Contagious yawning occurs in humans and a
few other highly social animals following the
detection of yawns in others, yet the factors
influencing the propagation of this response
remain largely unknown. Stemming from earlier
laboratory research, the authors conducted five
experiments to investigate the effects of social
presence on contagious yawning in virtual
reality (VR).
They show that, similar to a traditional
laboratory setting, having a researcher present
during testing significantly inhibited
contagious yawning in VR, even though
participants were viewing a virtual environment
and unable to see the researcher. Unlike
previous research, however, manipulating the
social presence in VR (i. e., embedding
recording devices and humanoid avatars within
the simulation) did not affect contagious
yawning.
These experiments provide further evidence
that social presence is a powerful deterrent of
yawning in humans, which warrants further
investigation. More generally, these findings
also have important applications for the use of
VR in psychological research. While participants
were quite sensitive to social stimuli presented
in VR, as evidenced by contagious yawning, their
results suggest a major difference in the
influence of social factors within real-world
and virtual environments. That is, social cues
in actual reality appear to dominate and
supersede those in VR.
Résumé
La contagion du bâillement existe chez
l'homme et quelques autres animaux sociaux
après la perception du bâillement
d'un congénère, mais les facteurs
qui influencent la propagation de cette
réponse restent en grande partie
inconnus. Issus de recherches antérieures
en laboratoire, les auteurs ont mené cinq
expériences pour étudier les
effets d'une présence sur la contagion du
bâillement en réalité
virtuelle (VR).
Ils montrent que, situation habituelle en
laboratoire, la présence d'un chercheur
au cours des essais a considérablement
inhibé la contagion du bâillement
en VR, alors même que les participants
visionnaient un environnement virtuel et ne
pouvaient pas voir le chercheur.
Contrairement aux recherches
précédentes, cependant, manipuler
une présence dans la
réalité virtuelle
(c'est-à-dire incorporer des appareils
d'enregistrement et des avatars humanoïdes
dans la simulation) n'a pas eu d'effet sur la
contagion du bâillement. Ces
expériences fournissent une preuve
supplémentaire qu'une présence est
un puissant moyen d'inhibition du
bâillement chez l'homme, ce qui
mérite une enquête plus
poussée.
Plus généralement, ces
résultats ont également des
conséquences importantes pour
l'utilisation de la RV dans la recherche
psychologique. Bien que les participants aient
été très sensibles aux
stimuli sociaux présentés dans la
réalité virtuelle, comme en
témoigne un bâillement contagieux,
les résultats suggèrent une
différence majeure entre l'influence des
facteurs sociaux dans les environnements
réels et virtuels. Les stimuli sociaux
réels semblent supplanter ceux du monde
virtuel.
Andrew
C. Gallup. Yawning and the thermoregulatory
hypothesis
All
articles about contagious
yawningYawning is characterized by a
powerful gaping of the jaw with deep
inspiration, followed by a temporary period of
peak muscle contraction and a passive closure of
the jaw during expiration. Physiologically,
yawns enhance intracranial circulation and
facilitate brain cooling, which in turn could
serve to promote cortical arousal6 and state
change during behavioral transitions. Contagious
yawning, which represents the reflexive
triggering or release of this response as a
result of sensing yawns in others, is a
well-documented phenomenon in humans, as seeing,
hearing and even thinking about yawning can
induce yawn contagion. Evidence for contagious
yawning is also present in a small number of
non-human animals, including some non-human
primates, birds and domesticated dogs.
Spontaneous and contagious yawns represent
intrinsically connected, yet distinct behaviors.
Although indistinguishable in the motor action
pattern described above, a number of important
factors, in addition to their triggers (i. e.,
physiological vs. social), differentiate these
two types of yawns. Spontaneous yawns, or
similar mandibular-gaping patterns, appear to be
a phylogenetically old and conserved across
vertebrate classes, while contagious yawning is
a more recently derived feature present in only
a few highly social species. Consistent with
this view, these two responses show distinct
ontogenies, with spontaneous yawns emerging
early on within intrauterine development in
humans, while contagious yawns do not appear
until early childhood. Furthermore, spontaneous
yawns seem to be a universal act across members
of a given species, whereas the expression of
yawn contagion appears to show much more
individual variability. For example, depending
upon the methods and stimuli used, only ~30-60%
of people show contagious yawning across
laboratory and semi-naturalistic settings, with
similarly variable response rates observed for
non-human primates.
The fundamental differences between these
two yawn-types have led some researchers to
propose that yawn contagion may reflect a form
of higher-level social-cognitive processing (e.
g., emotional contagion or state matching).
Indeed, a large body of research has explored
the proposed connection between yawning and
empathy, with lines of supporting evidence
coming from correlational studies,
in-group/out-group comparisons, neuroimaging
investigations, and clinical studies. Consistent
with this view, one study even showed a negative
association between contagious yawning and
measures of psychopathy in a non-clinical
population. However, a recent review of this
literature concluded that the empathy/contagious
yawning link is weak and inconsistent, with the
majority of studies failing to observe the
predicted effects derived from this empathy
modeling hypothesis. In fact, recent experiments
have shown that the susceptibility to contagious
yawning among humans is tied to the perceptual
encoding of the act in others and is unrelated
to psychological traits linked with empathy or
emotional processing. Others have argued that it
is not necessary to implicate a link between
contagious yawning and higher-order cognitive
processes. For example, given the physiological
consequences of this motor action pattern (i.
e., enhanced intracranial circulation and brain
cooling), the spreading of this response through
contagion could have evolved to coordinate group
arousal or activity patterns and promote
collective vigilance. Further research in these
areas is needed, but the fact that individual
differences in contagious yawning may predict
important aspects of psychological and
perceptional functioning deserves further
attention.
Despite the involuntary and impromptu nature
of yawns, and the potential benefits from this
action, the expression of yawning is often
stigmatized in social settings. In fact, yawns
are interpreted as a sign of boredom and
disrespect across different cultures. This
seemingly widespread, negative public perception
surrounding yawning is presumably a result of
the temporal associations between reduced states
of alertness and sleep/wake cycles, i. e., yawns
are known to occur when individuals are bored or
drowsy so this act might indicate one's
disinterest or reduced mental status. Some
researchers have therefore speculated that
yawning serves a primary communicative role, in
which yawns signal internal states to others.
However, a central signaling perspective fails
to take into account that spontaneous yawns are
widespread among solitary species, and often
occur when alone even among gregarious animals.
Moreover, yawns are triggered by a multitude of
factors, and are associated with a markedly
variable array of contexts, stimuli and internal
states (i. e., not just when we are bored or
sleepy), and as a result could not serve as
reliable signals. Nonetheless, the presence of
others, i.e., audience or mere presence effects,
tends to diminish the expression of yawning.
Early research has shown that spontaneous yawns
are indeed less common among people in crowded
environments, as well as when being observed by
a researcher in a laboratory. It is unclear
whether the reduced tendency to yawn in these
settings is due to the negative signaling value
of this response to others (i. e., a top-down
mechanism), or from heightened arousal levels
produced by these contexts that would act to
naturally reduce yawning.
Similar socially-modulated effects have been
observed for the expression of contagious
yawning. The administration of intranasal
oxytocin, a hormone and neuropeptide that
enhances social perception and awareness, has
been shown to increase the likelihood that
people actively inhibit and conceal their yawns,
while also decreasing the tendency for any
expressed yawns to be accompanied by overt cues
of the response. In comparison to those
receiving a placebo saline solution,
participants administered 30 IU of intranasal
oxytocin were more likely to stifle their urge
to yawn and required a greater length of
exposure to a contagious stimulus prior to
yawning. Moreover, among those that actually
showed yawn contagion, individuals that received
oxytocin were more likely to cover their mouth
and clench their jaw during the act of yawning,
and were less likely to display bodily
stretching and produce auditory sighs during
this response.
In a more recent and targeted investigation
of audience effects on yawn contagion, we
explicitly examined how varied degrees of social
presence in a laboratory setting altered
contagious yawning. Given the wide array of
stimuli that elicit social presence effects, we
tested how both implied (i. e., a "recording"
webcam facing the participants) and actual
social presence (i. e., a researcher in the
testing room) influenced contagious yawning to
video stimuli presented on a computer monitor.
The results from this experiment showed that
both implied and real social presence
significantly diminished the self-reported
expression of yawn contagion in comparison to
control settings (e. g., a webcam that is off or
a testing room that does not contain a
researcher). Participants in the "recording"
webcam or researcher present conditions also
reported a greater urge to yawn without doing so
when viewing the video stimuli. Thus, the
prospect, or actual presence, of others watching
seems to inhibit even this inherently social
form of yawning.
As an extension of this line of research,
here we conducted five experiments to
investigate the effects of real-world social
presence on contagious yawning in virtual
reality (VR). Although not without its
limitations, utilizing VR in experimental
psychology and the cognitive sciences offers
numerous advantages, providing greater levels of
control and manipulation over traditional
experimental procedures/settings. Furthermore,
the capability of VR to potentially block out
the external environment and engage the audience
with its own stimuli allows for a level of
immersion that typical displays cannot provide.
This is the first study to our knowledge which
leverages VR to study the factors influencing
yawning.
We first assessed how actual social presence
in the testing environment (i. e., a live
researcher present) altered contagious yawning
in VR, and then how both implied and actual
forms of social presence within the virtual
environment (i. e., embedding recording devices
and humanoid avatars within the scene) modified
this response. In particular, the first
experiment compared participant responses to
contagious yawning stimuli presented in VR
between conditions where a researcher either
indicated that they would be sitting in the
testing room during the virtual testing or not.
The second and third experiments examined how
the presence of an implicitly or explicitly
recording webcam embedded within VR altered
contagious yawn response rates within this
setting. In the second experiment, the virtual
webcam was either accompanied by a red light or
not, while in the third experiment the
researcher specifically instructed participants
in advance that the webcam in the simulation
would be recording their behavior during
testing, or there was no mention of this device.
The fourth experiment assessed how the presence
of a motionless humanoid avatar within the scene
influenced contagious yawning responses in VR.
In particular, contagious yawning was compared
between conditions when this figure was directly
facing and looking towards the participants or
was oriented 180 degrees in the opposite
direction facing away. Lastly, the fifth
experiment examined how contagious yawning was
influenced by the presence of an onlooking
humanoid avatar using a more realistic and
representative virtual simulation. In
particular, the humanoid avatar in this
experiment swayed slightly and moved its limbs
in a realistic way, mimicking natural behavior,
and the virtual scene was designed to represent
a more typical laboratory environment similar to
our past research with a smaller presentation
screen positioned on a desktop. Thus, the design
of this experiment allowed us to test the effect
of avatar realism and display/room size and
appearance. This final experiment also measured
the immersive properties of the VR experience
(including a sense of presence) to assess
whether this varied between conditions and
influenced yawning rates. In short, our
operationalization of social presence both in
the real world and in VR involved reproducing
stimuli that are known to affect performance in
actual reality (e. g., an active webcam, a
person) in a VR environment. Note, however, that
whether these stimuli will have the same effect
in VR as in actual reality is very much an open
question as the social presence of items in VR
may require some level of psychological
involvement and/or behavioral engagement.
To the extent that placing a participant in
a VR situation diminishes awareness or concern
regarding the world beyond the VR setting, by
inducing a mental state in which the user feels
as though they are present within and can act
upon the computer environment, we made the
following predictions based on the conditions of
each experiment. For Experiment 1, unlike Gallup
et al., we predicted that the presence of a
researcher in the testing room would be
equivalent to having no researcher in the room.
In contrast, we predicted that placing a
recording webcam in the VR environment
(Experiment 2) or explicitly referring to the
recording VR webcam (Experiment 3), would create
an implied social presence and reduce the number
of yawns and/or increase the urge to yawn
without doing so. Similarly, we predicted that
the actual presence of a humanoid avatar in the
VR setting or having that person look at the
participant (Experiments 4 and 5) would yield
behaviors indicative of yawn inhibition, i.e.,
reduced yawn frequency and/or an increased urge
to yawn.
Conditions where others are around and
potentially watching often changes the behavior
of humans and non-human animals. This report
sought to extend previous research examining
these so-called audience, or social presence,
effects on the expression of contagious yawning
in humans. By utilizing VR to study yawn
contagion, we investigated the effects of both
actual and simulated real-world social presence
manipulations on this response through five
separate experiments.
Based on previous research we had thought
that placing participants within a VR
environment would diminish their awareness of
actual reality, and with it, abolish our
previous finding that a researcher in the
testing room reduces the number of yawns people
produce and increases the tendency that they
fought off the urge to yawn. The opposite result
was obtained. Replicating, and extending our
previous findings, we found that even when a
person is placed in a VR environment, the real
presence of a single researcher is sufficient to
diminish the expression of contagious yawning in
this setting. In other words, yawning behavior
was affected by the live researcher presence
even though participants were exposed to the
virtual environment and unable to see this
individual. In sum, Experiment 1 yielded an
unexpected discovery that the presence of a
researcher in the testing room significantly
inhibits subsequent yawn contagion in VR, even
though participants are occluded from the
researcher and within a virtual environment when
presented with the yawning stimuli.
One explanation for this unexpected outcome
is that VR setup was not sufficient to shield
participants from the real context and thus the
social presence on the outside permeates their
VR experience and alters socially contagious
responses within. In light of the fact that
Gallup et al. found that an implied social
presence in the form of a live recording webcam
is sufficient to inhibit yawning behaviors, we
reasoned that a recording webcam placed within
the participant's virtual world would similarly
serve to reduce the expression of yawn
contagion. Surprisingly this was not what our
investigations revealed. Participants in a
virtual environment were immune to a webcam
embedded in VR both when the recording state of
the webcam was implicit (i.e., labeled with or
without a red light) as in Experiment 2, or made
explicit by the researcher prior to testing as
in Experiment 3. Moreover, and even more
surprising, Experiments 4 and 5 revealed that
the actual presence of a humanoid avatar within
the participant's VR environment did not lead to
a modulation of the contagious yawning response
rates. This was true even when that person was
directly facing the participants and displaying
naturalistic movements, creating a more
immersive experience within VR.
Collectively these data indicate that
participants are quite sensitive to the social
contagion of yawning within a virtual
environment, which is consistent with recent VR
research on other forms of contagious/reflexive
behavior (i.e., gaze following). And yet,
surprisingly, the participants were immune to
the visual social signals of explicit
(Experiments 4 and 5) and implied (Experiments 2
and 3) social presence displayed within this
setting; but were extraordinarily sensitive to
the unseen real presence of an experimenter
outside of VR, i.e., in the testing room. The
dynamics and social effects of virtual reality
appear to be profoundly different from those in
actual reality, with the latter superseding and
dominating those of VR. These are evident both
in the failure of varied real-world social
presence effects within VR to have any
significant impact on a social contagion
trigger, which was true even for the more
realistic VR simulations of Experiment 5, and by
the effect of actual social presence outside the
VR environment in Experiment 1 to alter social
contagion to stimuli presented within this
setting. Thus, within the domain of behavioral
contagion, the social effects of VR are very
unlike reality itself.
These findings suggest a major difference in
the perception of social factors within and
behind the scenes in VR, yet limitations to this
study should be acknowledged. First, we suspect
that the time spent in VR prior to testing and
stimulus exposure is an important factor. As
with our previous study, it remains unknown
whether participants would habituate to the
researcher present condition over time and
resume a more typical contagious yawning pattern
thereafter. For example, past research has shown
that social presence effects from wearing
eye-trackers diminishes if participants wear
these devices for a short period of time prior
to experimentation. Although results from
Experiment 5 show no change in total immersion
across the length of testing (from experimental
block 1 to block 2), further research is needed
to assess how any socially-mediated reduction in
yawn contagion may dissipate over extended
periods of time following various manipulations
in VR. In addition, although the mere presence
of an onlooking humanoid avatar was insufficient
to alter yawning, it is possible that
modifications to this figure could change the
results. We attempted to construct a humanoid
avatar that possessed realistic features in
Experiment 4, and added naturalistic movement
properties to this individual in Experiment 5,
but further research could include even more
advanced behavioral features to enhance avatar
realism and interaction. Such modifications
could help avoid the "uncanny valley" of
affinity towards entities of increasing but not
complete human likeness and enhance the social
presence, or sense of being with another via
psychological involvement and/or behavioral
engagement, experienced from these conditions in
VR. Related to these issues, it is possible that
participants could be "present" and experience a
sensation of being in a real place even more
during the VR simulations. The results from
Experiment 5 indicate that average presence
scores were near the midpoint of the 7-point
Likert scale used. Thus, future research in this
area could aim to improve place and plausibility
illusions and/or suspensions of disbelief among
participants within VR through modifications to
the procedures and/or simulations. These issues
notwithstanding, the current findings
demonstrate the power of actual human
observation on responses to social stimuli
presented in VR, and provide a baseline for
further work in this area.
It remains unclear whether the results in
Experiment 1 are due to the enhanced recognition
and social awareness of the negative perceptions
of yawning when in the presence of others, or if
this was a result of more bottom-up processes
(i.e., heightened arousal) inhibiting the
mechanisms that naturally trigger the primitive
form of this response11. To date, the evidence
for socially-mediated decreases in yawn
contagion seem to be due at least in part to the
former. These factors could potentially be
disentangled in the current setup by recording
both gaze direction and physiological measures
from participants exposed to social presence
manipulations when engaged in VR56. It would be
interesting to explore whether social perception
or arousal factors have differential effects on
the diminished expression of contagious versus
spontaneous yawns. It seems plausible that since
spontaneous yawns are triggered by internal
neurophysiological processes3, the inhibition of
this response within crowded environments or
under direct observation is most likely a result
of the latter.
Methodologically, this study provides a
novel approach to the study and manipulation of
contagious yawning. While previous comparative
research has used computer animations to induce
contagious yawning in chimpanzees, this is the
first study we know of to trigger this behavior
within the context of a virtual environment. The
rate of contagious yawning in VR without any
social presence (i.e., low social presence
condition in Experiment 1: 38.1%) was quite
similar to yawning rates when using the same
stimulus in traditional laboratory procedures,
and thus serves as a valid method for future
research. Given the potential advantages of VR
over traditional laboratory manipulations,
refining this method in follow-up experiments
could lead to a standardized approach to
studying yawn contagion. Further research could
create humanoid avatars that yawn within the
simulation, building from recent work examining
human-robot interactions. Previous studies have
already begun to use VR in comparative research,
and thus similar methods could be applied to
investigate yawn contagion in non-human
animals.
Overall, this report adds to a growing
literature showing that social presence can be a
powerful deterrent of yawning in humans. In
addition to improving upon methodological
limitations to the current experiments, future
studies should specifically examine the
mechanism(s) contributing to the reduced yawning
rate when in the presence of live onlookers.
Given the potential adaptive function(s) and
psychological correlates to yawning in humans,
we believe the inhibition to yawning in some
social settings deserves further investigation.
Moreover, these experiments offer two insights
into human behavior and cognition in VR. First,
social presence in actual reality can have a
profound impact on the perception of, and
responses to, the virtual environment. Second,
social factors influencing our behavior in the
real world can have both similar (in the case of
contagious yawning) and differing (in the case
of social presence) effects when present in VR.
Together, these findings could have broad and
important applications for VR research across
the psychological and cognitive sciences.
