Anna Kis, KatinkaTóth, Orsolya
Kanizsár, József Topál
Abstract
Previous studies have found that human
yawning is contagious to dogs, but the results
are still controversial. It is also debated
whether contagious yawning is a sign of empathy,
and the physiological mechanisms behind this
phenomenon are also unknown. Our goal was to
further investigate if human yawning is
contagious to dogs, and how it is affected by
intranasally administered oxytocin. We also
tested if contagious yawning was related to
dogs_ empathetic skills measured via an
owner-completed questionnaire. Dogs (N_=_33)
were presented with human yawns (yawning
condition &endash; YC) or mouth openings (gaping
condition &endash; GC) on two test occasions
after intranasal administration of 12 IU
oxytocin or placebo. Contagious yawning (defined
as a significant increase in the number of yawns
in YC as compared to GC) was not found in either
of the treatment groups and the number of yawns
was not related to the owner-reported emotion
contagion of the dogs. However, oxytocin
pre-treatment significantly decreased the number
of yawns in dogs (in general, but more
particularly during the phase when human yawns
and gaping were presented). Furthermore, there
was a positive relation between signs of stress
and anxiety (mouth licking) and the number of
yawns (in general, but more particularly during
the post phase when human yawns and gaping were
not any more presented). We conclude that dogs'
yawning observed in this situation may largely
be due to situational stress. Thus the
difference between the oxytocin and the placebo
group is possibly caused by the decreased social
stress and not by oxytocin's hypothesised effect
on empathy.
1. Introduction
Behavioral non-conscious synchronization is
widespread among animals, including humans. It
occurs when individuals engage in the same motor
action during a very tight time window and when
they are spatially close [1,2].
Synchronizing with others has a highly adaptive
value and produces many different benefits as a
function of the type of synchronization. For
example, the synchronization in reproduction,
movements, and vigilance of group living animals
can increase their effectiveness in the defense
from predators [1]. Synchronization can
occur not only at group level but also at dyadic
level [3]. During their dyadic social
interactions, human and non-human animals can
synchronize their motor actions such as facial
expressions and body postures [4]. For
example, it has been recently demonstrated that
dogs synchronized their resting and locomotor
activity with that of their owners when they
were freely acting in an open area
[2].
The behavioral synchronization domain
includes two phenomena that are still debated
topics in human and non-human animal studies:
Rapid Mimicry (RM) and Yawn Contagion (YC)
[5]. Even though they are often
considered as different processes, they seem to
share similar neural and social mechanisms
[6&endash;8]. It has been proposed that
these motor resonance behaviors are grounded in
the automatic perception&endash;action coupling
in the sensorimotor areas [5,9] The
mirror neurons, discovered in monkeys and
humans, provided neurophysiological support to
the perception&endash;action model, because they
fire when the motor action is both observed and
perceived. For this reason, a large array of
studies indicates that RM and YC may be
automatic, fast, reflex-like mechanisms which go
beyond the intentional control
[10].
RM and YC seem to intermingle with emotional
contagion which is defined as "the tendency to
automatically mimic and synchronize expressions,
vocalizations, postures, and movements with
those of another person's and, consequently, to
converge emotionally" [11]. Emotional
contagion is a well-known basic building-block
of empathy [5] and implies that two (or
more) subjects share the same affective state
[12]. The perception of others' action
leads the observer to automatically reproduce
the same action. Through this unconscious
mirroring response, the observer can experience
the same affective emotional state underpinning
such action [13]. The expression and
perception of emotions in non-human animals seem
to be adaptive because these skills allow them
to respond quickly and appropriately to
unpredictable situations thus increasing their
survival and fitness [14,15]. According
to the perception&endash;action coupling, the
capacity to mirror others' behaviors acquires an
even more adaptive value in animals whose
relationships are not inhibited by rank rules
and that build and maintain their bonds through
cooperation and social affiliation
[16].
Beyond primates, cohesive and cooperative
societies also characterize several social
carnivores such as wolves (Canis lupus lupus)
and dogs (Canis lupus familiaris)
[17,18]. Compared to wolves, dogs have
expanded towards humans their propensity to
synchronize and affiliate [2]. Dogs
develop strong affinitive bonds with owners,
which represent their favorite social partners
[19]. One of the aspects of this
preferred affectional bond seems to be the
empathically predisposition of dogs towards
humans' emotions [20]. The capacity of
dogs to understand and experience humans'
emotions (i. e., empathy) [5,12]can be
evidenced by their (1) preferential responses to
the cry of humans compared to their humming or
talking [21], (2) ability to
discriminate emotions that are more positive for
owners [22] and (3) high propensity to
be infected by humans' yawns
[23&endash;25]. The question arising
from these evidence deals with the origin of the
human&endash;dog emotional sharing: is it a
phenomenon shaped by artificial selection or
evolutionarily rooted in the line of social
carnivores? Trying to answer the question, we
focus on rapid mimicry (RM) and yawn contagion
(YC) in wolf and dog groups to investigate if
the ability to emotionally engage with
conspecifics can provide the basis for the
development of inter-specific emotional
sharing.
2.2. Yawn Contagion in Dogs and Wolves:
When the Partner Matters
Spontaneous yawning is a primitive
stereotyped motor action pattern, that once
triggered, is uncontainable and unstoppable
[54]. It cannot be considered as a
facial expression because it recruits facial,
oral, laryngeal, pharyngeal, thoracic and
abdominal muscles. Indeed, depending on the
species, yawning may be also accompanied by eye
closing, vocalizations, body stretching,
pandiculation and even tongue protrusion
[63]. Yawning is a widespread phenomenon
in mammals and birds [64,65]. In humans,
the new ultrasound technologies evidenced the
presence of spontaneous yawning just starting
from the fetal phase [66]. The
occurrence of fetal yawning seems to contribute
in assessing the correct development of
brainstem and understanding the neural
underpinnings of sleep and arousal systems
[66]. Several hypotheses on the
functions of spontaneous yawning in non-human
animals have been formulated and these involved
both physiological and social explanations.
Indeed, yawning could act as a homeostatic
restoring and brain cooling mechanism, an
anxiety and drowsiness signal, and it has also a
role in the social communicative system (e. g.,
threat or alertness yawn)
[63].
Contagious yawning, a behavioral act
involuntarily induced by viewing or listening
other's yawns [64], is considered as a
different phenomenon which has been widely
demonstrated in human and non-human primates
[67&endash;71].
In primates, yawn contagion can be
considered as a proxy of emotional contagion and
it can be associated with the level of social
attachment between partners [5]. Even
though in human and non-human primates the
social modulation of contagious yawning is
supported by ethological [68&endash;71],
neurological [72,73], and psychological
evidence [74], contradictory results
about proximate mechanisms of contagion emerged
from the studies on domestic dogs. In 2008,
Joly-Mascheroni and co-authors [23]
evidenced that dogs responded to humans' yawns
in the 72% of experimental cases. This first
evidence of contagious yawning between two
different species led to hypothesize a possible
association between the phenomenon and empathic
involvement between partners. Later, two studies
found neither evidence of yawn contagion in dogs
nor its linkage with empathy [75,76]. In
2013, Madsen and Persson [77]
demonstrated the occurrence of dog&endash;human
yawn contagion. However, the empathic
interpretation of the phenomenon was not
supported because the relationship quality
shared by the interacting subjects (dog-owner vs
dogunfamiliar human) did not affect the level of
contagion [77]. Silva and colleagues
[25] provided evidence in favor of the
empathic hypothesis of contagious yawning in
dogs by using auditory instead of visual
stimuli. During the experiment, dogs yawned more
frequently when the auditory stimulus came from
familiar (the owner) rather than unfamiliar
humans. This first evidence of social modulation
of yawn contagion was quickly challenged. In
many species, including dogs, spontaneous
yawning can emerge in response to environmental
or social stressors [63, 78&endash;80].
In this light, it has been suggested that the
finding obtained by Silva and colleagues
[25] could be simply ascribed to a
distress response by dogs. When dogs heard, but
not saw, their owner, they experienced high
anxiety levels that determined an increase in
their yawning response.
In order to discriminate between the two
possible proximate causes (empathy vs anxiety)
of yawn contagion in dogs, Romero and co-workers
[24] carried out an elegant study
integrating ethological and physiological
approaches. During the research, dogs observed
familiar (the owner) and unfamiliar humans while
yawning (experimental condition) or simply
making mouth movements (control condition).
Concomitantly, the heart rate and heart rate
variability, two reliable physiological
indicators of stress and well-being in animals
[81], were monitored. The authors
demonstrated that dogs yawned more frequently in
experimental than in control condition and they
were more infected by the yawns of familiar than
unfamiliar humans. Intriguingly, the
physiological parameters did not differ between
experimental and control condition, thus
suggesting that yawn contagion was not related
to environmental or social stressors, but it was
modulated by the affinitive attachment between
dogs and their owners.
Recently, by investigating the possible
relation between yawn contagion and the empathic
gradient Kis and colleagues [82]
provided contrasting results. The authors tested
33 dogs in both yawning and control condition by
previously treating them with intranasal
oxytocin or placebo. Contagious yawning did not
occur at all and the number of yawns performed
by dogs were not related to the degree of social
closeness with owners but rather with the levels
of anxiety measured by mouth licking. Moreover,
the oxytocin pre-treatment significantly reduced
the yawning events. Nevertheless, it is worth
noting that intranasal administration of
oxytocin can activate the endogenous
hypothalamic oxytocin system, thus the still
lack of information about the neural,
physiological and behavioral consequences of
this activation implies to interpret with
caution the observed effects of exogenous
oxytocin [83&endash;85]. Accordingly,
Romero and colleagues [50] found that
the effects of exogenous oxytocin in dogs were
fine-tuned by the individual variability in
endogenous oxytocin: the dogs showing high basal
levels of oxytocin were less responsive to the
effects of intranasal oxytocin. In order to
clarify the "social role" of oxytocin in
dog&endash;human emotional contagion, future
studies should take into account the interaction
between the applied hormone doses and endogenous
secretion in modulating behavioral changes.
If the occurrence of contagious yawning and
its proximate causes have been investigated in
the dog&endash;human relationship, no studies
are available on the phenomenon of yawn
contagion between dogs. Is the phenomenon of
yawn contagion exclusively linked to
dog&endash;human interactions because of the
domestication process or is this motor resonance
phenomenon deeply rooted in social carnivores
and simply transferred from conspecifics to
human partners in the domestic dogs? A study
carried out by Romero and colleagues
[86] on captive wolves may answer the
question. The authors observed the animals in
their environmental and social setting without
altering their ordinary habits. The wolves
yawned more frequently when they were exposed to
conspecific yawning and the propensity to be
infected was positively linked to the strength
of social relationship shared between the
trigger (the initial yawner) and the observer.
These findings suggest that contagious yawning
is an ancestral phenomenon firstly naturally
selected for intra-specific social communication
and group synchronization and later artificially
selected for inter-specific communication. To
complete this framework, it is therefore crucial
exploring the occurrence of the phenomenon of
yawn contagion and its potential roles in groups
of dogs characterized by a certain level of
variance in the distribution of relationship
quality.
3. Conclusions
Wolves, the ancestors of dogs, constitute
cohesive family group whose members actively and
cooperatively participate to the pack life
creating a sort of "division of labor" system
[87,88]. As humans, wolves engage in
strong social and emotional bonds with
conspecifics and these relations could be at the
basis of the evolution of dog&endash;human
affective attachment [62]. The brain
areas involved in reward, stress, imitation, and
prosocial systems have been evolutionary
conserved along with the developmental pathway
of mammals [89&endash;91].
This common neurological state could have
permitted wolves and (ancient) humans to
develop, during the progressive increase of
their cohabitation, a shared representation of
actions and emotions. Motor and emotional
sharing probably promoted wolf&endash;human
social interactions and improved interspecific
empathic predisposition [92] that is at
the basis of the long history of
dog&endash;human affinitive relationship. In
conclusion, to understand the role of the
domestication process and artificial selection
operated by humans in shaping the emotional
contagion of dogs (rapid mimicry and yawn
contagion), we should focus not on the possible
presence and distribution of these phenomena
between hand-reared wolves and humans but,
instead, between the different wolf pack members
[93]. Long-term studies should quantify
(dyadic variance, seasonality, hormonal
variations) and qualify (affiliation, agonistic
support) the relationship quality shared by the
subjects and verify whether the incidence of
rapid mimicry and yawn contagion is socially
modulated in wolves. A further important point
is whether the occurrence and the distribution
of rapid mimicry and yawn contagion covariate
with some empathy-driven and prosocial behaviors
such as consolation and helping. This could
provide an indirect evidence of the potential
empathic nature of these motor resonance
phenomena.
As for dogs, we should focus on the
different breeds. By selecting the most recent
(less humanoriented) and ancient breeds (more
human-oriented) as models we can evaluate to
what extent the artificial selection has
operated in shaping the mimicry phenomena. To
reach the goal, observational studies on groups
of dogs belonging to the same breed and
experimental studies including dogs of different
breeds and their owners are necessary.
