Failing to take the moral high ground: Psychopathy and the vertical ...

Abstract
Personality and Individual Differences 43 (2007) 757–767
Failing to take the moral high ground: Psychopathy
and the vertical representation of morality
Brian P. Meier a, *, Martin Sellbom b , Dustin B. Wygant b
a Department of Psychology, Gettysburg College, Gettysburg, PA 17325, United States
b Kent State University, United States
Received 1 August 2006; received in revised form 17 December 2006; accepted 2 February 2007
Available online 21 March 2007
Morality is explained in metaphors that use descriptions of verticality (e.g., ‘‘an upstanding citizen’’). It
is unknown, however, if these metaphors simply aid communication or indicate a deeper mode of knowledge
representation. In two experiments, we sought to determine the extent to which verticality is used
when encoding moral concepts. Furthermore, because psychopaths are characterized by a lack of moral
concern, we believed this personality dimension could act as an important moderator. Experiment 1 established
that people have implicit associations between morality and vertical space. Experiment 2 extended
this finding by revealing that people low in psychopathy encoded moral-related (vs. immoral-related) concepts
faster if they were presented in a high (vs. low) vertical position. This effect did not occur for participants
high in psychopathy. Our results indicate that morality is partially represented on the vertical
dimension, but not for individuals with little concern for morality.
Ó 2007 Elsevier Ltd. All rights reserved.
Keywords: Psychopathy; Metaphor; Affect; Verticality
* Corresponding author. Tel.: +1 717 334 6173.
E-mail address: bmeier@gettysburg.edu (B.P. Meier).
0191-8869/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2007.02.001
www.elsevier.com/locate/paid

758 B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767
1. Introduction
Morality is an abstract concept used to describe behavior or beliefs that an individual considers
to be right and moral (e.g., fairness) or wrong and immoral (e.g., intolerance; Haidt & Algoe,
2004; Lakoff & Johnson, 1999). When discussing morality, people make use of metaphors that
tap vertical space (i.e., moral is up; immoral is down). For example, a person who is moral might
be described as ‘‘high minded’’ or ‘‘on the up and up,’’ whereas a person who is immoral might be
described as ‘‘down and dirty’’ or ‘‘underhanded’’ (Lakoff & Johnson, 1999).
One reason metaphors for morality use descriptions of vertical space is likely because morality
is an abstract concept. Lakoff and Johnson (1999) and Gibbs (2006) contend that abstract thought
is possible because of our capacity for metaphor representation. They believe that metaphors
allow people to communicate and represent (i.e., to depict or make sense of) abstract concepts
that would otherwise have no reference to physicality. In many contexts, we learn about what
things are like through our senses. For example, cherries are red and taste sweet, whereas snowflakes
are cold and make us wet. Morality is different in that it cannot be directly perceived
through the senses. Thus, to appreciate the nature of morality, it makes sense that people use
metaphors when describing it.
Lakoff and Johnson (1999) further contend that human thought processes are structured on
metaphors. The manner in which people encode, store, and retrieve information is grounded in
metaphors. Thus, metaphors are not simply communication devices, but are used to represent
concepts. That is, thinking about abstract concepts is not possible without activating or simulating
the sensations and perceptions relevant to metaphor.
More generally, a number of researchers argue for an embodied mode of cognition. Proponents
of this view contend that cognition, rather than being abstract and amodal, is inherently linked to
sensation and perception (e.g., Niedenthal, Barsalou, Winkielman, & Krauth-Gruber, 2005). In
this view, cognition is not a process that is separated from bodily states, but involves the simulation
of modality-specific actions.
Meier and Robinson (2004) presented evidence for an embodied mode of cognition in the affective
realm. They examined the metaphor that ties affect to verticality (i.e., good is up; bad is
down). For example, good things are described as being up (e.g., ‘‘thumbs up’’ for a good movie)
and bad things are described as being down (e.g., ‘‘thumbs down’’ for a bad movie). Meier and
Robinson (2004) found that participants evaluated words with a positive meaning faster if they
were presented in a high vertical position, whereas participants evaluated words with a negative
meaning faster if they were presented in a low vertical position. Participants encoded the vertical
location of the words even though it was unnecessary for the task. This research indicates that
people partially represent affect on a vertical dimension. That is, when encoding good and bad
stimuli, people simulate perceptions of vertical space. Thus, it appears that vertical space is a perceptual
cue for affect. In the current context, we seek to determine if morality is similarly represented
on this same vertical dimension.
1.1. Psychopathy and morality
In addition to examining morality and vertical space, a central purpose of the current project is
to explore the role of individual differences. Little if any research has examined the extent to which

individual differences moderate the embodiment of concepts. Although theorists might consider
embodiment to be a fundamental process (e.g., Niedenthal et al., 2005), it is reasonable to assume
that personality differences exist. More specifically, it might be that concepts not central to one’s
personality are not embodied or represented to a great extent. If an individual does not frequently
invoke a concept, then it seems likely that this concept would not need representation. In the present
case, morality is one such concept that individuals high in psychopathy seem to have little use
for. Psychopathy is typically described as a cluster of personality characteristics and behaviors,
including grandiosity, manipulativeness, impulsivity/norm-violation, and a lack of empathy or remorse
(e.g., Cleckley, 1941; Hare, 1996). Historically, some have suggested that immorality is a
primary underlying characteristic (Krafft-Ebing, 1904).
Research reveals that psychopathy can be characterized on a continuous dimension (e.g.,
Edens, Marcus, Lilienfeld, & Poythress, 2006) that includes individuals at the high extreme
(e.g., violent offenders, Hare, 1991, 1996). Although individuals in non-institutional samples
(e.g., college students) have a lower central distribution of scores on this dimension than prison
inmates, individual differences at this level are nonetheless meaningful. For instance, researchers
have found that psychopathy can be successfully conceptualized in normal personality space (Benning,
Patrick, Hicks, Blonigen, & Krueger, 2003).
Psychopaths appear to be deficient in processing affective stimuli. For example, Williamson,
Harpur, and Hare (1991) found that people high (vs. low) in psychopathy had less behavioral
and neural response differences when processing words with an affective versus neutral meaning.
Herve, Hayes, and Hare (2003) found that people high (vs. low) in psychopathy had difficultly
determining the valence of metaphoric statements. For example, one psychopath rated the statement
‘‘love is an antidote for the world’s ills’’ as strongly negative.
1.2. The present experiments
In the present experiments, we examined whether this deficiency would extend to affectiveknowledge
representation. A lack of moral concern is one characteristic of psychopathy. Therefore,
we expected that psychopathy would impact the extent to which morality is represented on
the vertical dimension. That is, it might be that people who are not concerned about morality do
not use vertical cues to represent it. In Experiment 1, we first sought to determine if people have
implicit associations between morality and vertical space. In Experiment 2, we examined the potential
moderating impact of psychopathy.
2. Experiment 1
B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767 759
We used an implicit association test (IAT; Greenwald, McGhee, & Schwartz, 1998) to examine
the association between morality and verticality. The IAT assesses associations in an implicit
manner by requiring participants to categorize stimuli as belonging to one of four concepts.
For example, Greenwald et al. (1998) had people categorize words from four categories: insects,
flowers, pleasant, and unpleasant. The trials of interest occur when two category names are
mapped onto one button. These researchers found that participants were faster to categorize

760 B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767
words when insect/unpleasant were mapped onto one button and flower/pleasant were mapped
onto another (rather than the reverse).
In Experiment 1, participants categorized words as having either a moral or immoral meaning
and they categorized a cue (***) as appearing either above or below screen center. We predicted
that participants would have implicit associations between morality and ‘‘up’’ and immorality and
‘‘down.’’ Such results would provide evidence that the vertical dimension is invoked during the
encoding of moral concepts.
2.1. Method
2.1.1. Participants
Participants were 69 undergraduates (17 males) with an average age of 19.19 years (SD = 1.32
years).
2.1.2. Materials and procedure
Word stimuli. We selected five words with a moral meaning (caring, charity, nurture, truthful,
and trustworthy) and five words with an immoral meaning (adultery, corrupt, dishonest, evil, and
molest). The number of letters and frequency did not differ, Fs < 1. Twelve individuals (not in
either experiment) rated the extent to which each word had an immoral or moral meaning (1 =
extremely immoral; 9 = extremely moral). Using word as the unit of analysis, the moral words
(M = 8.02; SD = 0.11) were rated as having a significantly more moral meaning than the immoral
words (M = 2.08; SD = 0.24), F(1,8) = 2522.58, p < 0.001, g2 p ¼ 1:00.
IAT. The IAT was comprised of two categorizations. One, participants categorized the words
as referring to an immoral or moral behavior/concept as the words randomly appeared in the center
of a 19-in. computer screen (in green, 16-point Arial font). Two, participants categorized a cue
(three asterisks: ***) as it appeared 1 in. above or below the center of the screen (in green, 20-
Point Arial font).
Participants were asked to perform quick and accurate categorizations. Category names (immoral,
moral, up, down) appeared on the upper left and right of the screen (in white, 16-point
Arial font). Participants were told to press the ‘‘Q’’ key if the word or cue belonged to the category
on the upper left and to press the ‘‘P’’ key if the word or cue belonged to the category on the upper
right. Incorrect categorizations were followed by the word ‘‘INCORRECT’’ in red for 1.5 s. Correct
categorizations were followed by a 150 ms blank screen. The words and cues remained on the
screen until participants responded.
The IAT included several practice blocks, but the main interest was two experimental blocks:
(1) 40 trials for the combined categories (moral or up/immoral or down); (2) 40 trials for the reverse
combined categories (immoral or up/moral or down). The presentation of these blocks was
counter balanced.
2.1.3. Results and discussion
We followed the data reduction procedure of Greenwald et al. (1998): deletion of the first two
trials, replacement of trials below 300 ms and above 3000 ms, deletion of inaccurate trials (3.13%),
and a log-transformation of the latencies. Analyses were performed on log-transformed latencies,
but means are reported in milliseconds.

We conducted a mixed-model ANOVA with two factors. The between-factor pertained to IAT
Order (moral + up/immoral + down in the first combined block vs. immoral + up/moral + down
in the first combined block). The within-factor pertained to IAT Block (moral + up/immoral +
down vs. immoral + up/moral + down). The main effect of IAT Order was not significant,
F < 1. The main effect of IAT Block, F(1,67) = 68.88, p < 0.001, g2 p ¼ 0:51, and the IAT Block
by IAT Order interaction, F(1,67) = 8.21, p = 0.006, g2 p ¼ 0:11, were significant. The main effect
of IAT Block revealed that participants were faster at categorizing the moral words when they were
paired with the asterisks that appeared up and the immoral words when they were paired with
asterisks that appeared down (M = 748 ms; SD = 150 ms) rather than when these pairings were
reversed (M = 887 ms; SD = 202 ms). The IAT Order by IAT Block interaction indicated the
IAT Block effect was stronger when the first combined block included moral + up and immoral +
down ðg2 p ¼ 0:75Þ rather than the reverse ðg2 p ¼ 0:24Þ.
The results of Experiment 1 indicate that people have associations in memory between the concepts
of morality and vertical space. These results establish that people associate morality with
areas of vertical space in the absence of figurative utterances.
3. Experiment 2
Experiment 2 had two goals: (1) we wanted to replicate Experiment 1 with a task that did not
use vertical descriptors (‘‘up’’ or ‘‘down’’), and (2) we wanted to investigate the potential moderating
role of psychopathy. Participants were asked to categorize words as having a moral or
immoral meaning as the words randomly appeared near the top or bottom of a screen. After
the completion of this task, participants completed Levenson’s Self-Reported Psychopathy scale
(LSRP; Levenson, Kiehl, & Fitzpatrick, 1995). We expected that vertical space would interact
with word type such that participants would be faster to categorize a word with a moral meaning
when it was presented on top (vs. bottom) whereas the reverse should occur for words with an
immoral meaning. Furthermore, because a lack of moral concern is one defining feature of psychopathy,
we believed that this effect might not occur for participants high in psychopathy.
3.1. Method
B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767 761
3.1.1. Participants
Participants were 112 undergraduates (59 males) with an average age of 19.48 years (SD = 2.60
years).
3.1.2. Materials and procedures
Word stimuli. We used the same ten words from Experiment 1 and we selected 32 additional
words. 1 The words were similar in number of letters and frequency, Fs < 1. The same 12 individuals
1 Additional moral words were: assist, donate, empathy, ethical, fairness, generous, goodness, goodwill, honesty,
honorable, integrity, kindness, noble, respectful, virtue, and volunteer. Additional immoral words were: bribe, crooked,
deceit, hurt, indecent, lying, manipulate, merciless, perverted, prostitution, ruthless, shoplift, sinfulness, villainous,
violate, and wicked.

762 B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767
from Experiment 1 rated the moral words (M = 7.37; SD = 0.47) as having a significantly more
moral meaning than the immoral words (M = 2.53; SD = 0.52), F(1,40) = 986.64, p < 0.001,
g2 p ¼ 0:96.
Categorization task. Participants categorized the words as referring to a behavior/concept that
was moral or immoral. A randomly chosen word randomly appeared approximately one-quarter
of an inch from the top or bottom of a 19-in. screen (in white, 22-point Arial font). Each word was
presented three times (all words were shown before a repetition began). Thus, there were 126 trials
with approximately 31 trials in each condition (i.e., moral/up, moral/down, immoral/up, immoral/
down).
Each trial began with the presentation of a small white circle in the screen center for 300 ms.
Participants were to attend to this circle. After the circle disappeared, a blank screen appeared
for 300 ms. Then, a randomly chosen word randomly appeared near the top or bottom of the
screen. Participants were to press the ‘‘Q’’ key for words with a moral meaning and to press
the ‘‘P’’ key for words with an immoral meaning. If participants were correct, a 500 ms blank
screen appeared until the next trial. If participants were incorrect, the word INCORRECT was
shown in red for 1.5 s before the next trial. The word remained on the screen until participants
pressed a key.
Levenson’s Self-Reported Psychopathy scale. (LSRP; Levenson et al., 1995). Participants completed
the LSRP after the categorization task. The LSRP includes 26 items (e.g., for me, what’s
right is whatever I can get away with) answered on a scale from 1 (strongly disagree) to 4 (strongly
agree). The LSRP has two subscales (primary and secondary), but a total score consisting of an
average of all items is typically used (which is what we used in the current experiment; a = 0.83).
Research suggests that this measure is at least moderately correlated with Hare’s (1991) psychopathy
checklist-revised (PCL-R; Brinkley, Schmitt, Smith, & Newman, 2001).
3.1.3. Results and discussion
For the categorization task, we deleted inaccurate trials (4.5% of trials), log-transformed correct
trials, and replaced trials above and below 2.5 SDs (with the 2.5 SD value; 3.0% of trials;
Robinson, in press). Analyses were conducted on log-transformed latencies, but means are reported
in milliseconds.
We performed a 2 (word type: immoral or moral) by 2 (word position: up or down) repeatedmeasures
ANOVA with psychopathy entered as a covariate (z-scores). The main effect of Word
Type was significant, F(1,110) = 78.99, p < 0.001, g2 p = 0.42. Participants were faster to categorize
the moral words (M = 919 ms; SD = 145 ms) than the immoral words (M = 968 ms; SD =
164 ms). The main effect of Word Position was significant, F(1,110) = 49.83, p < 0.001, g2 p = 0.31.
Participants were faster to categorize words appearing on top (M = 925 ms; SD = 153 ms) than
on the bottom (M = 963 ms; SD = 157 ms). The main effect of Psychopathy as well as the interactions
between Word Type and Psychopathy and Word Position and Psychopathy were non-significant,
Fs < 1.9.
The interaction between Word Type and Word Position was significant, F(1,110) = 11.23,
p = 0.001, g2 p ¼ 0:09. As shown in Table 1, participants were faster to categorize both word types
when words appeared near the top (vs. the bottom), but this effect was more pronounced for
words with a moral meaning. This interaction was qualified by a significant three-way interaction
among Word Type, Word Position, and Psychopathy, F(1,110) = 4.07, p = 0.046, g2 p ¼ 0:04. To

B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767 763
Table 1
Descriptive statistics for both experiments
Experiment Variable Statistic
M SD
1(N = 69) RT for moral-up/immoral-down 748 ms 150 ms
RT for moral-down/immoral-up 887 ms 202 ms
2(N = 112) RT for moral-up 895 ms 150 ms
RT for moral-down 943 ms 150 ms
RT for immoral-up 954 ms 165 ms
RT for immoral-down 982 ms 174 ms
LSRP 1.93 0.38
Note. LSRP = Levenson’s Self-Reported Psychopathy.
investigate this interaction, we performed a median split on the psychopathy scores and conducted
a Word Type by Word Position repeated-measures ANOVA for participants scoring low and high
in psychopathy (seven participants at the median were excluded).
For participants low in psychopathy, the main effect of Word Type was significant,
F(1,54) = 31.75, p < 0.001, g2 p ¼ 0:37, indicating that participants were faster to categorize the
moral (M = 906 ms; SD = 144 ms) than the immoral (M = 955 ms; SD = 165 ms) words. The
main effect of Word Position was significant, F(1,54) = 32.58, p < 0.001, g2 p ¼ 0:38, indicating that
participants were faster to categorize words appearing near the top (M = 909 ms; SD = 146 ms)
than the bottom (M = 952 ms; SD = 161 ms). These effects were qualified by a significant Word
Type by Word Position interaction, F(1,54) = 12.46, p = 0.001, g2 p ¼ 0:19. As shown in the top
portion of Fig. 1, both word types were categorized faster when they appeared near the top of
the screen, but this effect was more pronounced for the moral words.
For participants high in psychopathy, the main effect of Word Type was significant,
F(1,49) = 48.34, p < 0.001, g2 p ¼ 0:50, indicating that participants were faster to categorize the
moral (M = 934 ms; SD = 154 ms) than the immoral (M = 986 ms; SD = 171 ms) words. The
main effect of Word Position was also significant, F(1,49) = 16.50, p < 0.001, g2 p ¼ 0:25, indicating
that participants were faster to categorize words appearing near the top (M = 943 ms;
SD = 166 ms) than the bottom (M = 976 ms; SD = 162 ms). The interaction between Word Type
and Word Position was not significant, F(1,49) = 2.21, p = 0.144, g2 p ¼ 0:04. As shown in the bottom
portion of Fig. 1, both word types were similarly categorized faster when they appeared near
the top of the screen.
Experiment 2 partially replicated Experiment 1 in revealing that the speed to encode words with
a moral and immoral meaning varied by vertical position. While participants categorized both
word types faster when they appeared near the top, this effect was pronounced for moral words.
Furthermore, as hypothesized, participants’ psychopathy scores predicted the extent of this effect.
The nature of the interaction between word type and word position for participants low in psychopathy
was not exactly as expected. Participants were faster to categorize both word types when
they appeared near the top of the screen (although the effect was stronger for moral words). It
might be that vertical representation is only relevant for moral concepts. The main effect of word
position, however, may have obscured an effect for immoral concepts. Main effects for word posi-

764 B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767
Mean Categorization Time (ms) Mean Categorization Time (ms)
960
940
920
900
880
860
840
820
1020
1000
980
960
940
920
900
880
860
Moral Immoral
Word Type
Moral lmmoral
Word Type
Top of Screen
Bottom of Screen
Top of Screen
Bottom of Screen
Fig. 1. Categorization times as a function of word type and word position for individuals low (top panel) and high
(bottom panel) in psychopathy.
tion occur in this research (e.g., Schubert, 2005), possibly because people have a history of reading
text from top to bottom.
We wanted to determine what the interactive pattern between word type and word position
would look like if we could remove the influence of the main effects. That is, what pattern would
occur if we removed the tendency for people to be faster to moral words and to words that
appeared near the top? Rosnow and Rosenthal (1991) have argued that two by two interactions
should be interpreted by examining residual scores. Residual scores reflect the nature of an interaction
after removing the influence of significant main effects. Although we did not originally plan
on examining the residual scores, this method is justified in the present case because the main
effects were large, uninformative, and may have obscured the predicted interaction.
To obtain residual scores, we computed the two by two cell means, the grand mean, the row
effects, and the column effects. Then, we subtracted the row effect, column effect, and grand mean
from each cell mean leaving the residual scores (Rosnow & Rosenthal, 1991). As shown in Fig. 2,
these residual scores indicate that participants low in psychopathy were faster to categorize the
moral words when presented near the top, but they were faster to categorize the immoral words
when presented near the bottom.

4. General discussion
B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767 765
Residual Scores
10
8
6
4
2
0
-2
-4
-6
-8
-10
Moral Immoral
Word Type
Top of Screen
Bottom of Screen
Fig. 2. Residuals scores for the word type by word position interaction for participants low in psychopathy.
Morality is often explained in metaphors that use descriptions of verticality. Lakoff and Johnson
(1999) contend that metaphor is not simply a communication device, but it is used to represent
abstract knowledge. More generally, theorists contend that our conceptual systems are
embodied (Niedenthal et al., 2005). In two experiments, we found that people have implicit associations
between morality and vertical space (Experiment 1) and that this effect might be specific
to individuals low in psychopathy (Experiment 2). Our results have implications for metaphorrepresentation
and psychopathy.
4.1. Implications for metaphor-representation theory and psychopathy
Metaphor-representation theory contends that abstract-knowledge representation is aided by
conceptual metaphors (Lakoff & Johnson, 1999; Meier & Robinson, 2005). The current findings
support this view by revealing that moral concepts are encoded faster when paired with upper
visual cues or when presented in a high vertical location. More importantly, the current findings
extend metaphor-representation theory by showing that this effect was particular to people low
in psychopathy. These findings suggest that there are individual differences in metaphor representation.
Specifically, it could be that concepts not central to a specific personality dimension are not
represented to the same extent. Metaphor representation does not appear to be an inevitable occurrence,
but might be specific to individuals who frequently invoke a concept. Future research might
consider the role other personality dimensions (e.g., trait anger) play in the metaphorical-representation
of personality-specific concepts (e.g., anger is described in metaphors that use ‘‘heat’’).
Although psychopathy did not affect the speed in which moral and immoral words were categorized,
psychopathy did affect the extent to which the vertical dimension was used in these categorizations.
Individuals high in psychopathy do not appear to use verticality as a cue when encoding
moral-related concepts. These results complement research showing that psychopaths are deficient
in the implicit processing of affective stimuli (e.g., Kiehl, Hare, & McDonald, 1999; Williamson,
Harpur, & Hare, 1991). Because morality is an affectively-ladened concept, this deficiency appears
to extend to knowledge-representation tasks such as the present one.

766 B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767
4.2. Caveats and additional considerations
Why morality and verticality? Although we found that morality is partially represented on the
vertical dimension, our experiments do not provide a reason for this representation. The origination
of this pairing is beyond the scope of this article. We can, however, provide some speculation.
Considering cognitive development, Tolaas (1991) suggests that infants spend much of their time
lying on their back peering up at caregivers. This low vertical position necessarily means that
rewarding stimuli (e.g., food) and moral behavior (e.g., love) arrive from a high vertical position.
A child’s caregiver frequently appears in the infant’s upper visual-spatial environment. As cognition
matures, humans use this sensorimotor scaffolding to develop abstract thought, as recognized
by developmental psychologists (e.g., Piaget & Inhelder, 1969). This early pairing leads adults to
use the vertical dimension when expressing and representing morality (Lakoff & Johnson, 1999).
These considerations suggest that the association between morality and vertical position may
develop early within one’s sensory experiences (see Gibbs, 2006; for additional considerations).
Assessing psychopathy. We used the LSRP (Levenson et al., 1995) to assess psychopathy because
we wanted a quick and reasonably valid measure. Although Brinkley et al. (2001) showed that this
scale was moderately correlated with Hare’s (1991) PCL-R, it is clear that these measures are not
tapping exactly the same construct. In particular, Brinkley et al. (2001) showed that the LSRP and
the PCL-R do not assess the same thing in non-violent offenders. Furthermore, the LSRP may not
pick up on information that can be gathered via the interview-driven PCL-R. The LSRP does,
however, perform as well as the PCL-R in predicting the amount of errors committed during a passive-avoidance
cognitive task (Brinkley et al., 2001). In addition, Lynam, Whiteside, and Jones
(1999) showed that the LSRP predicted these same cognitive errors when examining college students.
Although the LSRP was a reasonable choice for the present cognitive investigation, future
research will be necessary to determine the extent to which psychopathy (as assessed by measures
like the PCL-R) moderates the relationship between morality and vertical position.
Morality, immorality, or both? The results of Experiment 2 suggest that individuals low (vs.
high) in psychopathy fall prey to vertical position when encoding moral concepts. This effect,
however, appeared to be specific to words with a moral meaning. As suggested earlier (and demonstrated
via residual scores), the main effect of vertical position may have obscured an effect for
the immoral concepts. Future studies might assess the vertical representation of moral/immoral
concepts in a manner that does not rely upon word stimuli. For example, one might use images
that depict moral and immoral acts that appear on the top and bottom of a screen. This design
may eliminate a main effect of position, which might allow a better test of the extent to which psychopathy
moderates the vertical representation of morality/immorality.
5. Conclusion
Two experiments revealed that people use the vertical dimension when processing moral-related
concepts and that psychopathy moderates this effect. Future research will be necessary to determine
if this effect is specific to psychopathy as measured by the LSRP (Levenson et al., 1995) or
whether it will generalize to more accepted measures (e.g., PCL-R; Hare, 1991). Moreover, additional
research will be needed to examine the impact of both moral- and immoral-related con-

cepts. Although we found that individual differences in psychopathy modify the embodied representation
of morality, future work will be needed to fully unpack the consequences psychopathy
(and other personality dimensions) has for embodied cognition and metaphor representation.
References
B.P. Meier et al. / Personality and Individual Differences 43 (2007) 757–767 767
Benning, S. D., Patrick, C. J., Hicks, B. M., Blonigen, D. M., & Krueger, R. F. (2003). Factor structure of the
psychopathic personality inventory: Validity and implications for clinical assessment. Psychological Assessment, 15,
340–350.
Brinkley, C. A., Schmitt, W. A., Smith, S. S., & Newman, J. P. (2001). Construct validation of a self-report
psychopathy scale: Does Levenson’s Self-report Psychopathy Scale measure the same constructs as Hare’s
psychopathy checklist-revised? Personality and Individual Differences, 31, 1021–1038.
Cleckley, H. (1941). The mask of sanity. St. Louis, MO: C.V. Mosby.
Edens, J. F., Marcus, D. K., Lilienfeld, S. O., & Poythress, N. G. (2006). Psychopathic, not psychopath: Taxometric
evidence for the dimensional structure of psychopathy. Journal of Abnormal Psychology, 115, 131–144.
Gibbs, R. W. Jr., (2006). Embodiment and cognitive science. New York: Cambridge University Press.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition:
The implicit association test. Journal of Personality and Social Psychology, 74, 1464–1480.
Haidt, J., & Algoe, S. (2004). Moral amplification and the emotions that attach us to Saints and Demons. In J.
Greenberg, S. L. Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 322–335).
New York: Guilford.
Hare, R. D. (1991). The Hare psychopathy checklist-revised. Multi-Health Systems: Toronto, Ont..
Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal Justice and Behavior, 23, 25–54.
Herve, H. F., Hayes, P. J., & Hare, R. D. (2003). Psychopathy and sensitivity to the emotional polarity of metaphorical
statements. Personality and Individual Differences, 35, 1497–1507.
Kiehl, K. A., Hare, R. D., & McDonald, J. J. (1999). Semantic and affective processing in psychopaths: An eventrelated
potential (ERP) study. Psychophysiology, 36, 765–774.
Krafft-Ebing, R. (1904). Textbook of insanity (C. G. Haddock, Trans.). Philadelphia: F.A. Davis.
Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied mind and its challenges to western thought. New
York: Basic Books.
Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing psychopathic attributes in a noninstitutionalized
population. Journal of Personality and Social Psychology, 68, 151–158.
Lynam, D. R., Whiteside, S., & Jones, S. (1999). Self-reported psychopathy: A validation study. Journal of Personality
Assessment, 73, 110–132.
Meier, B. P., & Robinson, M. D. (2004). Why the sunny side is up: Associations between affect and vertical position.
Psychological Science, 15, 243–247.
Meier, B. P., & Robinson, M. D. (2005). The metaphorical representation of affect. Metaphor and Symbol, 20, 239–257.
Niedenthal, P. M., Barsalou, L. W., Winkielman, P., & Krauth-Gruber, S. (2005). Embodiment in attitudes, social
perception, and emotion. Personality and Social Psychology Review, 9, 184–211.
Piaget, J., & Inhelder, B. (1969). The psychology of the child. New York: Basic Books.
Robinson, M. D. (in press). Lives lived in milliseconds: Using cognitive methods in personality research. In R. W.
Robins, R. C. Fraley, & R. Krueger (Eds.), Handbook of research methods in personality psychology. New York:
Guilford Press.
Rosnow, R. L., & Rosenthal, R. (1991). If you’re looking at the cell means, you’re not looking at only the interaction
(unless all main effects are zero). Psychological Bulletin, 110, 547–576.
Schubert, T. W. (2005). Your highness: Vertical positions as perceptual symbols of power. Journal of Personality and
Social Psychology, 89, 1–21.
Tolaas, J. (1991). Notes on the origin of some spatialization metaphors. Metaphor and Symbolic Activity, 6, 203–218.
Williamson, S., Harpur, T. J., & Hare, R. D. (1991). Abnormal processing of affective words by psychopaths.
Psychophysiology, 28, 260–273.