Considerable variation exists in the
contagiousness of yawning, and numerous studies
have been conducted to investigate the proximate
mechanisms involved in this response. Yet,
findings within the psychological literature are
mixed, with many studies conducted on relatively
small and homogeneous samples. Here, the authors
aimed to replicate and extend upon research
suggesting a negative relationship between
psychopathic traits and yawn contagion in
community samples. In the largest study of
contagious yawning to date (N_=_458), which
included both university students and community
members from across 50 nationalities,
participants completed an online study in which
they self-reported on their yawn contagion to a
video stimulus and completed four measures of
psychopathy: the primary and secondary
psychopathy scales from the Levenson Self-Report
Psychopathy Scale (LSRPS), the psychopathy
construct from the Dirty Dozen, and the
Psychopathic Personality Traits Scale (PPTS).
Results support previous findings in that
participants that yawned contagiously tended to
score lower on the combined and primary measures
of psychopathy. That said, tiredness was the
strongest predictor across all models. These
findings align with functional accounts of
spontaneous and contagious yawning and a
generalized impairment in overall patterns of
behavioral contagion and biobehavioral synchrony
among people high in psychopathic traits.
Résumé
Il existe des variations
considérables de la contagiosité
du bâillement, et de nombreuses
études ont été
menées pour étudier les
mécanismes immédiats
impliqués dans cette réponse.
Pourtant, les résultats de la
littérature psychologique demeurent
confus, après de nombreuses études
menées sur des échantillons
relativement petits et homogènes. Ici,
les auteurs ont cherché à
reproduire et à étendre les
recherches suggérant une relation
négative entre les traits psychopathiques
et la contagion du bâillement. Dans la
plus grande étude sur la contagion du
bâillement à ce jour (N_=_458), qui
comprenait à la fois des étudiants
d'universités et un public de 50
nationalités différentes, les
participants ont complété une
étude en ligne dans laquelle ils ont
autodéclaré leur contagion du
bâillement à un stimulus
vidéo et ont complété
quatre mesures de la psychopathie : les
échelles de psychopathie primaire et
secondaire de l'échelle de psychopathie
d'auto-évaluation de Levenson (LSRPS), la
construction de psychopathie de la Dirty Dozen
et l'échelle des traits de
personnalité psychopathiques (PPTS). Les
résultats corroborent les conclusions
précédentes selon lesquelles les
participants qui bâillaient de
manière contagieuse avaient tendance
à obtenir des scores inférieurs
aux mesures combinées et primaires de la
psychopathie. Cela dit, la fatigue était
le prédicteur le plus fort dans tous les
modèles. Ces résultats
correspondent aux critères des
bâillements spontanés et contagieux
et à une réduction de la
sensibilité à la contagion
comportementale et de la synchronie
bio-comportementale chez les personnes
présentant des traits psychopathiques
élevés.
The automatic and reflexive tendency to yawn
in response to sensing the yawns of others,
i.e., contagious yawning, is well-documented in
humans1,2, and has been observed in a growing
number of non-human species including non-human
great apes3,4, dogs5, pigs6, and birds7,8, among
others. Unlike spontaneous yawning, which begins
in utero9, is ubiquitous across vertebrates10,
and appears to be a human universal11,
contagious yawning does not emerge until early
childhood12,13, is limited to social species14,
and psychological studies reveal considerable
variability in the tendency for people to yawn
contagiously. In particular, the percentage of
participants that yawn in response to video
and/or audio stimuli of others yawning typically
ranges between 30 and 60%
(e.g.1,2,15,16,17,18,19).
This variability in yawn contagion has drawn
considerable attention over the last two decades
(e.g.2,20,21), with numerous published works
examining factors that contribute to this
response. Overall, the primary emphasis of
research in this area has been to explore
associations between contagious yawning and
empathy or emotional contagion (reviewed
by22,23). The proposed link between contagious
yawning and empathic processing originates from
a monograph by Lehmann24, and has been
elaborated more recently in its inclusion in the
Perception&endash;Action-Model (PAM) proposed by
Preston and de Waal25,26. Accordingly, the motor
mimicry of yawn contagion results from a
perception&endash;action mechanism that permits
the rapid synchronization of states between
individuals27. Extending from the PAM,
contagious yawning has been proposed to
represent a basic form of emotional contagion,
whereby the yawns, and accompanied emotional or
mental state of the yawner, are passed on to
another individual22. Similarly, automatic
facial mimicry is hypothesized to enhance
emotional recognition28, and has been considered
to be a critical feature of emotional
contagion29. While the motor action pattern of
yawning is clearly contagious, as referenced by
works above, this theory lacks empirical
evidence when it comes to the transfer of
emotional states during yawn contagion23.
Moreover, in general, the role of facial mimicry
in the transfer of emotional states is not
unequivocal30. Therefore, this phenomenon might
be better explained by simple behavioral
contagion or facial mimicry31,32,33. In contrast
to emotional contagion, whereby the synchrony of
behaviors and emotions are neurologically
linked, simple behavioral contagion represents
the copying of the behavior itself. Yawn
contagion could be adaptive in the absence of
emotional coupling in facilitating collective
vigilance and coordinated group behavior23.
Nonetheless, the idea that contagious yawning is
perhaps reflective of empathy or emotional
contagion has drawn considerable interest and
investigation.
Platek et al.2 garnered initial support for
a connection between contagious yawning and
mental state attribution (a form of cognitive
empathy) among samples of university students.
These authors found that contagious yawning was
positively correlated with performance on
self-face recognition and faux pas theory of
mind tests, and negatively correlated with
measures of schizotypal personality traits.
Since then, psychological studies examining
associations between individual differences on
empathy measures and yawn contagion have
revealed mixed findings (e.g.20,21,34,35,36),
with most studies reporting no clear
relationship (reviewed by23). For example, in
what was the largest study at the time
(N_=_328), Bartholomew and Cirulli20 found
highly consistent individual differences in yawn
contagion to video stimuli in an online format,
but when accounting for the age of the
participants, yawn contagion was not related to
empathy as measured by the Interpersonal
Reactivity Index (IRI)37. More recently,
however, Franzen et al.21 conducted two large
laboratory studies (N_=_171; N_=_333) and found
that participants who yawned contagiously
reported significantly higher empathy levels as
measured by the IRI. Yet, in a smaller study
(N_=_97) published recently, contagious yawning
was unrelated to the total IRI score, or any of
the subscales (cognitive: perspective taking and
fantasy; affective: empathic concern and
personal distress)19. In sum, results from
studies linking individual differences in yawn
contagion to measures of empathy remain
inconsistent.
As an indirect measure of empathy, studies
have also examined ingroup or familiarity biases
in yawn contagion. Observational studies suggest
that naturalistic instances of yawn contagion
appear to occur more regularly among friends and
family compared to acquaintances and
strangers38,39. Yet, enhanced attention towards
familiar individuals and the tendency to avoid
the gaze of strangers or less familiar
acquaintances, could have contributed to this
effect40. To date, the only experimental
research to examine in-group or familiarity
biases in yawn contagion among humans failed to
show an effect41. Researchers have also explored
gender differences in yawn contagion, as there
are marked differences in empathy between men
and women42. In support of this notion, Norscia
and colleagues43 reanalyzed data from the
aforementioned observational study38 and found
that women yawned contagiously more so than men.
These authors have also noted a similar effect
for auditory yawn contagion39. However, an
analysis across the broader literature
demonstrated no differences in contagious
yawning between men and women44.
Another approach to assessing the link
between contagious yawning and empathy has been
to study clinical populations with deficits in
empathy and to assess variability in other
psychological attributes predictive of empathic
processing in non-clinical populations. This
area of research has led to the examination of
(1) group differences in contagious yawning
among children with autism spectrum disorder
(ASD), and (2) correlational studies measuring
individual differences in psychopathic traits
and the susceptibility to yawn contagiously.
Individuals with ASD tend to be characterized by
impairments in cognitive empathy or perspective
taking (i.e., theory of mind), while
psychopathic traits are associated with reduced
affective empathy, including diminished primary
emotions such as fear and sadness45.
Interestingly, both ASD and the prevalence of
psychopathic traits are more common among
males46,47. These populations are also of
interest to study with regards to the
hypothesized functional significance of yawn
contagion in promoting collective vigilance and
synchronized group
behavior/movement48,49,50,51,52,53. Impairments
in imitation and joint attention among
individuals with ASD diminish cooperation and
coordinated actions54. Individuals with ASD also
show diminished facial mimicry55,56 and less
synchronization during interpersonal
coordination57. Similarly, psychopathy is marked
by deceit and a lack of cooperation58, and
individuals that score high on
callous-unemotional traits also display reduced
facial mimicry30,59 and diminished group
cohesion60. In addition, a general impairment in
biobehavioral synchrony has been implicated in
psychopathy61.
The first such study on contagious yawning
involving individuals with ASD was conducted by
Senju et al.62. Using standard laboratory
procedures, i.e., the presentation of video
stimuli with people yawning, children with ASD
showed diminished yawn contagion&emdash;an
effect that garnered significant interest in the
scientific community and was subsequently
replicated by Giganti and Ziello63 and Helt and
colleagues64. While initially taken as strong
support for a link between contagious yawning
and empathy, further work refined this view
showing that the diminished contagion among
children with ASD results from reduced attention
to yawning stimuli. When instructed to focus
their attention on the eyes of the yawning
target in the stimuli, individuals with ASD
displayed rates of contagious yawning equivalent
to aged-matched control samples (e.g.18,65).
Moreover, when using eye-tracking to confirm
that visual attention is allocated to the
stimuli during testing, both typically
developing children and those with ASD yawn more
often to depictions of yawning than to control
clips66. In another study, Mariscal et al.67
measured contagious yawning along with blood
oxytocin levels from both children with ASD and
those that were typically developing. Again, no
group differences in yawn contagion were found,
i.e., children with ASD yawned just as often as
typically developing children. However, a
positive relationship was reported between
contagious yawning and oxytocin among children
with ASD, while no such relationship was present
in the control group. Although links between
contagious yawning and oxytocin continue to be
discussed68, attempts to experimentally
manipulate oxytocin through intranasal
administration have found no effect on yawn
contagion69,70.
To date, only two studies have examined the
link between contagious yawning and psychopathic
traits. Rundle, Vaughn, and Stanford71 had
university students (N_=_135) from the United
States complete the Psychopathic Personality
Inventory-Revised (PPI-R)72 and watch a series
of yawning videos to obtain physiologically
defined measures of yawn contagion using facial
electromyography and galvanic skin response.
When comparing participants that did and did not
yawn, there was no difference in overall scores
on the PPI-R. However, when examining the
individual subscales of the PPI-R, which include
fearless dominance, self-centered impulsivity,
and cold-heartedness, the latter predicted yawn
contagion. That is, participants who scored
higher on measures of cold-heartedness were less
likely to yawn contagiously. Males from the
sample (N_=_57) also took part in an eyeblink
startle paradigm, and consistent with a link to
psychopathy, contagious yawning was less common
in men with a lower startle response71. The
negative association between contagious yawning
and psychopathic traits was also recently
examined by Helt et al.19. Drawing from a
slightly smaller sample of university students
in (N_=_97) also from the United States,
participants completed the PPI-R, the Autism
Spectrum Quotient (AQ), and the IRI (see above).
Unlike Rundle et al.71, a negative relationship
was revealed between the combined PPI-R and
video confirmed yawn contagion (yes/no), i.e.,
participants that scored higher on psychopathic
traits were less likely to yawn contagiously,
though no analyses were conducted across the
subscales of this measure. By employing
eye-tracking, the researchers were also able to
show that the negative relationship between
contagious yawning and scores on the PPI-R was
not moderated by visual attention. A similar
negative relationship was found between the AQ
and contagious yawning, but unlike measures of
psychopathy, the relationship between yawn
contagion and the AQ was moderated by eye
gaze19. Lastly, contagious yawning was not
significantly correlated with total empathy or
any of the subscales of the IRI.
Based on the overall mixed findings within
the psychological literature, and the relatively
limited investigation into the link between
psychopathic traits and yawn contagion in
particular, the current study aimed to provide
resolution to this anticipated association. In
particular, this investigation sought to
replicate and extend upon the findings from
Rundle et al.71 and Helt et al.19, but with a
larger and more heterogeneous sample of online
participants. Four distinct measures of
psychopathy were included to assess the
generalizability of this association, and given
the importance of physiological variables in
influencing yawn contagion
(e.g.15,48,73,74,75,76,77,78), measures of sleep
and fatigue were taken into consideration.
Lastly, given recent debates on the importance
of attention in contributing to yawn
contagion22,23,34,65, both objective and
subjective measures of attention to the
contagious yawning video stimulus were
included.
Discussion
In a large-scale conceptual replication of
two previous publications19,71, the findings
from this study demonstrate a small, but
significant negative relationship between yawn
contagion and psychopathic traits in a community
sample. While these previous studies included a
single measure of psychopathy (PPI-R) among
university student populations in the United
States, the current study assessed the
generalizability of this relationship across
different scales in a much larger and more
heterogenous sample. Consistent with previous
reports, we show that participants who failed to
show contagious yawning (37.3%) tended to score
higher on various measures of psychopathy,
including the primary psychopathy scale of the
LSRPS79, the psychopathy construct of the Dirty
Dozen80, and the combined PPTS81. However, the
effect sizes from these comparisons were small
(Rank biserial
correlations_=_0.126&endash;0.141). When
considering participant gender, age, prior
sleep, current tiredness, and objective and
subjective measures of attention, only the
primary psychopathy scale of the LSRPS and the
combined PPTS remained significant, with
tiredness of participants serving as the best
predictor of yawn contagion across all models.
Consistent with a previous review of the
literature44, and replicating the findings of
Rundle et al.71, we show no difference in
contagious yawning between men and women.
Given that psychopathy is characterized by
deficits in emotional empathy, or the ability to
be affected by&emdash;or share&emdash;the
emotional states of others45, the impetus for
examining the connection between psychopathic
traits and contagious yawning has focused on
purported links between contagious yawning and
empathy19,71. In particular, Rundle et al.71
found that scores on the social-emotional
component of the PPR-I, i.e., cold-heartedness,
negatively predicted yawn contagion. Since
cold-heartedness represents the inconsideration
of the emotional state of others72, this result
was consistent with the notion that contagious
yawning represents a basic form of emotional
contagion22,24,25,26. In line this view, Helt et
al.19 found an inverse relationship between
combined PPR-I scores and yawn contagion, and
that the overall measure of psychopathy was
negatively correlated with personal distress in
social interactions. Based on these findings,
the authors proposed that callous and
unemotional traits producing less personal
distress might contribute to the inability to
achieve bodily resonance with others, and that
diminished yawn contagion is due to a
malfunctioning of the empathic system.
With regards to the current study, the
primary psychopathy scale of the
LSRPS&emdash;which was the best predictor of
contagious yawning from all psychopathy
measures&emdash;captures indices of individual
callousness, egocentricity, manipulation,
selfishness, and deceit79, and is thus in line
with these previous reports. The four questions
within the psychopathy construct of the Dirty
Dozen measure these dimensions as well80, and
individuals that scored higher on this
instrument were also less likely to yawn
contagiously. In addition, the results revealed
a significant negative relationship to the
combined PPTS, which also emphasizes primary
psychopathic personality characteristics81. Low
internal consistency precluded the assessment of
the affective responsiveness subscale of the
PPTS, which is most closely related to the
cold-heartedness subfactor of the PPI-R found to
be significant in Rundle et al.71. However, the
results did reveal a significant negative
relationship to the combined PPTS. The secondary
psychopathy measure of the LSRPS was the only
instrument unrelated to yawn contagion. Unlike
the other measures, secondary psychopathy is
associated with antisociality, anxiety-driven
impulsivity, and irresponsibility, and reflects
more of an antisocial-behavioral dimension of
psychopathy82. In particular, the PPTS was
designed with the intention to place a greater
emphasis on personality rather than behavior81.
Thus, yawn contagion appears to be largely
associated with affective-interpersonal
personality characteristics.
These findings are consistent with the view
that contagious yawning is associated with
empathy or emotional contagion22. In line with
this perspective, and as it relates to the
current findings, prior works have shown that
adolescent males that score high on
callous-unemotional traits both score lower on
self-report empathy and have reduced facial
mimicry, as measured by electromyography, to
empathy-inducing video clips59 and dynamic
emotional expressions30. However, it is
important to acknowledge that psychopathic
traits only represent an indirect measure of
empathy, such that yawn contagion could merely
reflect a simple feature of behavioral contagion
or facial mimicry31,32,33. While empathy
represents a superordinate category of PAM25,26,
features explained by this model, and subclasses
of phenomena with the same mechanism, such as
imitation and state matching, do not require
empathy. This is similar to the Russian-doll
model for the evolution of empathy, whereby
components of empathy are layered on top of and
dependent upon one another27. In this case,
motor mimicry is at the core within a
perception&endash;action mechanism, but in and
of itself does not reflect empathy.
Therefore, as an alternative interpretation,
the characteristic features of psychopathy might
represent an impairment in attachment, social
affiliative behaviors, and social connectedness
that affects more general features of behavioral
contagion61. Consistent with the view, Helt et
al.19 found a similar negative relationship
between psychopathic traits and itch contagion,
which is not implicated in empathic processing.
In addition, a growing number of studies have
found that yawn contagion does not correlate
significantly with measures of empathy20,35,36.
In an attempt to delineate between the role of
emotional and behavioral contagion in the
contagious yawning reflex, Chan and Tseng34
compared the effects of empathy, emotional
processing, and detection sensitivity to yawning
in typically developing populations, finding
that only the latter predicts contagious
yawning. This same study found that people more
likely to detect yawning, but not emotional
expressions, are most likely to show contagious
yawning. As it relates to group dynamics,
psychopathy is thought to originate from
disrupted biobehavioral synchrony61, which
represents the coordination of biological and
behavioral processes during social contact and
is considered critical for attachment,
affiliative bonds, and promoting survival
activities in groups83. Thus, the reduced
tendency to yawn contagiously among individuals
scoring high in psychopathic traits could
reflect a generalized impairment in
biobehavioral synchrony. Moreover, it has been
theorized that contagious yawning evolved to
enhance collective vigilance48,49 and coordinate
or synchronize group behavior50,51,52,53, and
these functional accounts for yawn contagion can
explain the negative association between
contagious yawning and psychopathy in the
absence of emotional contagion. For example,
groups characterized by higher levels of
psychopathy have previously been shown to have
more dysfunctional interactions and lower levels
of cohesion60.
An often-neglected area of study when
examining variation in contagious yawning
concerns underlying physiologic and circadian
factors that contribute to non-social (i.e.,
spontaneous) yawning. Spontaneous yawns
represent the primitive feature of this
stereotyped action pattern, and are
physiologically driven. Contagious yawning,
however, represents a derived characteristic
that has appeared more recently among a select
group of social species14. Since the action
patterns of both yawn types appear
indistinguishable, the neurological mechanisms
governing yawn contagion are likely mediated by
the same physiological variables affecting
patterns of spontaneous yawning. In support of
this view, and consistent with the hypothesis
that the motor action pattern of yawning evolved
to increase arousal and state change1,84 via
intracranial circulation and brain cooling85,86,
socially-elicited forms of contagious yawning
can be modulated by different methods of
breathing48, time of day74, cooling/heating to
the surface of forehead and neck48,78, chewing
on gum76, ambient temperature variation and
seasonal climactic conditions15,73,77, and acute
physical stress75. Collectively these studies
support the view that spontaneous and contagious
forms of yawning share fundamental mechanistic
pathways87. In line with this perspective,
previous research has shown that both
spontaneous and contagious forms of yawning are
positively correlated with subjective ratings of
sleepiness74,88, and the best predictor of
contagious yawning here was self-reported
tiredness during testing. Thus, when assessing
variation in yawn contagion, we propose more
attention should be given to a combination of
both psychological traits as well as internal
physiologic states. Franzen et al.21 provides a
good example of this approach, measuring pulse
rate as an indicator of sleepiness during
contagious yawning trials.
With regards to the overall levels of
psychopathy reported in this study, the mean
scores and measures of variability are similar
to previous community samples79,80. Previous
studies have also used these instruments within
incarcerated populations. In comparison to a
sample of male inmates in a minimum-security
prison, the current levels on the LRSPS were
appreciably lower, particularly for the primary
measure89. In a separate study90, which used a
different Likert scale, psychopathy levels on
the Dirty Dozen were also significantly higher
among an incarcerated population compared to a
community sample. The PPTS was designed to
measure psychopathic personality regardless of
respondents' criminal history, but was initially
measured in two incarcerated samples81,91. In
this case, the levels in these previous studies
were quite similar to those reported here. Thus,
future research could examine whether the
negative association between psychopathy and
contagious yawning would replicate in clinical
or incarcerated populations.
While the advantages to this study include
the large sample size and relatively diverse
composition, there remain limitations. The
reliance on self-reported contagious yawning,
rather than recording physiologic- or
video-confirmed instances of yawning, as was the
case in Rundle et al.71 and Helt et al.19, is a
notable limitation. Unfortunately, the Covid-19
pandemic dictated the online format for this
study, and prevented the possibility to obtain
other measures of yawn contagion. That said,
there are two reasons to remain confident in the
findings presented here. First, the current
results largely replicate earlier studies, an
effect not expected if we had obtained
inaccurate measures of contagious yawning.
Second, prior research has thoroughly
demonstrated that self-report is a valid measure
of contagious yawning based on substantial
agreement between self-report and
video-confirmed yawns41,69,92. That said, only a
binary measure of contagious yawning was
assessed, and capturing self-reported yawn
frequency could have provided an important
source of variability. Another limitation to
this study was the lack of a non-yawning control
condition. As a result, we were unable to
distinguish between spontaneous yawns that could
have been reported as contagious. However, we do
not view this as a major concern given prior
research has shown a very low rate of
spontaneous yawning in response to control
stimuli (3.3%)21. Overall, this study also
shares limitations with most research on
contagious yawning with regards to the external
validity of the contagion stimuli93. In this
case, participants were shown 50 consecutive
yawning clips in short duration and without
sound. Thus, in the future, researchers could
work to develop and incorporate more
ecologically valid stimuli for contagious
yawning research.
Our measures of attention were also a
limitation. We found no relationship between
contagious yawning and the subjective measure of
attention to the video stimulus, and an
unexpected inverse association between
contagious yawning and the objective measure of
attention. That is, participants that correctly
recalled more details from the video were less
likely to report contagious yawning. These
findings are difficult to reconcile with
previous studies showing a significant effect of
visual attention to yawning stimuli on yawn
contagion19,65. It is possible that rather than
providing an index of attention towards yawning,
as it was intended, our measure of objective
attention inadvertently captured participant
attention towards other aspects of the video,
such as the running tally of the total number of
yawns in the upper-righthand portion of the
screen. Thus, participants scoring higher on
this measure may have been more attuned to
outside features and actually spent less time
focused on yawns, which could explain the
pattern of results here. Given these
limitations, and the continued debate over the
roles of visual attention compared to
social-connectedness in contributing to
contagious yawning22,34,40,65, it is recommended
that future studies use eye-tracking to assess
visual attention to different features of the
yawning stimulus19.
In summary, the current results, which to
our knowledge come from the largest and most
diverse sample in the study of contagious
yawning to date, show a small, but significant
negative relationship between psychopathic
traits and the susceptibility to yawn
contagiously. Limitations notwithstanding, this
study provides a conceptual replication of
previous research using multiple measures of
psychopathy, suggesting that variability along
this dimension is a true and reliable
contributor to individual differences in
contagious yawning in the general population.
While these findings are consistent with an
indirect link between contagious yawning and
empathy, they also align with functional
accounts of spontaneous and contagious yawning
and a generalized impairment in overall patterns
of behavioral contagion among people high in
psychopathic traits19. Given the mixed and
inconsistent psychological literature on
contagious yawning, we hope this research spurs
further replication efforts within this
area.
