The ability to recognize dog emotions depends on the cultural milieu in which we grow up

Abstract

Inter-specific emotion recognition is especially adaptive when species spend a long time in close association, like dogs and humans. Here, we comprehensively studied the human ability to recognize facial expressions associated with dog emotions (hereafter, emotions). Participants were presented with pictures of dogs, humans and chimpanzees, showing angry, fearful, happy, neutral and sad emotions, and had to assess which emotion was shown, and the context in which the picture had been taken. Participants were recruited among children and adults with different levels of general experience with dogs, resulting from different personal (i.e. dog ownership) and cultural experiences (i.e. growing up or being exposed to a cultural milieu in which dogs are highly valued and integrated in human lives). Our results showed that some dog emotions such as anger and happiness are recognized from early on, independently of experience. However, the ability to recognize dog emotions is mainly acquired through experience. In adults, the probability of recognizing dog emotions was higher for participants grown up in a cultural milieu with a positive attitude toward dogs, which may result in different passive exposure, interest or inclination toward this species.

Full text

1
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
The ability to recognize dog
emotions depends on the cultural
milieu in which we grow up
Federica Amici1,2,3*, James Waterman4, Christina Maria Kellermann3,5, Karimullah Karimullah2
& Juliane Bräuer6,7
Inter-specic emotion recognition is especially adaptive when species spend a long time in close
association, like dogs and humans. Here, we comprehensively studied the human ability to recognize
facial expressions associated with dog emotions (hereafter, emotions). Participants were presented
with pictures of dogs, humans and chimpanzees, showing angry, fearful, happy, neutral and sad
emotions, and had to assess which emotion was shown, and the context in which the picture had been
taken. Participants were recruited among children and adults with dierent levels of general experience
with dogs, resulting from dierent personal (i.e. dog ownership) and cultural experiences (i.e. growing
up or being exposed to a cultural milieu in which dogs are highly valued and integrated in human lives).
Our results showed that some dog emotions such as anger and happiness are recognized from early
on, independently of experience. However, the ability to recognize dog emotions is mainly acquired
through experience. In adults, the probability of recognizing dog emotions was higher for participants
grown up in a cultural milieu with a positive attitude toward dogs, which may result in dierent passive
exposure, interest or inclination toward this species.
e physiological foundations of basic emotions are shared by humans and other mammals1–5. Emotions are
oen expressed through behavioural and somatic displays4, which serve as signals for other individuals and may
have a crucial communicatory and social function in several species6–9. rough the expression of emotions, for
instance, individuals may communicate their intentions and motivations7, facilitating conspecics’ responses and
the establishment and maintenance of long-term relationships6.
Recognizing others’ facial expressions of emotions, therefore, clearly provides tness benets6,7. For instance,
an animal may become alert when another individual displays fear, as a predator or aggressive conspecics may
be nearby. Recognizing others’ emotions may also be advantageous in inter-specic interactions, such as mutual-
ism or predator-prey interactions10. However, inter-specic emotion recognition may be especially challenging,
as the same emotion may be displayed dierently in dierent taxa10. erefore, it is expected to be favoured by
evolution when two species spend a signicant amount of time in close association with each other, and each
species gains tness benets through recognition of the other species’ emotions.
Close association between humans and domestic dogs (Canis familiaris) has occurred since dogs’ domes-
tication, at least 30 000 years ago11,12. Dogs show remarkable communicative skills: they may use eye gaze as
a communicative act13,14, and decipher humans’ communicative intent15. ey can also make use of words16,
iconic signs17 and human gestures18–20. Moreover, dogs can use acoustic and visual cues to recognize human
emotions. Dogs, for instance, take the emotional expression of their owner into account when deciding whether
to approach a novel object21. Dogs can also recognize the emotional expressions of human faces e.g.10,22, and
integrate bimodal sensory information to discriminate positive and negative emotions from dogs and humans23.
Importantly, dogs’ ability to recognize human emotions appears to exceed the ability of other taxa, including
1Research Group “Primate Behavioural Ecology”, Department of Human Behavior, Ecology and Culture, Max Planck
Institute for Evolutionary Anthropology, Leipzig, Germany. 2Behavioral Ecology Research Group, Institute of Biology,
Faculty of Life Science, University of Leipzig, Leipzig, Germany. 3Leipzig Research Center for Early Child Development,
University of Leipzig, Leipzig, Germany. 4School of Psychology, University of Lincoln, Lincoln, UK. 5Faculty of Social and
Behavioral Sciences, Friedrich Schiller University, Jena, Germany. 6Department of Linguistic and Cultural Evolution, Max
Planck Institute for the Science of Human History, Jena, Germany. 7Friedrich Schiller University, Department of General
Psychology and Cognitive Neuroscience, Jena, Germany. *email: amici@eva.mpg.de
OPEN
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
wolves and chimpanzees, and it may be the result of the domestication process having selected for dogs that most
prociently communicate with humans24–26.
In contrast, the human ability to recognize dog emotions has received only limited attention. Studies using
auditory input demonstrate that humans can recognize some dog emotions, like aggressive barks to strangers27–30.
While dogs may display their emotions through auditory signalling31, they also use a large range of body and
facial signals, which are a primary channel for emotional transmission in several species e.g.6,32–34. However,
several studies suggest that children and adults do not reliably understand the body signals of dogs35–38, and
that children oen mistake angry dog facial displays for happy ones39. Indeed, not all emotions may be equally
easy to recognize. Overall, people are generally more successful at recognizing positive dog emotions, like hap-
piness38,40–42, while oen confusing negative emotions, like fear38,40,41; but see42. More contrasting results have
emerged on the human ability to recognize dogs’ aggressive behaviour from visual cues, with positive38,40 and
negative evidence42.
Furthermore, it is not clear whether previous experience with dogs is necessary for humans to recognize dog
emotions10. According to the co-domestication hypothesis, human ability to recognize dog emotions (or at least
some especially relevant ones, like angry emotions) may be supported by specially adapted mechanisms. In par-
ticular, convergent evolution would have led humans and dogs to evolve emotional displays and cognitive skills
that favour reciprocal understanding and inter-specic communication, with humans selecting dogs based on
their working abilities and communication skills, and humans evolving an ability to read dog emotions13,20,43–46.
erefore, even though direct experience with dogs (e.g. dog ownership) may still increase the ability to recog-
nize dog emotions, this ability should be partially present also in the absence of experience. Also in this respect,
experimental evidence provides contrasting results. In some studies, inexperienced humans (i.e. non-owners)
were better than humans with dog experience (e.g. dog owners) at reading dog emotions38, reliably recognizing
positive (i.e. curiosity and play) and negative (i.e. fear and social isolation) emotions34,47. In other studies, the abil-
ity to recognize dog emotions did not dier between dog-owners and non-owners28,29,42, although in some cases,
and for some emotions, it did increase slightly with age and experience30,34,41,47; see25.
Here, we conducted the rst comprehensive study of the human ability to recognize the facial expressions
associated with dog emotions (hereaer, emotions). Firstly, we thoroughly assessed the eect of experience with
dogs on the ability to recognize their emotions. “Experience with dogs” is a general concept that may encompass
a variety of relationships with dogs, such as (i) ownership of a dog, (ii) growing up in or (iii) being exposed to a
cultural milieu with a dog-positive attitude. By a cultural milieu with a dog-positive attitude we refer to a society
in which dogs are highly integrated in human lives, and in which there is a general positive attitude toward them.
In Europe, for example, dogs are generally seen as part of the family, are walked on a lead, enter homes and spend
substantial amount of time with people. In contrast, in traditional communities in Muslim countries, dogs are
oen viewed as impure and rarely integrated as part of the family see48,49. Clearly, this has nothing to do with the
mistaken notion that Muslims would hate dogs, and simply implies that dierent societies may importantly dier
in their general attitude to dogs. erefore, our study included (i) non-Muslim European dog-owners (i.e. owners
who grew up in and were exposed to a cultural milieu with a dog-positive attitude), (ii) non-Muslim European
non-owners (i.e. non-owners who grew up in and were exposed to a cultural milieu with a dog-positive attitude),
(iii) Muslim non-owners from countries in which Islam is the majority religion, but living in Europe for at least 3
years (i.e. non-owners who grew up in a cultural milieu with no dog-positive attitude, but who were extensively
exposed to one with a dog-positive attitude), and (iv) Muslim non-owners living in Morocco (i.e. non-owners
who grew up in and were exposed to a cultural milieu with no dog-positive attitude). If the co-domestication
hypothesis is to be supported, high performance would be expected also in groups who grew up in a cultural
milieu with no dog-positive attitude, and/or were not exposed to a cultural milieu with a dog-positive attitude.
However, experience with dogs may nonetheless increase human ability to recognize dog emotions. erefore,
being culturally exposed to a dog-positive attitude, growing up in such a cultural milieu and owning a dog should
have an increasingly stronger positive eect on the ability to read dog emotions.
Secondly, we compared participants’ ability to recognize dog, chimpanzee and human emotions, to assess
whether participants’ ability to read human emotions is more similar to their ability to read dog emotions (as pre-
dicted by the co-domestication hypothesis) or chimpanzee emotions (if emotions are recognized on the basis of
shared phylogenetic history). Moreover, by testing whether chimpanzee emotions were recognized by all partici-
pants in a similar way, independently of their general dog experience, we could rule out that potential dierences
in the ability to recognize dog emotions were simply reecting more general dierences in participants’ overall
ability to read animal emotions.
irdly, we assessed the eect of age on emotion recognition, by testing both children and adults. If the ability
to recognize dog emotions has been selected through evolution (in line with the co-domestication hypothesis),
performance in children should be similar to performance in adults, and similar in all participant groups.
Methods
Participants. As adult participants, we recruited 24 non-Muslim European dog-owners; 24 non-Muslim
Europeans who did not own a dog and did not live in close contact with one (hereaer, non-owners); 18 Muslim
non-owners from a country in which Islam is the majority religion, but that had been living in Europe for more
than 3 years; and 23 Muslim non-owners living in Morocco. rough formal educational establishments and
local kindergartens, we further recruited 5- and 6-year-old children. In particular, we included 23 non-Muslim
European dog-owners; 31 non-Muslim European non-owners; and 23 Muslim non-owners living in Morocco.
Participants belonged to both sexes, and diered in terms of their individual attitude to dogs (i.e. how much they
liked dogs and considered them to be important for humans). All Muslim participants were practicing Muslims,
except for one. Both in adults and children, non-Muslim European dog-owners (hereaer, EGO, with E stand-
ing for Extensive exposure to a dog-positive cultural milieu, G standing for having Grown-up in such a cultural
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
milieu, and O standing for dog-Ownership) liked dogs the most and found them most important, followed by
non-Muslim European non-owners (EGo, with o standing for no dog-ownership), Muslim non-owners in Europe
(Ego, with g standing for not having grown-up in a dog-positive cultural milieu) and Muslim non-owners in
Morocco (ego, with e standing for not having been exposed to a dog-positive cultural milieu; see Supplementary
Material for more details).
Materials. Stimuli consisted of frontal facial photographs of 20 dierent dogs, 20 dierent chimpanzees, and
20 dierent humans, on a white background. Dog pictures were taken in a park in Leipzig, Germany, noting the
context in which the pictures were taken (see below), or pre-selected from real-life images from the internet,
ascertaining the context by contacting the owners of the pictures. All dogs had a wolf-like face (i.e. German
Shepherd, Husky and similar), with upright ears and relatively short hair. is is because other breeds may have a
reduced social signaling capacity, due to brachycephaly, oppy ears or long fur in the face. Chimpanzee pictures
were real-life images taken at the Wolfgang Koehler Primate Research Center in Leipzig, Germany, also noting the
context in which the pictures had been taken. In this way we dened all animal emotions depending on the objec-
tive context in which the pictures had been taken, anchoring the pictures to behaviourally dened situations38,50.
Finally, human pictures were instructed images downloaded from the AR Face Database51.
All the pictures were selected by the rst and last authors if they both considered them typical for the contexts
listed below. “Typical” referred to the fact that these facial expressions were regularly displayed in these contexts
(e.g. in chimpanzees, play face during playful interactions with conspecics; see e.g. Parr et al. 2006). Pictures
were also sent to other seven expert colleagues, who classied them into the ve dierent categories described
below. Researchers’ agreement with the authors’ choice was very good (rs = 0.91; N = 320, p < 0.001).
In line with previous literature e.g.28–30, we included the following 5 dierent emotions (displayed by 4 dif-
ferent individuals per species): (a) happy/playful, (b) sad/distressed, (c) angry, (d) fearful and (e) neutral. For
dogs and chimpanzees, pictures had been respectively taken in the following contexts: (a) the individual was
together with a trusted conspecic partner, playing with him/her; (b) the individual had been abandoned or was
observing a stressful/undesirable event, like a ght; (c) the individual was in a state of excitement directly before/
while attacking a conspecic; (d) the individual was afraid of being directly attacked by some stronger conspecic
partner; (e) none of the previously described situations had been happening/happened in the last/next 3 minutes.
Procedures. Research was approved by the University of Lincoln Psychology Research Ethics Committee
(soprec@lincoln.ac.uk) and by the University of Jena, and the methods were carried out in accordance with the
relevant guidelines and regulations. Informed consent was directly signed by adult participants. In case of chil-
dren, informed consent was obtained from their parents and/or legal guardians through the kindergartens. All
adults were tested with the following procedure. Before being tested, participants provided biographic informa-
tion, were asked to state whether they owned a dog or had had close contact with dogs during their lives, and
provided their opinion on dogs on a scale from 1 to 5 (i.e. “how much do you like dogs”, and “how important are
dogs for humans”). In contrast, none of the participants had ever owned or had had extensive experience with
chimpanzees (e.g. working with them, or having friends who owned a chimpanzee). Aer that, the Experimenter
(E) instructed participants in the procedure for the two tasks.
In the rst task, we adapted the procedure from Pongracz and colleagues29. E sat in front of the participants
and provided them with a pen and coding sheets with which to note their answers. For each participant E had a
set of 30 pictures (see Supplementary Material for more information). Each one was shown to the subject for up to
30 s, or until the subject rated the picture on the coding sheet, on a 1–5 scale see e.g.29, for each of the 4 emotions
above: happiness, sadness, anger, fear. In the second task E repeated the procedure, showing participants the same
set of 30 pictures, in the same order. However, participants had to specify in which of the 5 contexts listed above
the picture had been taken. Adults’ responses in both tasks were analysed together in the same model (see below).
To make it more age appropriate see40,52, children were tested with a simplied version of the rst task, in which
they were shown 15 pictures and had to attribute each picture to one of the ve emotions (happiness, sadness,
anger, fear, and neutral), as read aloud by E.
Statistics. Analyses were conducted using generalized linear mixed models53 with the lme4 package in R
soware (version 3.2.3)54. Continuous variables were z-transformed to facilitate model convergence. We used
a likelihood ratio test55 to compare full models with null models. When full models diered signicantly from
null models, likelihood ratio tests were conducted to obtain the p values for each test predictor via single-term
deletion56. Post-hoc comparisons were then conducted using Tukey tests (below, only signicant post-hoc tests
are reported). No convergence issues were detected. To rule out collinearity, we used variance ination factors
(VIF57), which were good (maximum VIFs across all models = 1.85). No random slopes were included to avoid
convergence issues.
Model 1 investigated how emotion recognition by children is aected by their general experience with dogs,
depending on the species and emotion observed. Given that the dependent variable was binary (i.e. 0 for an
incorrect choice and 1 for a correct choice), models were run with a binomial structure. General experience (i.e.
EGO, EGo, ego) emotion (anger, fear, happiness, neutrality, sadness), species (dogs, chimpanzees, humans), and
their 2- and 3-way interactions were test predictors. As control predictors we entered participants’ gender (male
or female), number of trials (1 to 15) and the proportion of trials in which each participant selected that specic
emotion in the species (to control for the fact that participants who answer with a certain emotion in most trials
will also have a higher probability to correctly recognize that emotion, despite having no greater sensitivity to that
emotional expression). We further included child ID as a random eect, to account for the non-independence of
data points.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Model 2 investigated how emotion recognition by adults is aected by their general experience with dogs,
depending on the species and emotion observed. We used the same binary dependent variable (i.e. individual
response at recognizing emotions) and test predictors as in Model 1 (i.e. general experience with dogs, emotion,
species and their 2- and 3-way interactions), further including task type (i.e. Task 1 or 2) among the test predictors
(as adults were administered two tasks), and a fourth factor level for experience (i.e. Ego). We also included the
same control predictors as xed eects (i.e. proportion of trials in which each participant selected that specic
emotion in the species, participants’ gender and number of trials, from 1 to 30), further including subject’s age
(as this varied across adult participants; see Supplementary Material). Furthermore, we included adult ID as a
random eect.
If three-way interactions were not signicant, they were downgraded to the two-way interactions experience
x species and emotion x species. Post-hoc comparisons in Models 1 and 2 assessed whether participants with
dierent general dog experience diered at recognizing specic emotions in dierent species. For each emotion,
we also analysed whether participants with dierent general dog experience recognized dog emotions dierently
than in the other two species. Only for dog emotions, we further addressed which emotions were easier to evalu-
ate by participant groups with dierent general dog experience.
Finally, to assess whether individual ability to recognize emotions is consistent across the three species
(humans, chimpanzees and dogs), we calculated the average of correct responses for each subject and species and
used Spearman exact tests, separately for adults and children.
Results
Model 1 – Children. e comparison between the full and null model was signicant (GLMM: χ2 = 209.76,
df = 44, p < 0.001), but not the three-way interaction experience*emotion*species (GLMM: χ2 = 19.35, df = 16,
p = 0.251). After downgrading the interaction, the two-way interaction emotion x species was significant
(GLMM: χ2 = 68.36, df = 8, p < 0.001), but not the interaction experience x species (GLMM: χ2 = 6.95, df = 4,
p = 0.139; see Table1).
Post-hoc analyses see58 were run to assess whether emotional expressions were more easily recognized in
certain species, across all participants (Fig.1). All ve emotions were recognized more easily in humans than in
chimpanzees (all p < 0.004). Neutral and sad emotions were recognized more easily in humans than in dogs (both
p < 0.001). Happy emotions were recognized more easily in humans than in the other species, but also in dogs
more than chimpanzees (both p ≤ 0.005). Finally, angry emotions were recognized in dogs like in humans, and
more than chimpanzees (p < 0.001).
Test category
Children Adults
χ2df Pχ2df P
Experience * Species 6.95 4 0.139 34.79 6 <0.001*
Emotion * Species 68.36 8 <0.001*165.69 8 <0.001*
Task type — — — 8.65 1 0.003*
Trial number 7.32 1 0.007*11.65 1 <0.001*
Proportion of trials in which the
emotion was selected 453.87 1 <0.001*1321.32 1 <0.001*
Subject’s gender 0.31 1 0.576 2.08 1 0.149
Subject’s age — — — 0.00 1 0.975
Table 1. Results of Models 1 and 2, for children and adults (respectively), with emotion recognition as the
dependent variable. Signicant results are marked with. *Signicant test predictors are also in bold.
0%
20%
40%
60%
80%
100%
angryfearful happyneutral sad
Children
chimpanzeesdogs humans
Figure 1. For each species (dogs, humans, chimpanzees), mean estimated probability (+SE) of children
recognizing dierent emotions (angry, fearful, happy, neutral, sad). Parentheses indicate signicant post-hoc
comparisons, and the continuous line chance level.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

5
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Model 2 – Adults. e comparison between the full and null model was signicant (GLMM: χ2 = 1045,
df = 60, p < 0.001) but not the three-way interaction experience*emotion*species (GLMM: χ2 = 30.59, df = 24,
p = 0.166). After downgrading the interaction, both the two-way interaction emotion x species (GLMM:
χ2 = 165.69, df = 8, p < 0.001) and the interaction experience x species (GLMM: χ2 = 34.79, df = 6, p < 0.001)
were signicant. Also the task type was signicant (GLMM: χ2 = 8.65, df = 1, p = 0.003), indicating that recog-
nizing the context in which the picture was taken was signicantly easier than directly naming emotions (see
Table1).
Post-hoc analyses see58 were run to assess whether experience with dogs aected adults’ ability to recognize
emotions in certain species (Fig.2). Non-Muslims (EGO, EGo) were better than Muslims (Ego, ego) at recogniz-
ing dog emotions (all p < 0.001). Non-Muslims (EGO, EGo) were also better than Muslims (Ego, ego) at recog-
nizing human emotions, which were displayed by Caucasian actors (all p < 0.001). Experience with dogs had no
eect on the ability to recognize emotions in chimpanzees.
Post-hoc analyses were further run to assess whether emotional expressions were more easily recognized in
certain species by all participants (Fig.3). As in children, all ve emotions were recognized more easily in humans
than in chimpanzees (all p < 0.001). Neutral and fearful emotions were recognized more easily in humans than in
dogs (both p < 0.001). Happy and sad emotions were recognized more easily in humans than in the other species,
but also in dogs more than chimpanzees (both p ≤ 0.023). Finally, like in children, angry emotions were recog-
nized in dogs like in humans, and more than chimpanzees (p < 0.001).
Spearman exact tests. Finally, we assessed whether individual ability to recognize emotions was consist-
ent across the three species. For adults, the individual average of correct responses correlated between species of
stimuli (dogs-chimpanzees: ρs = 0.323, n = 89, p = 0.002; dogs-humans: ρs = 0.644, n = 89, p < 0.001). In contrast,
children’s average of correct responses did not correlate between species of stimuli (dogs-chimpanzees: ρs = 0.002,
n = 77, p = 0.987; dogs-humans: ρs = 0.189, n = 77, p = 0.098).
Discussion
Our results indicate that the ability to recognize dog emotions is mainly acquired through experience. In particu-
lar, children’s ability to recognize dog emotions was similar across all participants, with no eect of general expe-
rience with dogs. Children had more trouble recognizing dog emotions than human emotions (except for anger),
and were equally bad at recognizing dog and chimpanzee emotions (except for angry and happy emotions, which
0%
20%
40%
60%
80%
100%
angryfearful happyneutral sad
Adults
chimpanzeesdogs
humans
Figure 2. For each species (dogs, humans, chimpanzees), mean estimated probability (+SE) of adults
recognizing dierent emotions (angry, fearful, happy, neutral, sad). Parentheses indicate signicant post-hoc
comparisons, and the continuous line chance level.
0%
20%
40%
60%
80%
100%
chimpanzeesdogshumans
Adults Non-Muslim European owners (EGO)Non-Muslim European non-owners (EGo)
Muslim non-owners in Europe (Ego)Muslim non-owners in Morocco (ego)
Figure 3. For each experience group (i.e. non-Muslim European owners EGO, non-Muslim European non-
owners EGo, Muslim non-owners in Europe Ego, Muslim non-owners in Morocco ego), mean estimated
probability (+SE) of adults recognizing emotions in dierent species (chimpanzees, dogs and humans).
Parentheses indicate signicant post-hoc comparisons, and the continuous line chance level.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

6
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
were recognized in dogs better than in chimpanzees). Except for anger and perhaps happiness, therefore, children
appear able to only limitedly recognize dog emotions, regardless of their general experience with dogs. In adults,
in contrast, the ability to recognize emotions strongly varied depending on their general experience with dogs,
but also on the emotion and the species of the stimuli observed. In particular, participants with more general dog
experience (i.e. EGO and EGo, who grew up in and were exposed to a cultural milieu with a dog-positive attitude,
regardless of whether they owned a dog) were overall more procient at recognizing dog emotions than partic-
ipants with less general dog experience (i.e. Ego and ego, who grew up in a cultural milieu with no dog-positive
attitude). ese dierences did not hold when assessing chimpanzee emotions. Finally, as in children, adults
recognized all dog emotions worse than human emotions (except for anger), but angry, sad and happy emotions
were recognized better than in chimpanzees, suggesting a possible increase through age in the ability to recognize
dog emotions.
All children recognized dog emotions in the same way, independently of their experience with dogs. Moreover,
except for anger and happiness, children had just as much trouble recognizing dog emotions as chimpanzee emo-
tions. ese results suggest that the ability to read dog emotions does not manifest spontaneously in young chil-
dren (but see below for further discussion), and it is mainly acquired over the course of development. In adults,
in contrast, general experience with dogs had a clear positive eect on the ability to understand dog emotions.
Participants growing up in a cultural milieu with a dog-positive attitude (EGO and EGo) were overall more pro-
cient at recognizing dog emotions than other participants (Ego and ego). ese results are noteworthy because
they suggest that it is not necessarily direct experience with dogs (i.e. dog ownership) that aects humans’ ability
to recognize their emotions e.g.28,29,38,42,59, but rather the cultural milieu in which humans develop. Growing up in
a cultural milieu in which dogs are viewed as highly important for humans, and are highly integrated in human
lives, may result in dierent passive exposure, or dierent interest and inclination toward this species. erefore,
possible cultural dierences in the ability to read dog emotions only emerge through development (at least aer
6years of age), when the eects of growing up in a cultural milieu with a dierent attitude toward dogs start
aecting human ability to recognize their emotions. Crucially, all adult participants recognized chimpanzee emo-
tions in a very similar way, independently of their general dog experience, suggesting that our results were not
simply reecting more general dierences in participants’ overall ability to read animal emotions.
An important exception to this pattern are anger and happiness, which were reliably recognized also
by children, regardless of their previous general experience with dogs. These results seem to support the
co-domestication hypothesis, in that even children with minimal experience (i.e. young age, no direct experience
with dogs, no cultural milieu with a dog-positive attitude) correctly interpret some dog emotions. e ability
to recognize anger is clearly adaptive, as it provides immediate tness benets (i.e. reduced risk of receiving
aggression) by conveying crucial information about possibly dangerous situations, and thus bears higher sur-
vival costs27,30,38. However, it is also possible that our results simply reect the fact that humans quickly learn to
recognize anger through experience. e fact that even young Muslim children in Morocco could successfully
recognize dog anger provides more support to the hypothesis of this ability being supported by specially adapted
mechanisms that operate largely independent of specic experiences. However, future studies should provide
stronger evidence in this sense, by for instance testing even younger children.
Also for adults, not all dog emotions were as easy to recognize. Adult participants were generally procient at
recognizing happy emotions, but not fearful ones. ese results are in line with previous studies, also suggesting
that dog fearful emotions may be especially hard to read e.g.30,40,41, while happy/friendly emotions are easier to
decode25,40,41. Adults were also generally procient at recognizing angry emotions, like children. In children,
angry and happy emotions were recognized more easily than sad and fearful ones. ese results seem to expand a
pattern, according to which humans recognize happiness and anger in other humans relatively early in life, while
fear recognition follows a much slower developmental trajectory60.
Overall, children’s performance only marginally matched adults’ performance, as only adults highly varied in
their response depending on general experience, species and emotion. is suggests that, although some emo-
tions are recognized early on independently of previous experience, learning is crucial to improve dog emotion
recognition. ese results conrm previous studies, showing that younger children are less procient at reading
dog emotions e.g.40. If the ability to recognize emotions is acquired through development, it is therefore no sur-
prise that culture may play a major role in the skills acquired, as not all emotions may be culturally relevant e.g.61.
Finally, our results raise four additional considerations. Firstly, inter-individual dierences in emotion rec-
ognition were maintained across species, but only in adults, with participants that were better able to recognize
emotions in one species also being better able to recognize them in the other ones. is is in line with other
ndings showing that humans perceive human and dog facial expressions in a similar way e.g.25. Secondly, suc-
cess at recognizing emotions diered depending on the task administered, with adults performing better overall
when asked to recognize the context in which the pictures were taken, as compared to when asked to name the
emotions observed see e.g.62. is suggests that tasks using context cues may be more ecient in catching human
ability to read emotions. irdly, emotions were generally recognized better in humans than in dogs by all partic-
ipants, but Muslim participants performed worse than non-Muslims at recognizing human emotions. is was
likely due to a limitation of our study, in that the human stimuli we used only depicted white Western models
see63–67. Fourthly, all participants were rather bad at recognizing chimpanzee emotions. Although this may seem
unexpected, it is important to note that chimpanzees and humans, despite being closely related and thus sharing
physical and functional similarities68–71, have facial emotions that dier in substantial ways (e.g. ear and head
movements play a larger role in chimpanzee facial expressions71,72).
Unfortunately, our study also presented several limitations. Firstly, pictures of emotional expressions were
selected based on the authors’ and other expert colleagues’ rating. Although the selection was based on objective
criteria (see above), future studies may benet from using dierent approaches for the stimuli selection, like FACS
e.g.68. Secondly, all dogs included in our stimuli had a German shepherd-like face, because morphological features
Content courtesy of Springer Nature, terms of use apply. Rights reserved

7
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
of these breeds (e.g. hair, ears) ensure that emotional expressions are well visible. However, future studies should
include a larger variety of dog breeds, and specically test whether experience with specic dog breeds easily
transfers to the ability of recognizing emotions in dierent breeds. irdly, future studies should also include
participants with a broader range of experience with dogs, also including participants not owning a dog, but with
extensive experience with dogs (e.g. experts working with dogs, non-experts living in close contact to dogs), and
possibly participants from other cultural milieus which may show more diverse attitudes to dogs. Fourthly, the
stimuli we used depicted dierent individual dogs for the dierent expressions, creating a potential confound
between the emotional expressions and the identity of the dog. As participants across all tested groups (EGO,
EGo, Ego, ego) were exposed to exactly the same stimuli, this limitation cannot explain the results obtained.
However, future studies should ideally present participants with pictures of the same individuals showing the
dierent emotional expressions. Finally, before being tested, participants were asked about their experience and
relationship with dogs. Although this approach was necessary to select participants for the dierent tested groups
(EGO, EGo, Ego, ego), we cannot rule out that these questions might have aected participants’ response in the
tasks. erefore, future studies should nd a dierent approach to recruit participants.
In conclusion, our study provides partial support for the co-domestication hypothesis, but also shows that
the ability to recognize dog emotions is largely acquired through general experience with dogs. More than direct
experience with dogs (i.e. ownership), the cultural milieu in which participants grew up likely determined the
interest with which humans attended to dogs and were therefore able to pick up subtle cues that facilitated emo-
tional recognition. Future studies should further investigate exactly which cultural aspects aect this ability.
Moreover, it will be important to use dierent procedures to conrm these results, by for instance comparing
performance with both instructed and real-life stimuli, and both facial and body expressions e.g.30,40; see73. is
will not only allow us to better understand inter-cultural variation in emotion recognition, but will also provide
us with useful hints to facilitate inter-specic communication and reduce the occurrence of harmful or negative
incidents between humans and dogs caused by humans’ inability to read dog signals e.g.74.
Received: 2 January 2019; Accepted: 26 October 2019;
Published: xx xx xxxx
References
1. Pansepp, J. The basic emotional circuits of mammalian brains: do animals have affective lives? Neurosci. Biobehav. Rev. 35,
1791–1804 (2011).
2. Paul, E. S., Harding, E. J. & Mendl, M. Measuring emotional processes in animals: the utility of a cognitive approach. Neurosci.
Biobe hav. R ev. 29, 469–491 (2005).
3. Mendl, M., Burman, O. H. P. & Paul, E. S. An integrative and functional framewor for the study of animal emotion and mood. Proc.
Biol. Sci. 277, 2895–2904 (2010).
4. Anderson, D. J. & Adolphs, . A framewor for studying emotions across species. Cell 157, 187–200 (2014).
5. Watanabe, S. & uczaj, S (Eds) Emotions of animals and humans: comparative perspectives. (Toyo: Springer, 2013).
6. Parr, L. A., Winslow, J. T., Hopins, W. D. & de Waal, F. B. ecognizing facial cues: individual discrimination by chimpanzees (Pan
troglodytes) and rhesus moneys (Macaca mulatta). J. Comp. Psychol. 114, 47–60 (2000).
7. Schmidt, . L. & Cohn, J. F. Human facial expressions as adaptations: evolutionary questions in facial expression research. Am. J.
Phys. Anthropol. 33, 3–24 (2001).
8. Blair, . J. . Facial expressions, their communicatory functions and neuro-cognitive substrates. Phil. Trans. R. Soc. Lond. B 358,
561–572 (2003).
9. Matsumoto, D. & Hwang, . S. Judgments of facial expressions of emotion in prole. Emotion 11, 1223–1229 (2011).
10. Mueller, C. A., Schmitt, ., Barber, A. L. A. & Huber, L. Dogs can discriminate emotional expressions of human faces. Curr. Biol. 25,
601–605 (2015).
11. almann, O. et al. Complete mitochondrial genomes of ancient canids suggest a European origin of domestic dogs. Science 342,
871–874 (2013).
12. Grimm, D. Dogs may have been domesticated more than once. Science 352, 1153–1154 (2016).
13. Soproni, ., Milosi, A., Topal, J. & Csanyi, V. Comprehension of human communicative signs in pet dogs. J. Comp. Psychol. 115,
122–126 (2001).
14. Teglas, E., Gergely, A., upan, ., Milosi, A. & Topal, J. Dogs’ gaze following is tuned to human communicative signals. Curr. Biol.
22, 209–212 (2012).
15. aminsi, J., Schulz, L. & Tomasello, M. How dogs now when communication is intended for them. Dev. Sci. 15, 222–232 (2012).
16. aminsi, J., Call, J. & Fischer, J. Word learning in a domestic dog: evidence for ‘fast mapping’. Science 304, 1682–1683 (2004).
17. aminsi, J., Tempelmann, S., Call, J. & Tomasello, M. Domestic dogs comprehend human communication with iconic signs. Dev.
Sci. 12, 831–837 (2009).
18. Bräuer, J., aminsi, J., iedel, J., Call, J. & Tomasello, M. Maing inferences about the location of hidden food: social dog, causal
ape. J. Comp. Psychol. 120, 38–47 (2006).
19. Milosi, A. & Soproni, . A comparative analysis of animals’ understanding of the human pointing gesture. Anim. Cogn. 9, 81–93
(2006).
20. aminsi, J. & Marshall-Pescini, S. e social dog. (Academic Press, 2014).
21. Merola, I., Prato-Previde, E. & Marshall-Pescini, S. Dogs’ social referencing towards owners and strangers. PLoS ONE 7, e47653
(2012).
22. Nagasawa, M., Murai, ., Mogi, . & iusui, T. Dogs can discriminate human smiling faces from blan expressions. Anim. Cogn.
14, 525–533 (2011).
23. Albuquerque, N. et al. Dogs recognize dog and human emotions. Biol. Lett. 12, 20150883 (2016).
24. Udell, M. A. ., Dorey, N. . & Wynne, C. D. L. What did domestication do to dogs? A new account of dogs’ sensitivity to human
actions. Biol. Rev. 85, 327–345 (2009).
25. ujala, M. V. Canine emotions as seen through human social cognition. Anim. Sent. 14(1) (2017).
26. Bräuer, J., Silva, . & Schweinberger, S. . Communicating canine and human emotions. Anim. Sent. 14(7) (2017).
27. Scheumann, M., Hasting, A. S., otz, S. A. & Zimmermann, E. e voice of emotion across species: how do human listeners
recognize animals’ aective states? PLoS ONE 9, e91192 (2014).
28. Pongracz, P., Molnar, C. & Milosi, A. Acoustic parameters of dog bars carry emotional information for humans. Appl. Anim.
Behav. Sci. 100, 228–240 (2006).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

8
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
29. Pongracz, P., Molnar, C., Milosi, A. & Csanyi, V. Human listeners are able to classify dog (Canis familiaris) bars recorded in
dierent situations. J. Comp. Psychol. 119, 136–144 (2005).
30. Pongracz, P., Molnar, C., Doa, A. & Milosi, A. Do children understand man’s best friend? Classication of dog bars by pre-
adolescents and adults. Appl. Anim. Behav. Sci. 135, 95–102 (2011).
31. Feddersen-Petersen, D. U. Hundepsychologie; Sozialverhalten und Wesen, Emotionen und Individualität (Stuttgart: Franch osmos
Verlag, 2004).
32. Eman, P. Gesichtsausdruck und gefühl (Paderborn: Junfermann, 1988).
33. Eman, P. Facial expression and emotion. Am. Psychol. 48, 384–392 (1993).
34. Schirmer, A., Seow, C. S. & Penney, T. B. Humans process dog and human facial aect in similar ways. PLOS ONE 8, e74591 (2013).
35. eisner, I. . & Shofer, F. S. Eects of gender and parental status on nowledge and attitudes of dog owners regarding dog aggression
toward children. J. Am. Vet. Med. Assoc. 233, 1412–1419 (2008).
36. erswell, . J., Bennett, P., Butler, . L. & Hemsworth, P. H. Self-reported comprehension ratings of dog behavior by puppy owners.
Behav. Process. 81, 376–382 (2009).
37. Mariti, C. et al. Perception of dogs’ stress by their owners. J. Vet. Behav. 7, 213–219 (2012).
38. Bloom, T. & Friedman, H. Classifying dogs’ (Canis familiaris) facial expressions from photographs. Behav. Process. 96, 1–10 (2013).
39. Meints, ., acca, A. & Hicey, N. How to prevent dog bite injuries? Children misinterpret dogs’ facial expressions. Inju ry P rev. 16,
1136 (2010).
40. L aestani, N. N., Donaldson, M. L. & Waran, N. Interpretation of dog behavior by children and young adults. Anthrozoös 27, 65–80
(2014).
41. Wan, M., Bolger, N. & Champagne, F. A. Human perception of fear in dogs varies according to experience with dogs. PLoS One 7,
e51775 (2012).
42. Tami, G. & Gallagher, A. Description of the behaviour of domestic dog (Canis familiari s) by experienced and inexperienced people.
Appl Anim Behav Sci 120, 159–169 (2009).
43. Hare, B., Brown, M., Williamson, C. & Tomasello, M. e domestication of social cognition in dogs. Science 298, 1634–1636 (2002).
44. Milosi, A. & Szabo, D. Modelling behavioural evolution and cognition in canines: Some problematic issues. Jap. J. Anim. Psychol.
62, 69–89 (2012).
45. Schleidt, W. M. & Shalter, M. D. Co-evolution of humans and canids: an alternative view of dog domestication: HOMO HOMINI
LUPUS? Evol. Cogn. 9, 57–72 (2003).
46. Coppinger, . & Coppinger, L. Dogs: a startling new understanding of canine origin, behavior, and evolution (New Yor: Scribner,
2001).
47. Bahlig-Pieren, Z. & Turner, D. C. Anthropomorphic interpretations and ethological descriptions of dog and cat behavior by lay
people. Anthrozoös 12, 205–210 (1999).
48. Jegatheesan, B. Inuence of cultural and religious factors on attitudes towards animals. In Handboo on animal-assisted therapy:
theoretical foundations and guidelines for practice (Ed. Fine, A. H.) (Amsterdam: Elsevier, 2015).
49. Foltz, . C. Animals in Islamic tradition s and Muslim cultures (Oxford: Oneworld Publications, 2005).
50. acca, A., Guo, ., Meints, . & Mills, D. S. eading faces: dierential lateral gaze bias in processing canine and human facial
expressions in dogs and 4-year-old children. PLoS ONE 7, e36076 (2012).
51. Martinez, A. M. & Benavente, . e A Face Database. CVC Techn. Rep. 24 (1998).
52. Pons, F., Harris, P. L. & de osnay, M. Emotion comprehension between 3 and 11 years developmental periods and hierarchical
organization. Eur. J. Dev. Psychol. 1, 127–152 (2004).
53. Baayen, . H., Davidson, D. J. & Bates, D. M. Mixed-eects modeling with crossed random eects for subjects and items. J. Mem.
Lang. 59, 390–412 (2005).
54. Bates, D. M. Lme4: Mixed-eects modeling with , http://lme4. r-forge. r-project. org/boo (2010).
55. Dobson, A. J. An introduction to Generalized Linear Models (Boca aton, Chapman & Hall/CC, 2002).
56. Barr, D. J., Levy, ., Scheepers, C. & Tily, H. J. andom eects structure for conrmatory hypothesis testing: eep it maximal. J. Mem.
Lang. 68, 255–278 (2013).
57. Field, A. Discovering statistics using SPSS (London: Sage Publications, 2005).
58. Gigerenzer, G. Mindless statistics. J. Socio-Econ. 33, 587–606 (2004).
59. Moss, S. P. & Wright, J. C. e eects of dog ownership on judgements of dog bite lielihood. Anthrozoös 1, 95–99 (1987).
60. omas, L. A., De Bellis, M. D., Graham, . & LaBar, . S. Development of emotional facial recognition in late childhood and
adolescence. Dev. Sci. 10, 547–558 (2007).
61. itayama, S., Marus, H. . & uroawa, M. Culture, emotion, and well-being: good feelings in Japan and the United States. Cogn.
Emot. 14, 93–124 (2000).
62. Hassin, . ., Aviezer, H. & Bentin, S. Inherently ambiguous: facial expressions of emotions, in context. Em ot. R ev. 5, 60–65 (2013).
63. Jac, . E., Blais, C., Scheepers, C., Schyns, P. G. & Caldara, . Cultural confusions show that facial expressions are not universal.
Curr. Biol. 19, 1543–1548 (2009).
64. Jac, . E., Garrod, O. G., Yu, H., Caldara, . & Schyns, P. G. Facial expressions of emotion are not culturally universal. PNAS 109,
7241–7244 (2012).
65. Stahl, J., Wiese, H. & Schweinberger, S. . Expertise and own-race bias in face processing: an event-related potential study.
Neuroreport 19, 583–587 (2008).
66. Gendron, M., oberson, D., van der Vyver, J. M. & Barrett, L. F. Perceptions of emotion from facial expressions are not culturally
universal: evidence from a remote culture. Emot. 14, 251–262 (2014).
67. Hughes, B. L. et al. Neural adaptation to faces reveals racial outgroup homogeneity effects in early perception. PNAS 116,
14532–14537 (2019).
68. Parr, L. A. & Waller, B. M. Understanding chimpanzee facial expression: insights into the evolution of communication. Soc. Cogn.
Aect Neurosci. 1, 221–228 (2006).
69. Cheva lier-Solnio, S. Facial expression of emotion in non-human primates. In Darwin and facial expression: a century of research
in review (Ed. Eman, P.) (Los Altos, CA: Malor Boos, 1973).
70. Parr, L., Cohen, M. & de Waal, F. B. M. Inuence of social context on the use of blended and graded facial displays in chimpanzees.
Int. J. Primatol. 26, 73–103 (2005).
71. Parr, L. A., Waller, B. M., Vic, S. J. & Bard, . A. Classifying chimpanzee facial expressions using muscle action. Emot. 7, 172–181
(2007).
72. Parr, L. A., Waller, B. M. & Heintz, M. Facial expression categorization by chimpanzees using standardized stimuli. Emot. 8, 216–231
(2008).
73. Abramson, L., Marom, I., Petraner, . & Aviezer, H. Is fear in your head? A comparison of instructed and real-life expressions of
emotion in the face and body. Emot. 17, 557–565 (2017).
74. Mertens, P. A. Canine aggression. In BSAVA Manual of canine and feline behavioural medicine (Eds Horwitz, D., Mills, D. & Heath
S) (Gloucester: British Small Animal Veterinary Association, 2002).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

9
SCIENTIFIC REPORTS | (2019) 9:16414 | https://doi.org/10.1038/s41598-019-52938-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Acknowledgements
We would like to warmly thank Prof. Alan Mikhail at Yale University, for his friendly input on the changing
and complicated relationship between humans and dogs in the Muslim world. anks to Bonaventura Majolo
and Emma Watkins at Lincoln University (UK), for help collecting data at Lincoln and for comments on
the manuscript; Daniel Haun at Leipzig University (Germany), for extensive help with data collection; Katja
Kirsche and the Leipzig Research Center for Early Child Development, for invaluable logistic support; and all
the participants in Europe and Morocco, for their precious time and eorts. anks to Filippo Aureli, Natacha
Mendes, Linda Oña, Katrin Schumann, Karine Silva, Sebastian Tempelmann, Zsoa Viranyi and Claudio Tennie
for generously helping with picture validation. anks to the Editorial Board of the Journal and three anonymous
Reviewers for providing feedback that signicantly enhanced the quality of this article. is work was conducted
while F.A. held a Humboldt Research Fellowship for Postdoctoral Researchers (Humboldt ID number 1138999),
and later on a research grant by the German Research Foundation (AM 409/4–1). We acknowledge support from
the German Research Foundation (DFG) and Leipzig University within the program of Open Access Publishing.
Author contributions
F.A. and J.B. designed and coordinated the study. F.A., J.W., C.M.K. and K.K. collected the data. F.A. carried out
the statistical analyses. F.A. wrote the M.S., with substantial input by all the other co-authors.
Competing interests
e authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-52938-4.
Correspondence and requests for materials should be addressed to F.A.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2019
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com