early
trials. Consistent with either view, later-reported emotions were also more
likely to involve imaginary sources, correct behavior, low activity, and
brevity. It should be noted, however, that analysis o f trial 1 data alone
tended to give smaller F's than with the combined data, and basically similar
patterns.

6Significance levels for differences from the neutral point are not presented in Table II. It
can be seen that most of them would be significant with the large numbers of observations
in Experiment2.
Table II. Means and Standard Deviations for Guilt a n d S h a m e on Each Scale, F and P for Guilt-shame C o m p a r i s o n s , and F
for the Trials E f f e c t - E x p e r i m e n t 2
Guilt Shame
Trials
Scale Mean SD Mean SD F p F
I felt pleasure (vs. pain) 3.55 2.12 3.48 2.22
In comparison to others I felt inferior (vs. superior) 5.96 2.21 6.40 2.24 7.03 < .01 6.35
The emotion began suddenly (vs. gradually) 5.09 2.79 5.34 2.83
I felt extremely self-confident a n d capable (vs. not at all) 4.06 2.34 3.60 2.22 7.95 < .01
I felt tense (vs. relaxed) 6.70 2.33 6.96 2.21 14.29
My face was not at all hot and flushed (vs. extremely) 5.20 2.73 4.39 2.64 14.72 < .001
t was dominant (vs. submissive) 4.97 2.45 4.22 2.50 16.12 < .001
The emotion lasted a short time (vs. a long time) 4.31 2.52 4.22 2.57 4.19
I was more concerned with how others viewed me (vs. how
lviewedmyselJ) 5.52 2.83 5.59 2.91
I was u n h a p p y with myself in general (vs. only with my
behavior) 5.07 2.91 4.94 3.02 3.90
After this event I felt not at all depressed (vs. extremely) 3.77 2.28 3.64 2.16
T h e emotion was caused m o r e by m y imagination (vs. things
that actually happened) 3.42 2.62 3.41 2.52 3.54
A short written description o f why I felt u n c o m f o r t a b l e
would be easy (vs. difficult) 5.66 2.79 5.37 2.88
I felt my status in the situation was too low (vs. toohigh) 5.15 2.15 5.58 2.35 7.17 < .01 4.54
My action was inappropriate to the situation (vs.
appropriate) 5.28 2.67 5.21 2.75
In this situation m y attention was scattered and distracted
(vs. focused and concentrated) 4.72 2.75 5.07 2.82 2,84 < .10
I was feeling cooperative (vs. competitive) 4,81 2.33 4,48 2,27 3.55 '--.05
I thought I would bepunished (vs. abandoned) 5.84 2.16 5,41 2.20 6.58 < .05
I wanted to hide (vs. make restitution) 4.96 2.84 5.71 2.90 9.86 < .01
I thought others would laugh extremely (vs. not at all) 3.99 2.47 4.85 2.70 25.75 < .001
I was not surprised (vs. surprised) 4.74 2,54 4.21 2.54 7,07 < .01
I felt extremely self-conscious or "exposed" (vs. not at all) 5.78 2.47 6.61 2.41 19.25 < .001 3.18
I felt my power in the situation was too little (vs. too e~

much) 5.22 2.43 5.98 2.45 16,80 < .001 8.86

I felt not inhibited (vs. inhibited) 4.04 2.32 3,52 2.30 10.23 < .01
I wanted to reward others (vs. punish others) 4.80 1.78 4,23 t,81 16.85 < .001
I felt, extremely, I had done an unjust thing (vs. not at all) 6.09 2.48 5.56 2,71 6.86 < .01
I was concerned with the future (vs, the present) 4.32 2,96 4.39 2.89
I was active (vs. passive) 5.99 2.65 5.52 2.73 7.11 < .01 4.68
Negative evaluation was coming m o r e f r o m myself(vs.
others) 6.05 2.82 5,75 2.89
The situation was ambiguous (uncertain) (vs. clear) 3.92 2.68 4.29 2.70 3,91 < .05
I felt I was exceeding boundaries o f correct behavior not
atall(vs, extremely) 3.79 2.40 3.98 2.53 3.27
I was not in control (vs. in control) 4,76 2.67 5.39 2.62 8.46 < .0t 3.23
M y body felt calm (vs, agitated and aroused) 3.64 2,26 3,4I 2.20 5.10
Others were rejecting me (vs. accepting me) 4.64 2.35 5.20 2.27 11,20 < .01
36 Wicker, Payne, and Morgan

DISCUSSION

Results were compatible with a number of claims in the literature

about how shame and guilt differ, but not with all of them.
Our subjects seemed to agree strongly with the view that shame is the
more incapacitating or overpowering emotion (e.g., Lewis, 1979; Lynd,
1958). They were more active and felt greater control when guilty; they felt
more submissive, inferior, inhibited, and lacking in status, power, and self-
confidence when ashamed. With shame they appeared in general to be
weaker, more helpless, and under the scrutiny and control of others.
Perhaps the active, controlled quality of guilt is consistent with the
metaphor of exceeding a boundary and the passive, helpless quality of
shame is a correlate of failure to achieve a goal (Piers & Singer, 1953). These
results are reminiscent of Erikson's (1963) statement, in a developmental
context, that "shaming exploits an increasing sense of being small" (p. 253),
and they document a sense in which shame can be said to be the greater
threat to self-esteem. They also suggest a basis for the claim that shame is
the "stronger" emotion, but it should be noted that our raters did not
consider it stronger in terms of the degree of pain, tension, or arousal
produced.
What can be said about the situation that produces the greater
incapacity of shame? It is seen as more ambiguous and more surprising,
both of which can increase the difficulty of reacting appropriately and
adequately. Therefore, the resulting response may be a weak one such as
"hiding" rather than an active purposeful one such as "making restitution."
There was also support for the claim that shame entails greater self-
consciousness and feeling of exposure to the evaluative view of others (e.g.,
Lewis, 1971). In this sense shame could be said to be more oriented toward
others than guilt, but there was no expressed difference in concern for
other-evaluation versus evaluation by the self in either study. In this there is
more agreement with theorists such as De Rivera (1977) and Rawls (1963),
who argued that other-orientation is characteristic of guilt and shame, than
with the many writers who argued that it is more characteristic of shame.
The difference between the two emotions may lie in the extent to which one
feels actively scrutinized by others and not in the attribution of importance
to evaluation from them. Or it may lie only in objective self-awareness
(Duval & Wicklund, 1972).
On the other hand, subjects did seem to express greater alienation
from others when ashamed. They reported a greater desire to punish others,
compete with them, or hide from them with shame. Other people were more
likely to reject the raters, laugh at them, or abandon them. Apparently the
incapacity that is associated with shame often comes at the hands of others.
Therefore, the greater extrapunitiveness with shame may reflect a desire for
Ratings of Guilt and Shame 37

revenge, to rectify the power imbalance and/or return the pain received. Or
perhaps this attitude is a defense to rationalize denial and protect self-
esteem from the "prying" and negative evaluating tendencies of the other
person. If so, it would appear that, even as they experienced shame, our
subjects were working at defending themselves against it (Lewis, 1971). As
Erikson says of the shamed person, "He would like to destroy the eyes of
the world ''7 (1963, pp. 252-253). With guilt, these extrapunitive mechanisms
may be either weaker or counterbalanced by the desire to make restitution
to others. 8
The view that guilt is more tied to ethical and moral issues than shame
was supported by differences on the "I had done an unjust thing" scale and
on expectation of punishment in Experiment 2. Evidence for this difference
was somewhat weak: It did not appear on "exceeding boundaries" or on any
of the relevant scales of Experiment 1. The problem may have been simply
that it was difficult to find wordings for scales that could not be interpreted
in terms of transgressions of propriety or appropriateness as well as ethical
transgressions. Thus, there is tentative evidence that raters felt a difference
in salience of ethical values or principles of justice.
With the two scales directed at perceived physiological correlates of
the emotions, one difference was found. The shamed face was remembered
as more hot and flushed, which is consistent with the literature (Izard, 1977)
and with the finding that more surprising, incapacitating, ego-threatening
situations elicit shame. Use of more bodily oriented scales in future studies
might reveal a fuller pattern of physiological differences.
Though our raters seemed to agree more than disagree with emotion
theorists, they did not always agree. Consider, for example, the claim that
guilt lasts longer than shame. Subjects reported the opposite difference in
Experiment 1 and no difference in Experiment 2. They also did not agree
with Kemper's (1978) association of guilt with excess power and shame with
excess status. Power ratings were not above the neutral point with guilt, and
shamed respondents reported a deficit of status. While the proposal that a
feeling of excess status would produce shame seemed palusible, shamed
subjects tended in fact to feel that they were getting less status than they
deserved or wanted. This result, however, may reflect the defensive esteem-

7Such defensiveness might have masked a greater concern for the evaluations of others in
shame situations, a point that should remind us o f the necessity for caution in interpreting
rating data.
sOther interpretations can be offered o f course, but the ones that have come to the authors'
attention appear less compelling than those presented in the text. For example, it is possible
that s h a m e is m o r e likely to follow the act o f punishing someone else (which is correlated
with desire to punish) than is guilt. But it seems probable that guilt would be the more com-
m o n consequence.
38 Wicker, Payne, and Morgan

supporting tendency proposed above. The rating procedure was also not
sensitive to proposed differences in orientation toward self versus thing
done, degree of articulation, or consequent depression (e.g., Lewis, 1971).
Finally, whereas differences between guilt and shame have been the
focus of this paper, Tables I and II suggest many commonalities too. In
only 6 of 68 comparisons in two studies were the two emotions significantly
different at p < .05 and on opposite sides of the neutral point. In the
majority of cases the difference was one of degree but not direction. In both
studies, for example, raters felt self-conscious and exposed when guilty but
even more so when ashamed. Also, while many differences were significant,
the absolute magnitude of the difference was usually not very large. In
Experiment 2 this may reflect a gradual weakening of recalled incidents over
trials, since absolute differences seem somewhat larger in Experiment 1. But
in addition, subjects in both studies probably often chose experiences in
response to one emotion word that theorists would consider more ap-
propriate to the other. And there may be source of confusion other
than limitations of raters. Each word could have several common
meanings-some independent, but others equivalent to a meaning of the
other word. Or both emotions may often be aroused by the same situation
and they may fuse (Lewis, 1979). Yet despite these potential sources of noise
(as well as rating biases, memory limitations, and effects of scale wordings),
differences emerged that suggest that undergraduate education majors
make many of the same distinctions between guilt and shame as do psycho-
logists, psychoanalysts, and philosophers who have studied these emotions.
The question of the extent to which ratings reflected observed features of
personal experience as opposed to previously learned, culturally trasmitted
conceptions of how shame and guilt differ is open to debate, but at least
these concepts had not been discussed in the introductory educational
psychology course subjects were taking. These ratings may be taken as
support for the generality and validity of several distinctions that have
emerged from prior clinical, introspective, and theoretical w o r k - f o r their
generality in our representations of experience, if not in our experience.
The standard deviations in Tables I and II show that there was con-
siderable variation among raters on each scale. This variability is not
surprising when we remember that the particular situation that represented
an emotion was different for each rater. It means, however, that we
probably need a large number of observations, as in Experiment 2, to detect
real average differences among emotions.
Nevertheless, it has been shown that a "critical-incident" technique of
structured naive introspection about remembered emotional experiences
can demonstrate differences in the usage of conceptually similar emotion
terms that are consistent with prior theory. Raters would probably have
great difficulty telling us how they use these terms if they were asked
Ratings of Guilt and Shame 39

directly. But apparently they can rate particular situations consistently and
meaningfully, and in so doing help determine which of the distinctions
made by psychological theorists have functional significance to nonpsycho-
logists in important life settings.

REFERENCES

Ausnbel, D. P. Relationships between shame and guilt in the socializing process. Psycho-
Iogical Review, 1955, 62, 378-390.
Darwin, C. R. The expression of emotions in man and animals. London: John Murray, 1872.
De Rivera, J. A structural theory of the emotions. New York: International Universities
Press, 1977.
Duval, S., & Wicklund, R. A. A theory of objective self awareness. New York: Academic
Press, 1972.
Erikson, E. H. Childhood andsociety (2nd ed.). New York: Norton, 1963.
Izard, C. E. Human emotions. New York: Plenum, 1977.
Kemper, T. D. A social interactional theory of emotions. New York: Wiley, 1978.
Lewis, H. B. Shame and guilt in neurosis. New York: International Universities Press, I971.
Lewis, H. B. Shame in depression and hysteria. In C. E. Izard (Ed.), Emotions in personality
and psychopathology. New York: Plenum, 1979.
Lynd, H. M. On shame and the search for identity. New York: Harcourt Brace Jovanovich,
1958.
McNemar, Q. Psychologicatstatistics (2nd ed.). New York: Wiley, 1955.
Morris, H. (Ed.). Guilt and shame. Belmont, California: Wadsworth, 1971.
Piers, G., & Singer, M. B. Shame and guilt: A psychoanalytic and cultural study. Spring-
field, Illinois: Charles C Thomas, 1953.
Rawls, J. The sense of justice. PhilosophicalReview, 1963, 72, 281-305.
Sartre, J. P. Being and nothingness: An essay on phenomenotogical ontology. New York:
Philosophical Library, 1956.
Tomkins, S. S. Affect, imagery, consciousness (Vol. 2) The negative affects. New York:
Springer, 1963.
Wickelgren, A. Cognitive psychology. Englewood Cliffs, New Jersey: Prentice-Hall, 1979.
Wicker, F. W., & Payne G. C. Fear and anxiety: Are they synonyms? Submitted for publica-
tion.