Professional Documents
Culture Documents
Patricia Rockwell
Department of Communication
Lafayette, LA 70504
par2323@louisiana.edu
in San Antonio, TX
2
Abstract
student respondents, was conducted to determine personality traits that influence sarcasm
production. Regression analysis indicated that two subscales (Negative Tactics and
and one subscale (Performance) of the Lennox and Wolfe Self-Monitoring Scale (SMS)
features of sarcastic utterances (Clark & Gerrig, 1984; Jorgensen, Miller, & Sperber,
1984; Kreuz & Glucksberg, 1989; Sperber, 1984). Scant studies have actually
that have examined users of sarcasm, most have explored demographic features such as
age, culture, gender, and relationship (Gibbs, 2000; Rockwell, 2001, 2003; Rockwell &
Theriot, 2001). Yet to be determined are individual personality traits that facilitate or
can have confusing if not devastating effects on communication (Glaser et al., 2000).
Rockwell (2005), for example, found a significant but small positive correlation between
speakers’ cognitive complexity and their expression of sarcasm. The present study hopes
to expand on this approach and determine individual traits that promote sarcasm
production with the ultimate goal of providing a clearer picture of the sarcastic speaker.
recognition (Clark & Gerrig, 1984; Gerrig & Goldvarg, 2000; Jorgensen, Miller &
Sperber, 1984; Kaufer, 1981; Kreuz & Glucksberg, 1989; Slugoski & Turnbull, 1988;
Sperber, 1984 Williams, 1984), but few to measure sarcasm production. Ivanko,
Pexman, and Olineck’s (2004) Sarcasm Self-Report Scale (SSS) is a recent attempt to
measure individuals’ reports of their own behavior regarding sarcasm use and the types of
situations that prompt these behaviors. The researchers report four subscales for the SSS:
4
The questions in the scale ask respondents to rate (on a scale from “1” for “not at all” to
“7” for “extremely likely”) the likelihood that they would use sarcasm in various
situations. The four factors reported were determined through principle components
analysis (PCA) in two separate studies conducted by the researchers. “General Sarcasm”
was found to measure overall tendency to use sarcasm. “Face-saving” measures the
“Embarrassment Diffusion” measures how speakers use sarcasm to downplay their own
decrease the annoyance of various frustrating situations. The researchers found that
participants’ responses on the SSS predicted their behavior in a sarcasm production task
and also predicted their ability to quickly and correctly recognize sarcastic messages.
postulated that individuals with certain personality types may be more inclined to
produce certain behaviors than those with other personality types. For example, one
early personality categorization system was the Myers-Briggs. Of the four personality
Test (artisans, guardians, idealists, and rationals), artisans are seen as most likely to use
sarcasm, and rationals are seen as least likely to use sarcasm. Artisans, according to
Keirsey (1998) are more prone to use sarcasm “because of their tendency to do
whatever is necessary, even if that means being rude” (p. 47). Rationals were argued to
be the least sarcastic personality type in Keirsey’s system because they maintain a
5
greater distance from people than other types (p. 163). These observations, however,
remain to be tested. Even so, they represent an initial attempt at connecting personality
have proven relevant in the study of individual behavior. Traits such as altruism,
Some, if not all, of these personality traits may influence individuals’ production of
sarcasm. However, in attempting to focus on those traits with the greatest possible
association to sarcasm production, three well-investigated traits have been selected for
rationale for inclusion of each of these three variables will be discussed individually in
Eysenck, 1991; Christie & Geis, 1970, McCutcheon, 2002). The High Mach is “a
emotional coldness, duplicity, and aggressiveness.” (Paulhus & Williams, 2002, p. 557).
High Machs are controlling and use others to their own advantage. They tend to have
“Type A” personalities and are less ethically-oriented than Low Machs (Rayburn &
Rayburn, 1996). Scientists have found that there is no significant correlation between
Machiavellianism and intelligence, and have also found that high levels of
Machiavellianism do not necessarily lead to social success (Wilson & Miller, 1996).
6
psychopathy (McHoskey et. al., 2001). In one study, male medical students rated
themselves higher than female medical students on Machiavellianism (Merrill et. al.,
1993).
There has been little research on the language behaviors of high and low Machs.
In one study, High Machs used more first person-singular pronouns than Low Machs
(Ickes et al, 1986). It is intuitively appealing to imagine High Machs as using language
features—including sarcasm--that they believe will aid them in achieving their goals.
The standard test for Machiavellianism is Christie and Geis’ (1970) Mach V
Scale. This 20-item survey has been modified from the former Mach IV test, by
replacing forced-choice items with Likert-type items having responses of “1” to “7.”
Corral and Calvete’s (2000) confirmatory factor analyses of the Mach IV produced four
Human Nature, and Negative View of Human Nature. The “Positive Interpersonal
their goals and the “Negative Interpersonal Tactics” subscale measures negative
behaviors reported by respondents used to achieve their goals. The “Positive View of
Human Nature” subscale measures positive attitudes about others that are held by
respondents, and the “Negative View of Human Nature” subscale measures negative
attitudes about others that are held by respondents. The survey includes items such as
“I handle people by telling them what they want to hear,” (Positive Interpersonal
Tactics), “I would be prepared to walk all over people to get what I want” (Negative
Interpersonal Tactics), “I believe there’s a sucker born every minute” (Negative View of
7
Human Nature), and “Honesty is the best policy in all cases.” (Positive View of Human
Various researchers have used and examined the Mach IV and Mach V, and
some have complained that the test does not produce reliable results (e. g., Ray, 1983).
Even so, the Mach tests remain the gold standard for determining Machiavellianism.
expressing cynicism about something, but for some reason is not willing to state this
cynicism directly. Possibly this indirect approach is a strategy of high Machs. Due to
measured by the four subscales of the Mach V) and sarcasm production (as measured
Self-monitoring is the trait of observing one’s own behavior and adjusting that
behavior when necessary to adapt to other people and surroundings (Day et.al., 2002;
Gudykunst, 1985; Lindsey & Green, 1987). The ability to self-monitor is considered
integral to effective social functioning (Riggio & Friedman, 1982). High self-monitors
tend to adapt to situations, letting the requirements of the situation dictate their
behavior. Low self-monitors behave relatively the same way in all circumstances and
are less aware or less concerned about their own behavior. One theorist (Behncke,
2005) notes that in actuality, Highs monitor the behavior of others rather than their own
behavior in order to determine how their own behavior compares and then typically
8
adjust their behavior based upon these observations. Lows do not monitor their own or
others’ behaviors because they have little or no interest in adjusting their own behavior.
One study found that students rated high self-monitoring teachers better than low self-
Little research has been conducted regarding the language behaviors of self-
monitors. In one study, high self-monitors used more second and third person pronouns
than did low self-monitors (Ickes et. al., 1986) implying a desire to not take
responsibility for one’s own views. In another study (Galbraith, 1996), high self-
monitors produced more new ideas after writing notes than did low self-monitors.
However, this result was reversed when subjects were asked to write finished texts,
with low self-monitors producing more new ideas than high self-monitors.
something one might expect of high self-monitors. On the other hand, low self-
monitors may use a more negative abusive form of sarcasm when they are angry
because they are not particularly concerned about what others think of them.
different ways. The Lennox and Wolfe Self-Monitoring Scale (SMS) (Schutte &
subscale measures individuals’ ability to sense how others view them, and the
“Performance” subscale measures their ability to put this sensitivity to use. Sample
items include: “I am often able to read people’s true emotions correctly through their
9
eyes,”(Sensitivity) and “When I feel that the image I am portraying isn’t working I can
The implication with self-monitoring is that people monitor their own behavior
so that they can change those behaviors that are not interpersonally successful. The
question with self-monitoring and sarcasm use is whether a high self-monitor, who is
more sensitive to others and more willing to adapt behaviors, would recognize when
sarcasm might be used successfully and thus, use it more or less frequently. Due to the
fact that it is uncertain how any correlation between self-monitoring and sarcasm
two subscales of the SMS) and sarcasm production (measured by the four subscales of
the SSS)?
Emotional Expressivity
Some individuals are more emotionally expressive than others (Bippus &
(Gross & John, 1997, p. 435). These researchers point out that individuals differ
greatly in their expression of emotion. They differ in the strength or valence of emotion
that is required to activate expression and also they differ in the way they modulate or
filter the emotions that they express. Some individuals allow their emotions expression
readily while others inhibit the expression of the emotions they experience. The
reasons for these differences include cultural “display rules” as well as individual
differences in the intensity of the emotional experience. Gross and John’s study
10
determined that individuals are capable of recognizing the emotions they experience as
In another study (Hess, Senécal, Kirouac, Herrera, Phillippot, & Kleck, 2000),
gender differences were found in stereotypes that people possess regarding the way
men and women experience and express emotions, with men being perceived of as
experiencing and expressing positive emotions more than women, and women being
perceived of as experiencing and expressing negative emotions more than men. In Hess
et. al’s study, men and women also reported similar reactions about their own
“Impulse Strength” subscale concerns how strong individuals perceive their emotions.
The “Positive Expressivity” subscale concerns whether and to what extent individuals
perceive that they express positive emotions, and the “Negative Expressivity” subscale
concerns whether and to what extent individuals perceive that they express negative
emotional expressivity. Possibly, more expressive individuals will present anger more
readily when they feel it by using sarcasm. On the other hand, individuals who are not
emotionally expressive may use sarcasm to attenuate their anger expression. Due to the
11
by the BEQ and its three subscales) and sarcasm production (measured by the four
Method
university during the summer of 2005. Student respondents completed the survey during
class. After completion of the student surveys, student participants were asked to
distribute surveys to non-students whom they knew (one male and one female—
maximum of two) so as to acquire a larger, more diverse sample. Students were given
information was asked on the student surveys; however, on the non-student surveys, we
asked for the respondent’s names and phone numbers so that we could call to verify that
the respondents were who they said they were. Following verification, the respondents’
names were removed from the survey. A removable cover letter was attached to all
surveys which explained the purpose of the survey and which included contact
removable consent form followed the cover letter and was removed by the research
assistant when the respondent returned the survey. The consent forms were kept only
long enough to ensure that the student respondents received their extra credit.
12
Instrument
(Ivanko, Pexman, & Olineck, 2004), a modified version of the Mach V (Christie & Geis,
1970), the Lennox and Wolfe Self-Monitoring Scale (SMS) (Schutte & Malouff, 1999),
and the Berkeley Expressivity Index (BEQ) (Schutte & Malouff, 1999).
Results
We collected data from 150 students and 68 non-students for a total of 218
respondents (males = 87, females = 131). Summary statistics of the sarcasm questions
revealed interesting results (see Table 1 for sarcasm question means). The question on
the SSS with the highest mean score (greatest likelihood) was “What is the likelihood you
would use sarcasm with your best friend?” (m = 5.66, sd = 1.72). The question with the
lowest mean score (least likelihood) was “What is the likelihood you would use sarcasm
The overall regression analysis was significant (see Table 2 for regression analysis
results), allowing further investigation of the individual variables. The analysis produced
significant results for components of two of the three scales: two subscales of
sarcasm use. The variation accounted for, as indicated by the R², was 26%. This
suggests that these three subscale factors predict over a quarter of total sarcasm usage.
Emotional expressivity did not predict sarcasm usage, nor did the Positive Interpersonal
Tactics or Negative View of Human Nature subscales of the Mach V, nor the impulse
Discussion
The fact that we found significance between only some subscales of these
commonly measured personality traits cautions researchers to remember that many of the
instruments used in social science research are multifaceted. Indeed, the Mach V’s four
attitude (Positive and Negative Views of Human Nature). In our study, one tactics
component and one attitude component were strongly correlated with sarcasm use. This
finding is unusual. One might expect the Negative Interpersonal Tactics subscale to
predict sarcasm use better than Positive Interpersonal Tactics. However, it is strange that
the Positive View of Human Nature subscale was a significant predictor of sarcasm use
and not the Negative View of Human Nature subscale. One possible explanation is that
two of the Positive View of Human Nature subscale’s questions emphasized honesty, and
those who see themselves as honest may be more willing to use sarcasm.
14
significantly predicted sarcasm use—but not the Sensitivity subscale. Here, the key
appears to be actual behavior rather than attitude serving as the predictor. Individuals
who reported that they actually act on the information they acquire from the monitoring
that they do were more likely to report sarcasm use. Individuals who are sensitive to how
others view them may not always act on this behavior and thus may be less likely to use
sarcasm.
We were surprised to find that emotional expressivity did not correlate with
reported sarcasm usage. Is the sarcastic individual attempting to control their true
feelings when they use sarcasm? Or do sarcastic remarks represent a genuine revelation
of speakers’ true feelings couched in socially acceptable terms? This study did not
We did not find that the three subscales of the BEQ were positively correlated
with the SSS as were two subscales of the MACH V and one of the SMS; we believed we
would see similar correlations with the BEQ’s subscales. This did not happen. It appears
then, that the behaviors measured by the BEQ have little connection to sarcasm
production. That is, it appears that heavy sarcasm producers do not necessarily see
expression, one assumes the speaker is expressing emotion accurately and willingly. This
may not be what actually happens. Indeed, some sarcastic individuals may, for some
reason, retain greater control and be less emotionally expressive. Finally, as some
researchers have noted, “. . . from an evolutionary perspective, language did not emerge
15
Thus, some of the traits of manipulative and monitoring behaviors seem to have a
strong relationship to sarcasm production. These findings help “flesh out” the picture of
the sarcastic speaker. Sarcasm users appear to practice negative manipulation of others
while maintaining a positive view of human nature. They are able and willing to adjust to
This study considered how psychological traits impact the language behavior
called sarcasm. A future study may reverse this approach and examine how sarcasm
usage influences psychological traits. As noted by other researchers, “The words people
use in their daily lives can reveal important aspects of their social, psychological worlds.”
References
Bippus, A. M., & Young, S. L. (2005). Owning your emotions: Reactions to expressions
Christie, R., & Geis, F. (1970). Studies in Machiavellianism. NY: Academic Press.
Clark, H. H., & Gerrig, R. J. (1984). On the pretense theory of irony. Journal of
Day, D. V., Schleicher, D. J., Unckless, A. L., & Hiller, N. J. (2002). Self-monitoring
Couzijn (Eds.). Theories, models, and methodology in writing research (pp. 121-
Gerrig, R. J., & Goldvarg, Y. (2000). Additive effects in the perception of sarcasm:
Situational disparity and echoic mention. Metaphor and Symbol, 15(4), 197-
208.
Glaser, R., Sheridan, J. F., Malarkey, W., B., McCollum, R. C., & Kiecolt-Glaser, J. K.
Gross, J. J., & John, O. P. (1997). Revealing feelings: Facets of emotional expressivity
52 (3), 203-218.
Hess, U., Senécal, S., Kirouac, G., Herrera, P., Philippot, P., & Kleck, R. E. (2000).
Ivanko, S. L., Pexman, P.M., & Olineck, K. M. (2004). How sarcastic are you?:
Jorgensen, J., Miller, G. A., & Sperber, D. (1984). Test of the mention theory of irony.
495-510.
Kreuz , R. J., & Glucksberg, S. (1989). How to be sarcastic: The echoic reminder
Lindsey, A. E., & Greene, J. O. (1987). Social tendencies and social knowledge: Self-
Merrill, J. M., Lorimor, R. J., Thornby, J. I., & Vallbona, C. (1998). Medical manners:
256-260.
Paulhus, D. L., & Williams, K. M. (2002). The dark triad of personality: Narcissism,
563.
19
natural language use: Our words, our selves. Annual Review of Psychology, 54,
547-577.
1209-1219.
Rockwell, P. (2001). Facial expression and sarcasm. Perceptual and Motor Skills, 93,
47-50.
Rockwell, P. (2003). Empathy and the expression and recognition of sarcasm by close
Rockwell, P., & Theriot, E. (2001). Culture, gender, and gender mix in encoders of
Schutte, N. S., & Malouff, J. M. (1999). Measuring emotional intelligence and related
Slugoski, B. R., & Turnbull, W. (1988). Cruel to be kind and kind to be cruel: Sarcasm,
banter, and social relations. Journal of Language and Social Psychology, 7(2),
101-121.
Williams, J. P. (1984) Does mention (or pretense) exhaust the concept of irony? Journal
285-299.
21
Table 1
Sarcasm Self-Report (SSS) Item Means and Standard Deviations
(scale: “1” = “not at all,” “7” = “extremely”)
Mean Standard Deviation
9. When you and roommate are arguing about chores 4.23 1.78
10. When you score winning point for team 3.48 1.97
12. When you are in long line at grocery store 4.68 1.92
13. When you just got engaged and telling friends 2.63 1.68
14. When you are celebrating promotion with family 2.88 1.82
15. When you are late to work and lock keys in car 4.44 2.14
Table 2
Regression Analysis for Effects of Machiavellianism, Self-Monitoring, and
Emotional Expressivity on Sarcasm Use
Variable B SEB t
Machiavellianism
Negative Tactics .24 .10 2.31*
Positive Tactics .07 .08 .81
Negative View.11 .08 1.66
Positive View .27 .07 3.54**
Self-Monitoring
Performance .14 .06 2.10*
Sensitivity -.02 .08 -.23
Emotional Expressivity
Positive .12 .10 1.26
Negative .00 .07 .02
Impulse Control -.10 .07 .02
Analysis of Variance
Source SS DF MS F
Regression 59.08 9 6.56 6.73**
Residual 156.07 160 .97
Total 215.16 169 1.27
R² = .26