Independent, but still observant—dog breeds selected for functional independence learn better from a conspecific demonstrator than cooperative breeds in a detour task

Abstract

Background
While complex dog–human coexistence has been deeply investigated, there is a relative scarcity of similar knowledge regarding dog–dog interactions. Social learning, a fundamental synchronizing mechanism between dogs and humans, was recently found to be influenced by the functional breed selection of dogs: with the cooperative breeds being more effective learners from a human demonstrator than the independent working breeds were. Here, we investigated whether these differences would also be present when dogs had to learn from another dog and how to effectively perform a detour around a transparent V-shaped obstacle. We tested dogs from 28 independent and 19 cooperative breeds in three consecutive trials. In the control groups, all dogs had to detour on their own the obstacle. In the dog demonstration groups, in trial 1, the subjects had to detour on their own, but before the next two trials, a trained dog showed them the solution.

Results
We found that the performance of the two breed groups was the same in the without demonstration groups. However, after observing the dog demonstrator, the independent dogs learned the task more successfully than the cooperative breeds did. In the case of the independent working breeds, detour latencies significantly dropped along the consecutive trials, and these dogs also showed higher rate of successful detours after observing the demonstrator dog’s action than in the control group.

Conclusions
This is the first study where the consequences of functional breed selection were confirmed in a scenario that involved conspecific social learning in dogs. The results fit well to the ecologically valid framework of the evolutionary past of dog breed formation, in which cooperative breeds were selected for their interactivity with humans, whereas independent breeds often had to work together with their conspecifics.

Graphical Abstract

Full text

Lugosietal. BMC Biology (2024) 22:245
https://doi.org/10.1186/s12915-024-02046-1
RESEARCH ARTICLE Open Access
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License, which permits any non-commercial use, sharing, 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 Creative Commons licence, and indicate if
you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or
parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted 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 licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
BMC Biology
Independent, butstill observant—dog
breeds selected forfunctional independence
learn better fromaconspecic demonstrator
thancooperative breeds inadetour task
Csenge Anna Lugosi1 , Kata Mária Udvarhelyi-Tóth1,2 , Petra Dobos1 and Péter Pongrácz1*
Abstract
Background While complex dog–human coexistence has been deeply investigated, there is a relative scarcity
of similar knowledge regarding dog–dog interactions. Social learning, a fundamental synchronizing mecha-
nism between dogs and humans, was recently found to be influenced by the functional breed selection of dogs:
with the cooperative breeds being more effective learners from a human demonstrator than the independent work-
ing breeds were. Here, we investigated whether these differences would also be present when dogs had to learn
from another dog and how to effectively perform a detour around a transparent V-shaped obstacle. We tested dogs
from 28 independent and 19 cooperative breeds in three consecutive trials. In the control groups, all dogs had
to detour on their own the obstacle. In the dog demonstration groups, in trial 1, the subjects had to detour on their
own, but before the next two trials, a trained dog showed them the solution.
Results We found that the performance of the two breed groups was the same in the without demonstration
groups. However, after observing the dog demonstrator, the independent dogs learned the task more success-
fully than the cooperative breeds did. In the case of the independent working breeds, detour latencies significantly
dropped along the consecutive trials, and these dogs also showed higher rate of successful detours after observing
the demonstrator dog’s action than in the control group.
Conclusions This is the first study where the consequences of functional breed selection were confirmed in a sce-
nario that involved conspecific social learning in dogs. The results fit well to the ecologically valid framework
of the evolutionary past of dog breed formation, in which cooperative breeds were selected for their interactivity
with humans, whereas independent breeds often had to work together with their conspecifics.
Keywords Working dogs, Functional breed selection, Social learning, Conspecific demonstrator
*Correspondence:
Péter Pongrácz
peter.pongracz@ttk.elte.hu
Full list of author information is available at the end of the article
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 2 of 15
Lugosietal. BMC Biology (2024) 22:245
Graphical Abstract
Background
Dogs are undeniably social animals [1], and one of the
main features of their origin is their oft mentioned
descent from a wolf-like, social predator [2]. According
to Kotrschal [3], the reason why these two species came
and stayed so close to each other is probably rooted in the
functional (ecological) similarity between humans and
wolves. Both were cursorial hunters and scavengers liv-
ing in relatively restricted family groups, characterized by
high levels of cooperation. Dogs became a popular model
species for investigation of the evolution of social cog-
nition due to their socio-cognitive capacities [4], which
are considered to be not only human-compatible [5] but
often analogous to human cognitive phenotypes [6, 7],
for example (interspecific) social learning [8], empathy
[9], and rule following [10]. ese socio-cognitive fea-
tures, along with the dependency-based, asymmetrical
dog–human relationship [11], strongly contributed to the
formation of a species that today is considered to be one
of the few ‘true companion animals’ [12].
In a series of comparative studies, it was shown that the
social nature of dogs shows an interesting bias towards
humans [13]. While socialized (tame) wolves were equally
cooperative with human and wolf partners, similarly kept
(group-living) dogs cooperated more often with humans
than with other dogs [14]. From an ecological perspec-
tive, this bias is easy to explain with the feeding biology of
the two species. Since their domestication, dogs became
an omnivorous species, feeding on resources provided
by humans—and parallel with this, another adaptation
of dogs involved a rich set of behaviours based on their
dependency on humans [15]. For instance, they became
capable of forming social bonds with humans easier than
tame wolves do [16], they learn to follow human commu-
nicative signals sooner during their ontogeny [17], and
they diversified barking for communicating with humans
[18].
However, not every dog engages in communica-
tive interactions with humans equally successfully [19].
Various levels of exposure to everyday interactions with
people can differently affect the capacity of dogs to
understand human communication, as was found in dogs
living with their owners or at a shelter [20]. Although
dogs have a rich repertoire of human-related behav-
iours, it was found that among other factors (e.g. train-
ing level [21–23], position in the hierarchy [24], age [25],
the specific pressures of selection that resulted in differ-
ent working dog breeds can have also an effect on the
human-related behaviours of dogs. ose behaviours that
show the strongest between-breed segregation belong to
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 3 of 15
Lugosietal. BMC Biology (2024) 22:245
clusters that characterize either the cooperative (‘bidda-
ble’) or the independent working dogs [26]. Dog breeds
that were selected for performing cooperative tasks (e.g.
herding, retrieving, pointing) with humans look sooner
into the eyes of a human which is considered an impor-
tant initial step for interspecific communication [27].
Furthermore, they usually perform better in such con-
texts where following human communicative cues [28] or
learning from humans [29] was the key. Comparatively,
dog breeds that were selected for working independently
from their handler (e.g. sled dogs, terriers, hounds) did
not perform as successfully as the cooperative breeds did
in the aforementioned contexts. However, in other tasks,
independent working dogs seem to be more inventive,
for instance, if their owner did not see them, they tried
to steal forbidden food [30], they show stronger reward
maximizing tendencies, they visit more readily the
ambiguous reward location in a cognitive bias scenario
[31], and they are more sensitive to reward omission [32].
e dogs in the two functionally different groups had
to meet with completely different expectations during
their work. Dogs selected for cooperative tasks could
benefit from the continuous visual and acoustic cuing
of their handler, which would inhibit decision making
based exclusively on their own assessment of the current
situation. erefore, they were probably more inclined
to observe human behaviour and were also less influ-
enced by other aspects of the environment. is can be
one explanation why Dobos and Pongrácz [29] found
that cooperative dogs achieved higher success rates when
they could observe a human demonstrator, because the
situation of independent dogs was totally opposite. ey
usually did their work without the presence/cueing of the
owner, so they had to make decisions on their own and
focus on the reward as well as other aspects of the envi-
ronment [28].
It is presently unknown whether the functional work-
related selection (i.e. their cooperative or independent
breed designation) of dogs would have an effect on their
willingness (or ability) to learn from another dog. In gen-
eral, intraspecific social learning is a rarely investigated
phenomenon in dogs (but see [33, 34]). In the ‘classic’
detour paradigm, where a V-shaped transparent fence
serves as the obstacle, it was found that dogs readily learn
from a previously unknown conspecific demonstrator
[35]. It was also shown that low-ranking dogs learn more
effectively from another dog than high-ranking dogs did.
At the same time, both subordinate and dominant dogs
learned equally well from a human demonstrator [36].
As recently we found that independent working dogs
did not improve after observing a human demonstration
in the detour task, while the cooperative dogs did [29],
the question arose: would functional work selection also
affect dogs’ ability to learn from another dog? Or, alterna-
tively, does the cooperative or independent working style
of dog breeds affect only their inclination to pay attention
and learn from humans? To answer these questions, we
used the well-known detour paradigm [37], both with
and without a conspecific demonstrator, and tested dogs
from the possible widest selection of breeds from both
the independent and cooperative working dogs.
As we have already tested independent and coopera-
tive working dogs in a without-demonstration condition
in our previous study [29], similarly to the earlier results,
here we expected no difference between their perfor-
mance in the control condition. Regarding the outcome
of the conspecific demonstration condition, we had two
hypotheses. (1) According to the first, functional breed
selection caused such socio-cognitive and behavioural
changes, which made the cooperative and independent
dogs more successful in their specific work, resulting in
different skills for different tasks. erefore, we predicted
that independent and cooperative dog breeds would
show different performance in the intraspecific social
learning task. Based on this hypothesis, we predicted that
cooperative dogs would perform worse in the intraspe-
cific social learning task, because they were selected for
paying attention mostly to a human partner.
(2) Our alternative hypothesis was that the conspecific-
related sociocognitive skills of dogs are fundamental
attributes, rooted in their social nature that goes back
even before their domestication. us, these skills would
only be weakly affected by breed selection, and conse-
quently cooperative and independent dogs would not
show different performance success in the intraspecific
social learning task.
Results
In the case of the dogs’ success (i.e. whether they could
obtain the reward behind the fence within 60 s), we
found a significant effect of the repeated factor (GEE
with binary logistics, trials (χ2(2) = 6.834; P = 0.033), and
the testing groups (χ2(3) = 8.627; P = 0.035). In the case
of the groups with dog demonstration, the success rates
were higher in trials 2 and 3 than in trial 1; however, this
effect was most noticeable in the case of the independ-
ent dogs (see Table1). e independent dogs were more
successful in the dog demonstration condition than in
the control (no demo) condition. However, the coopera-
tive dogs’ success rate did not differ between the control
and the dog demonstration conditions. We did not find
significant association between the dogs’ success rate and
the following potential confounders: sex (χ2(1) = 2.289;
P = 0.130), keeping conditions (χ2(1) = 02.575; P = 0.109),
and training level (χ2(1) = 3.540; P = 0.060).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 4 of 15
Lugosietal. BMC Biology (2024) 22:245
In the case of between-group comparisons of the turn-
ing and reward latencies, we did not find any significant
effect of the four testing groups in trial 1. It means that
there was no initial difference between cooperative and
independent dogs in their problem-solving ability neither
in the control nor in the dog demonstration groups. e
potential confounders (dogs’ sex, age, keeping conditions
and training level) did not have any significant associa-
tion with the two types of latency data (Table2).
By comparing the detour and reward latencies among
the trials separately in the four testing groups (Table3),
we did not find significant differences among the laten-
cies in the two control groups (Fig.1a, b). However, we
found a significant effect of dog demonstration in the
case of the independent working dog group (Fig.1c). No
significant effect of dog demonstration was found in the
case of the cooperative working dog group (Fig.1d).
Task focus did not show significant association with the
testing groups (task focus frequencies: F(3, 112) = 1.130;
P = 0.340; task focus relative durations: F(3, 112) = 0.887;
P = 0.450) and with the repetition of the trials (task focus
frequencies: F(1, 112) = 0.674; P = 0.413; task focus rela-
tive durations F(1, 112) = 0.240; P = 0.625).
In the case of frequency of owners’ encouraging
utterances during the tests, we found a significant effect
both of the repeated trials (F(1, 111) = 8.774; P < 0.004)
and the testing groups (F(3, 111) = 20.764; P < 0.001).
According to the Tukey post hoc test, owners more
often encouraged their dogs in the two control (no
demo) groups than in the conditions with the dog dem-
onstration. In the case of the repeated trials, owners
encouraged their dogs more often in trial 1 than they
did in trial 3.
e frequency of looking back towards the humans
during the test (Fig.2) showed a significant association
with both the repeated trials (F(1, 112) = 16.705; P < 0.001)
and also with the testing groups (F(3, 112) = 147.229;
P < 0.001). Looking back towards the humans was least
frequent in the case of independent dogs with dog dem-
onstration, and dogs looked back significantly more fre-
quently in the two control (no demo) groups, where the
cooperative dogs looked back most frequently. e fre-
quency of looking at the humans decreased during the
repeated trials.
Frequency of side alternations at the corner of the fence
showed significant association with the repeated trials
(F(1, 112) = 10.432; P = 0.002). According to this, dogs
decreased their side alternation frequency in trial 3. We
did not find significant association between the side alter-
nation frequency and testing groups (F(3, 112) = 0.724;
P = 0.540).
Table 1 Ratio of successful dogs in the four testing groups across the three trials (success = the dog obtained the reward within 60 s)
Trial 1 Trial 2 Trial 3
Independent/no demonstration 50% (11/22) 40.1% (9/22) 50% (11/22)
Cooperative/no demonstration 68% (17/25) 64% (16/25) 68% (17/25)
Independent/dog demonstration 55.6% (20/36) 72.2% (26/36) 77.8% (28/36)
Cooperative/dog demonstration 48.5% (16/33) 60.6% (20/33) 57.6% (19/33)
Table 2 Results of the between-group Cox regression analysis
in the case of trial 1 for the latencies of turning at the rear end
of one of the wings of the fence and latencies of reaching the
target
Dependent variable Variable (xed
factors) Chi-square Df P
Turn (detour) latency Testing group 4.594 3 0.204
Keeping 1.022 1 0.312
Training 6.121 5 0.295
Sex 1.567 1 0.211
Reward latency Testing group 5.290 3 0.152
Keeping 1.151 1 0.283
Training 6.279 5 0.280
Sex 1.520 1 0.218
Table 3 Results of the Cox regression analysis in the case of the
four testing groups. Significant effects are highlighted with bold
letters
Dependent variable Testing group Chi-square df P
Turn (detour) latency Independent/no
demo 0.596 2 0.742
Cooperative/no demo 0.535 2 0.765
Independent/dog
demo 8.807 2 0.012
Cooperative/dog
demo 2.066 2 0.356
Reward (detour)
latency Independent/no
demo 0.558 2 0.757
Cooperative/no demo 0.520 2 0.771
Independent/dog
demo 7.985 2 0.018
Cooperative/dog
demo 1.635 2 0.441
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 5 of 15
Lugosietal. BMC Biology (2024) 22:245
In the case of dogs that were successful in trial 1, we
checked whether they followed their own direction from
trial 1 in the subsequent trials. Additionally, we analysed
whether the dogs in the two groups with the dog demon-
strator followed the demonstrator’s (inward) side choice
in the two demonstration trials (trials 2–3). Based on the
results of one-sample Wilcoxon signed-rank tests (hypo-
thetical value = 1), subsequent concordance with the own
original direction was significant in both cooperative
and independent dogs (cooperative (T = 3.530; N = 32;
P < 0.001); independent (T = 2.600; N = 31; P = 0.009)).
Neither the cooperative (T = − 0.426; N = 27; P = 0.670)
nor the independent (T = − 1.095; N = 33; P = 0.273)
dogs followed the demonstrated side while attempting to
detour the fence.
Discussion
e main goal of our experiment was to investigate
whether functional dog breed selection could also affect
the performance of dogs in an intraspecific social learn-
ing context. We tested several breeds from both inde-
pendent and cooperative working dogs in a transparent
obstacle detour task. According to our main finding,
while detour latencies in trial 1 did not differ across all
four testing groups, independent breeds learned more
effectively from observing a dog demonstrator who
showed them how to effectively detour the V-shaped
fence than the cooperative dogs did. Unlike the dogs
that belong to cooperative working breeds, independent
dogs performed the detours faster after the demonstra-
tion compared to their first trial. Detour latencies did not
decrease in the control groups. Additionally, independ-
ent breeds reached higher success rates in the dog dem-
onstration group than they did in the control condition.
Interestingly, although cooperative dogs in the control
group showed relatively high success rate in trial 1, this
did not improve along the subsequent trials, and this was
also true for the control group of independent dogs. A
previous study [29] that used the same detour task found
that cooperative working dogs improved their detour-
ing speed after observing a human demonstrator, while
independent dogs did not become faster in their detours
after witnessing human demonstration. Pongrácz et al.
[35] showed that in this situation dogs (at least the ones
who were living with conspecific companions at home)
were able to learn equally well from an unknown dog
and human demonstrator. However, while [35] tested an
inclusive sample of subjects (irrespective of breed), in
our present study, we took into consideration the work-
related selection history of the breeds, which allowed us
to discover more intricate details regarding the breed-
related differences in social learning of dogs. At first, it
Fig. 1 Cumulative proportions of independent (a N = 22 and c N = 36) and cooperative working dogs (b N = 25 and d N = 33) that per formed
a successful detour (reaching the reward) in the three trials of the experiment. a and b show the reward latencies in the control (no demonstration)
condition, and c and d show the results in case of dog demonstration
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 6 of 15
Lugosietal. BMC Biology (2024) 22:245
was found that dogs belonging to cooperative working
breeds learned better from an (unknown) human [29],
while according to our new results, independent dog
breeds benefitted more from observing the demonstra-
tion of an unknown dog.
ere is a long series of publications about dogs’ per-
formance in the detour paradigm (see for a review [8]),
where dogs did not improve neither in their success
rates nor in their detour speed across the consecutive
no-demonstration trials. is phenomenon was recently
confirmed by Dobos and Pongrácz [29] in cooperative
and independent working dogs as well as in the control
groups of present study. Compared to the similar per-
formance of the two breed groups across the consecu-
tive trials in the control (no demonstration) condition, if
we see a difference in the improvement of the two breed
groups during the trials preceded with dog-demonstra-
tion, we can attribute the result to social learning.
Training levels [28] and keeping conditions [38] were
both found earlier as being potentially important modify-
ing factors in various problem-solving tasks, where dogs
had to rely on human communicative cues. Scandurra
[23] found that Labrador Retrievers (a highly cooperative
hunting dog breed) could effectively learn from a conspe-
cific demonstrator—a result that seemingly contradicts
our findings. However, in that study, the training level
of the subjects proved to be a decisive factor, as only the
Labradors from the advanced level of water rescue train-
ing benefited from observing a dog demonstrator, while
the Labradors from the basic training level did it less.
In the current study, these and additional potential con-
founders (sex, age, keeping conditions and training level
of the subjects) did not have any significant association
with the detour latencies or the success rate of dogs.
We also analysed such behavioural parameters that
could mirror the potentially existing differences among
Fig. 2 The frequency of looking at the humans during the test. Coop/nodemo = cooperative dogs without demonstration (N = 25);
Ind/nodemo = independent dogs without demonstration (N = 22); Coop/demo = cooperative dogs with demonstration (N = 33); Ind/
demo = independent dogs with demonstration (N = 36). EMM, estimated marginal means
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 7 of 15
Lugosietal. BMC Biology (2024) 22:245
the dogs’ motivation or persistence to solve the task (‘task
focus’, side alternation) or the efforts that the owners
put into the encouragement of the dogs during the tri-
als. However, task focus and side alternation showed no
significant association with the testing groups, and side
alternation dropped only with the repetition of trials,
suggesting that as dogs generally became more effective
with the detour, they struggled less at the corner of the
fence. According to earlier papers (e.g. [39, 40]), the more
frequent side alternations coincided with a weaker per-
formance in the detour test. e encouraging utterances
provided by the owners also became less frequent during
the consecutive trials, which can be also the sign that as
owners saw their dogs’ improving performance, they were
less inclined to help them verbally. is notion is further
supported with the result that showed an increased fre-
quency of encouraging in the control groups compared
to the demonstration condition. As dogs performed less
successfully/improved less in the control groups, again,
here the owners could be more motivated to help them
out with verbal encouraging. Summing it up, the weaker
performance (lack of improvement in trials 2 and 3) can-
not be specifically explained with the lack of motivation
or encouragement in the case of the cooperative breed
group in the dog demonstration condition.
In the case of dogs that were successful in trial 1, subse-
quent concordance with their own original direction was
significant in both cooperative and independent dogs,
which could be expected after the similar results of our
earlier papers [39, 40]. Neither the cooperative nor the
independent dogs followed the demonstrated side while
attempting to detour the fence—which is no surprise as it
was found earlier that dogs only copied the demonstrated
side if the demonstrator used the same wing of the fence
for the inward and outward travel, and the dog itself did
not have a priori experience with its’ own successful
detour [40].
We only tested subjects from single-dog households;
thus, the potential effect of subjects’ dominance rank
could be excluded. Pongrácz etal. [36] found that in the
case of subjects from multi-dog households, the subor-
dinate dogs learned better from a (previously unknown)
dog demonstrator, compared to the dominant dogs, while
dominant and subordinate dogs learned equally well
when they observed a human demonstrator. In that study,
the social learning performance of the singleton dogs fell
between the dominant and subordinate dogs, regard-
less of whether the demonstrator was a dog or a human.
erefore, to avoid the confounding effect of social rank
in the case of dog demonstration, the use of individually
kept dogs was the obvious option.
Currently, the differences between dog breeds have
been highlighted in behavioural sciences and there are
several approaches to the grouping and comparing of
breeds. Some researchers use convenience samples
(i.e. to compare the most popular/available breeds at
a given location and time [41, 42]), which are a favour-
able method for explorative research, but it makes the
formulating of hypotheses about the potential factors
behind the found differences difficult. Ancestry-based
approaches, aided by the recent advances of molecular
genetics and the subsequent construction of evolution-
ary ‘trees’ for dog breeds (e.g. [43]), are another prom-
ising method, with the advantage of using biologically
relevant hypotheses based on the genetic distance from
the common ancestor of all dog breeds (e.g. [44, 45]). At
the same time, we think that in terms of ecological rel-
evance, investigations that are based on functional breed
selection (e.g. [27, 29, 46]) could provide one of the most
relevant approaches to the analysis of breed-related
behavioural differences in dogs. is approach does not
focus on the specific working task of the dogs but con-
centrates on how they are supposed to do it—together, in
strong interactivity with a human, or alone and indepen-
dently. Especially for the investigation of such socio-cog-
nitive traits that include interactions with humans (e.g.
communication [28]; social referencing [27]; cooperation
[47]), the categories that are built on functional breed
selection could provide excellent insight into those abili-
ties which were adaptive for success in a given situation.
A commonly encountered limitation of breed-spe-
cific comparative behavioural investigations is the non-
representative breed composition of the breed groups.
Researchers often test only a few (over-represented)
breeds per breed group; therefore, the interpretation
of the results leaves open the opportunity for several
alternative hypotheses (e.g. [31, 48]). To avoid this, we
planned to test as many breeds as possible in both groups
(28 breeds from the independent working dogs and 19
breeds from the cooperative working dogs), without
overrepresentation of any breed.
Conclusions
As often genetically distant related dog breeds were
selected for either cooperative or independent work
tasks, the division of dog breeds to independent and
cooperative working dogs provides an overarching sys-
tem for behavioural comparisons. e resulting catego-
ries are not confounded by a strong overlap with the
genetic clades of breeds [29, 49]; thus, this method allows
us to categorize and compare dog breeds regardless of
their genetic distance or popularity.
Our findings strongly suggest that functional breed
selection could influence the ability or willingness of
dogs to learn from their conspecifics. ese results fell
in line with our prediction that cooperative dogs would
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 8 of 15
Lugosietal. BMC Biology (2024) 22:245
perform weaker in the intraspecific social learning task,
because they were selected for paying attention mostly
to their human partner. Cooperative dogs were usually
more successful in tasks where they had to pay attention
to humans (gaze following [17]; learning from a human
demonstrator [29]). In our study, cooperative dogs more
often looked back to the owners during the test, than
independent dogs did, supporting the view that in the
case of a difficult task, cooperative dogs rather rely on
the human and wait for the visual or ostensive help or
instruction from their owner. Cooperative breeds were
selected to solve tasks and make decisions based on the
instruction of their human handler. In contrast, inde-
pendent dogs could benefit from concentrating on the
reward, potentially because of their stronger persistence
in problem-solving tasks [26, 50] and their tendency
for reward-maximizing strategies [30]. Most impor-
tantly, because working together with other dogs (e.g.
scent hounds, dachshunds, sled dogs) was more com-
monly included to their original working conditions, we
can expect more focused attention towards the actions
of other dogs from the independent breeds. A further
explanation for the better performance of independ-
ent dog breeds in the dog demonstration group could be
connected to the more ‘competitive’ (or less ‘amicable’)
nature of some of these breeds (e.g. some of the terrier
breeds). According to this idea, these dogs could become
more interested in the demonstrator dog’s action when
they noticed that the demonstrator was about to obtain
the reward. In the future, this would provide an interest-
ing possibility for detailed follow up studies.
Methods
Subjects
We tested N = 120 adult (1 year or older) companion
dogs in the presence of their owners. Dog owners were
recruited through advertisements placed on social
media. e subjects were all purebred dogs that belonged
to either a cooperative or an independent working dog
breed. We tested subjects that were the only dog of the
given owner, to avoid the confounding effect of a dogs’
rank in the social learning test [36]. Most importantly,
we recruited a large variety of breeds, thus avoiding the
overrepresentation of popular breeds in our sample.
None of the included breeds had more than three rep-
resentatives in any given experimental group. We tested
28 breeds from the independent working dog group and
19 breeds from the cooperative working dog group. e
basic demographic details of the subjects are provided
in Table4. Furthermore, we collected data about their
keeping conditions (indoor only, indoor-outdoor and
outdoor-only) as well as the level of training the dogs
received (none; training at home; course at dog school;
regular dog school; assigned trainer; specific sports/work
training). We only recruited such companion dogs for the
experiments, where the owners were comfortable with
letting their dogs off leash in a safe outdoor environment,
and they confirmed that their dogs have no problems
with being in the company of other (friendly) dogs.
Equipment
All tests were performed outdoors, in a grassy area at
the campus site of the Eötvös Loránd University. We had
an open, empty area for the tests that was far away from
buildings, roads, and frequented places, which provided a
calm, undisturbed environment. We ran all tests between
September 2023 and April 2024. Our equipment was
identical to the V-shaped fence described in the article of
[39], which consists of transparent wire mesh, stretched
tightly over a light steel frame. e fence was firmly
fixed into the ground with protruding steel pegs, so that
its lower edge was just above the soil, which prevented
the dogs from digging or crawling under the fence. e
intersecting angle of the fence was set to 80 degrees. Each
wing of the V-shaped fence was 3-m long, and its height
was 1m. e starting point was 2m away from the corner
of the V-shaped fence in the midline. e reward was a
favourite toy or food, selected and brought by the owner,
based on the preference of the dog. e experimenter
always put the chosen reward on a white plate which was
placed to the middle of the inner corner, 10cm from the
fence. We recorded the tests with a video camera (BLOW
Go Pro4U) that was positioned on a tripod and placed
slightly aside of the midline of the V-shaped fence. e
outlay of the testing area with the V-shaped fence can be
seen in Fig.3.
General procedure
e testing procedure was similar to the one used
in [29]. After arriving to the testing site, dog own-
ers (O) gave their written informed consent; they
entered the testing area with the experimenter (E),
which was always a young woman (C. A. L. or P. D.).
e E explained to them what to do and what not to do
during the test. e dog was able to walk around the
area on leash, but it was not allowed to go behind the
V-shaped fence at this time. We asked the O to keep the
dog on leash and position it to the starting point. e
O stood right behind the dog, and both of them faced
the fence. e E called the dog’s attention (by calling
its’ name, and saying for example, ‘Look’). en, the E
walked to the intersecting angle of the fence, conspicu-
ously holding a piece of food (or the toy) in her hand,
leaned over the fence, and dropped the reward to the
inner corner of the fence onto the plate. e E showed
her empty hands to the dog, then she returned to the
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 9 of 15
Lugosietal. BMC Biology (2024) 22:245
Table 4 Basic demographic data of those participating dogs who were included in the statistical analyses. Age is given in years.
Demographic data are missing in case of two subjects. We indicated test group assignments as well. Coop-d = cooperative dogs with
detour demonstration; Coop-n = cooperative dogs without demonstration; Ind-d = independent dogs with detour demonstration;
Ind-n = independent dogs without demonstration
Dog’s ID Breed Sex Reproductive status Age Test group
1 Australian Cattle Dog Male Neutered 3 coop-d
2 Australian Cattle Dog Female Intact 7 coop-d
3 Australian Shepherd Female Spayed 4 coop-d
4 Australian Shepherd Male Intact 1 coop-d
5 Border Collie Male Neutered 4 coop-d
6 Border Collie Female Spayed 2.5 coop-d
7 Border Collie Female Spayed 10 coop-d
8 English Cocker Spaniel Male Neutered 3 coop-d
9 English Cocker Spaniel Female Spayed 2 coop-d
10 English Cocker Spaniel Female Spayed 1 coop-d
11 German Wirehaired Pointer Male Intact 1 coop-d
12 Golden Retriever Female Intact 4 coop-d
13 Golden Retriever Female Intact 1.5 coop-d
14 Golden Retriever Male Intact 14 coop-d
15 Labrador Retriever Female Intact 2 coop-d
16 Labrador Retriever Male Neutered 3 coop-d
17 Labrador Retriever Female Spayed 3 coop-d
18 Lagotto Romagnolo Female Spayed 3 coop-d
19 Lagotto Romagnolo Female Spayed 5.5 coop-d
20 Lagotto Romagnolo Female Spayed 6 coop-d
21 Malinois Male Intact 2.5 coop-d
22 Malinois Male Intact 3 coop-d
23 Malinois Male Intact 9 coop-d
24 Mudi Female Spayed 3 coop-d
25 Mudi Male Neutered 1 coop-d
26 Mudi Male Intact 2 coop-d
27 Puli Female Intact 1.5 coop-d
28 Puli Male Intact 2.5 coop-d
29 Pumi Female Spayed 1 coop-d
30 Pumi Male Intact 3 coop-d
31 Pumi Male Intact 4 coop-d
32 Shetland Sheepdog Female Spayed 6 coop-d
33 Vizsla Female Spayed 2 coop-d
34 Australian Shepherd Male Neutered 2 coop-n
35 Border Collie Male Neutered 3 coop-n
36 Border Collie Female Spayed 2.5 coop-n
37 Border Collie Female Spayed 6.5 coop-n
38 Border Collie Female Intact 5 coop-n
39 Briard Female Intact 3 coop-n
40 Cardigan Welsh Corgi Male Neutered 5.5 coop-n
41 Golden Retriever Male Neutered 10 coop-n
42 Labrador Retriever Male Intact 4 coop-n
43 Labrador Retriever Female Spayed 10 coop-n
44 Labrador Retriever Female Spayed 5 coop-n
45 Lagotto Romagnolo Female Spayed 1.5 coop-n
46 Mudi Female Spayed 2 coop-n
47 Mudi Male Intact 11 coop-n
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 10 of 15
Lugosietal. BMC Biology (2024) 22:245
Table 4 (continued)
Dog’s ID Breed Sex Reproductive status Age Test group
48 Mudi Female Spayed 5.5 coop-n
49 Mudi Female Spayed 5 coop-n
50 Mudi Male Intact 1 coop-n
51 Pumi Male Intact 9 coop-n
52 Rottweiler Male Neutered 6 coop-n
53 Rough Collie Male Intact 9 coop-n
54 Shetland Sheepdog Male Neutered 5 coop-n
55 Shetland Sheepdog Female Intact 8 coop-n
56 Tervueren coop-n
57 Vizsla Male Intact 7 coop-n
58 Weimaraner Male Neutered 5.5 coop-n
59 Airedale Terrier Female Intact 3 ind-d
60 Airedale Terrier Male Neutered 4 ind-d
61 Akita Inu Male Intact 6 ind-d
62 Akita Inu Female Intact 2 ind-d
63 American Pitbull Terrier Female Intact 2 ind-d
64 American Pitbull Terrier Female Spayed 4.5 ind-d
65 American Staffordshire Terrier Male Intact 1 ind-d
66 Basset Hound Female Spayed 2 ind-d
67 Borzoi Male Intact 2.5 ind-d
68 Dachshund Female Spayed 3 ind-d
69 Dachshund Female Spayed 1.5 ind-d
70 Fox Terrier Male Neutered 10 ind-d
71 Fox Terrier Female Intact 1 ind-d
72 Hungarian Greyhound Male Neutered 2.5 ind-d
73 Hungarian Greyhound Male Intact 1.5 ind-d
74 Ibizan Hound Female Spayed 4 ind-d
75 Irish Terrier Male Intact 4.5 ind-d
76 Irish Terrier Male Neutered 3 ind-d
77 Jack Russell Terrier Female Spayed 3 ind-d
78 Jack Russell Terrier Male Neutered 3 ind-d
79 Jack Russell Terrier Female Intact 3 ind-d
80 Komondor Female Intact 2.5 ind-d
81 Patterdale Terrier Male Intact 1 ind-d
82 Patterdale Terrier Male Intact 3 ind-d
83 Rhodesian Ridgeback Male Intact 4 ind-d
84 Shiba Inu Male Neutered 3 ind-d
85 Shiba Inu Female Spayed 2 ind-d
86 Shiba Inu Male Neutered 7 ind-d
87 Siberian Husky Female Spayed 4 ind-d
88 Staffordshire Terrier Female Spayed 4 ind-d
89 Transylvanian Hound Male Neutered 7 ind-d
90 Transylvanian Hound Male Neutered 4 ind-d
91 Welsh Terrier Female Spayed 5.5 ind-d
92 Whippet Male Intact 3 ind-d
93 Yorkshire Terrier Male Intact 3 ind-d
94 Kuvasz Male Intact 5 coop-d
95 American Pitbull Terrier Female Spayed 9 ind-n
96 Basset Hound Female Spayed 1 ind-n
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 11 of 15
Lugosietal. BMC Biology (2024) 22:245
starting point and stood about 1m behind the O. After
this, E told the O to release the dog. e O was told to
encourage the dog to solve the task (i.e. ‘bring it’, ‘get
it’), but they were not allowed to use commands such
as ‘Go left/right’, ‘Go around’. Gestural commands were
also disallowed. e dog had 60s to solve the task—if
it performed a successful detour within this time limit
and obtained the reward from behind the fence, the O
had to recall the dog to the starting point, and the next
trial started. If the dog did not perform a successful
detour in 60s, the trial ended, and the O had to call the
dog back to the starting point. We tested every subject
in three consecutive trials. In each case, the first trial
was exactly how we described it above. In the case of
Table 4 (continued)
Dog’s ID Breed Sex Reproductive status Age Test group
97 Borzoi Female Intact 1 ind-n
98 Cane Corso Male Neutered 6 ind-n
99 Dachshund Female Spayed 2.5 ind-n
100 Dachshund Male Neutered 11 ind-n
101 Dachshund Male Neutered 5 ind-n
102 Fox Terrier Male Neutered 3 ind-n
103 Galgo Español Female Spayed 7.5 ind-n
104 Hovawart Male Intact 1.5 ind-n
105 Hungarian Greyhound Female Intact 3 ind-n
106 Hungarian Greyhound Male Intact 1.5 ind-n
107 Hungarian Greyhound Female Spayed 2 ind-n
108 Irish Terrier ind-n
109 Jack Russel Terrier Male Neutered 3.5 ind-n
110 Komondor Female Spayed 7.5 ind-n
111 Ogar Polski Female Spayed 5 ind-n
112 Parson Russell Terrier Female Spayed 8 ind-n
113 Siberian Husky Male Neutered 5.5 ind-n
114 Transylvanian Hound Male Neutered 6 ind-n
115 West Highland White Terrier Female Spayed 1 ind-n
116 Yakutian Laika Male Intact 1 ind-n
Fig. 3 Schematic drawing of the experimental setup. Measurements of the V-shaped fence and the distance of the starting point are
given in meters. The route of the dog demonstrator is symbolized with the ‘white dogs’
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 12 of 15
Lugosietal. BMC Biology (2024) 22:245
the dogs tested with demonstration, before the next
two trials, they observed the successful detour done by
the demonstrator dog.
Dog demonstration
e demonstrator dog was an intact female Border Col-
lie (5years old), trained to solve the detour task fast and
easy, repeatedly, without ceasing motivation levels. e
dog’s owner was present, and she handled the demon-
strator dog. While waiting for the demonstration task,
the handler kept the demonstrator dog approximately
2 m away from the fence, behind the starting point.
In the case of trials 2 and 3, the O positioned the sub-
ject dog to the starting point, then E placed the reward
behind the fence as in trial 1. en, the handler released
the demonstrator dog, and encouraged it to detour the
fence. e demonstrator dog performed the detour and
subsequently either ate the reward or picked up the toy
then returned to its handler. e handler put the demon-
strator dog back on leash. After this, the E again placed
the reward behind the fence, and the trial continued as
described in trial 1. e handler stood 2m behind the
starting point with the demonstrator dog, while the test
subject finished its trial. is procedure was repeated
once again in trial 3. We always used the same type of
reward for the demonstrator and subject dog in a given
test.
Control (no demo) groups
In the control groups, the dogs had to perform three
identical trials that were exactly the same as described in
the ‘General procedure’ section.
Experimental groups
Each subject participated only once in the experiment.
We assigned the dogs to the experimental groups by
paying attention to the balanced distribution of sex, age,
keeping condition, and training level of the subjects. e
following experimental groups were formed:
– Independent dogs/control (no demonstration)—
N = 22.
– Cooperative dogs/control (no demonstration)—
N = 25.
– Independent dogs/detour demonstration—N = 36.
– Cooperative dogs/detour demonstration—N = 33.
We determined the desired sample size by using the
equation for finite populations
z (z-score) = 1.96 for 95% confidence level, ℇ (margin
of error) = 0.05; p (population proportion) = 0.50. We
expected that the population of suitable dogs (N) for our
test would be 150 (based on previous subject recruiting
campaigns in social media, it means that within a rea-
sonable time frame we could invite no more than 150
subjects that belong to the targeted breed groups). e
calculated sample size thus was N = 108. We actually
had the opportunity to test a slightly higher number of
subjects (N = 120), expecting that some of the subjects
would need to be excluded for various reasons (e.g. lack
of motivation, technical issues, non-complying owner).
e exact details of exclusions and the actual number of
excluded subjects are provided in the next section.
Exclusions
We excluded those subjects that were not motivated to
perform any trials or lost interest for further performance
during the test. A dog was considered as having lost
interest if it did not approach the V-shaped fence upon
its release from the starting point or only did it once. We
had to exclude altogether N = 4 dogs for this reason, each
from the independent/demo group.
Behavioural coding
Each test was video recorded. We used Solomon Coder
(beta 19.08.02, Copyright by András Péter) for the extrac-
tion of data from the video sequences. Table5 shows the
behavioural variables we used for the analysis. We used
the ethogram that was described by Dobos and Pongrácz
[29]. For inter-coder reliability analysis, 15% of the videos
were re-coded by a second experimenter, who was una-
ware of the experimental hypotheses.
Statistical analyses
Raw data, used for statistical analyses, can be found in
Additional_file1.csv. All statistical analyses were per-
formed with the SPSS (version 29) software. e suc-
cess rate of dogs was analysed with GEE (generalized
estimating equations, with binary logistics). Trials were
included as repeated factor, testing group as independent
factor, and we used the dog’s sex, keeping conditions and
training level as covariates. We included the biologically
relevant 2-way interactions (breed group x training and
keeping) to the model.
Turn and reward latencies were analysed separately.
Latencies were analysed with Cox regression models. At
first, we ran a separate between-group comparisons in
the case of trials 1. As in trial 1, the dogs did not receive
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 13 of 15
Lugosietal. BMC Biology (2024) 22:245
demonstration, and their latencies here served as base-
line, and by comparing them between the four groups, we
could assess whether the dogs in the groups were equally
effective (or ineffective) problem solvers on their own.
Besides the testing group, the dog’s sex, keeping condi-
tions, and training level were also added to the model. In
separate Cox regression models, we analysed whether the
dogs in the four groups improved their detour efficiency
across the trials. is analysis would reveal if the dogs (1)
could improve their detour latencies on their own (across
the three control trials) and (2) if they have benefited
from the dog demonstration and performed the detour
with shorter latencies along the consecutive trials.
General linear model with repeated measures was used
for the analysis of task focus frequency and relative dura-
tion as well as for the frequencies of owner encourage-
ment, dog’s side alternation, and looking back at humans.
Trials were used as repeated factor and the testing group
as fixed factor. We also included the 2-way interaction of
trial and testing group to the model.
Finally, concordance between the dog’s side choice in
trial 1, and in the subsequent two trials, it was analysed
with Wilcoxon signed-rank test. We set the hypothetical
value at 1. e minimum value was 0 and the maximum
concordance was 2 (when the dog used the same side of
the fence in all three trials). We only included those dogs
to this analysis who had a successful detour attempt in
trial 1. Additionally, in the case of the dog demonstra-
tion condition, we also analysed whether the dogs fol-
lowed the inward direction of the demonstrator dog in
trials 2 and 3. e hypothetical value again was 1 in the
Wilcoxon signed-rank test. Each dog was included to this
analysis.
Whenever we analysed the two-way interactions, we
performed a backward model selection, removing the
non-significant interactions. In each instance, results of
the final (simplest) models were reported.
To check the reliability of the coding method, an inde-
pendent observer (who was unaware of the test hypothe-
ses) coded video footage from 15 randomly chosen dogs.
Latency and frequency data were analysed by Spear-
man’s rho correlation. Based on the analysis, our cod-
ing procedure was reliable (detour latency: R(45) = 1.000;
reward latency: R(45) = 0.999; P < 0.001; ‘ta sk fo cus’
duration: R(45) = 0.999; P < 0.001; ‘encouragement’ fre-
quency: R(45) = 0.894; P < 0.001; ‘looking back’ frequency:
R(45) = 0.875; P < 0.001; ‘side alternation’ frequency:
R(45) = 0.986; P < 0.001).
Table 5 The list of behavioural variables we used during the video coding, based on [29]
Behavioural variable Unit Description
Success Occurrence (binary) The dog reached the reward after performing a successful detour; then, it touched/consumed
the reward
Reward (detour) latency (s) The time elapsed between the moment of releasing the dog by the owner at the starting point
and the dog’s arrival at the reward (i.e. after a successful detour). In the case of an unsuccessful
trial, 60 s was assigned
Turn (detour) latency (s) The time elapsed between the moment of releasing the dog by the owner at the starting
point and the dog’s arrival at the rear end of one of the wings of the fence. ‘Arrival’ happened
when the dog turned in at the rear end of the fence
Detour direction (inward) Left or right or 0 The side of the fence where the dog went during its preceding successful detour attempt. In
unsuccessful trials, 0 was assigned
Concordance (own direction) 1 or 0 per trial Whether the dog performed the detour (inward) in trials 2–3 on the same side that it used
in trial 1
Concordance (demo direction) 1 or 0 per trial Whether the dog performed the detour (inward) in trials 2–3 on the same side that the demon-
strator dog used (N/A in the case of the control condition)
Task focus duration Relative duration Task focus describes the dogs’ tendency to leave the close vicinity of the fence during its
attempts to detour around it. The dog stepped over the boundary line drawn 1 m from the cor-
ner of the fence while moving away from the fence and back towards the owner. The value
was calculated from the total time the dog spent away from the fence, divided by the reward
detour latency
Task focus frequency Frequency per trial The value was calculated as the number of events when the dog was away from the fence
divided by the reward detour latency
Looking back 1/s During attempts to detour, the dog turned towards the owner/experimenter (by turning its
head only or with full body orientation) and looked at them. The number of looking back events
was then divided by the reward detour latency
Side alternation (at corner) 1/s The number of swapping the side events at the corner of the fence during the dog’s attempts
to detour divided by the reward detour latency
Encouragement (by owner) 1/s The number of distinct verbal utterances (at least 1 s between two) given by the owner dur-
ing the dog’s attempts to detour divided by the reward detour latency
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 14 of 15
Lugosietal. BMC Biology (2024) 22:245
Abbreviations
Coop-d Cooperative dogs with detour demonstration
Coop-n Cooperative dogs without detour demonstration
Ind-d Independent dogs with detour demonstration
Ind-n Independent dogs without detour demonstration
E Experimenter
O Owner (of the dog)
Supplementary Information
The online version contains supplementary material available at https:// doi.
org/ 10. 1186/ s12915- 024- 02046-1.
Additional file 1.
Acknowledgements
The authors are grateful to Celeste R. Pongrácz for proofreading the manu-
script and to Alexa Iván for the graphical abstract and illustrations.
Authors’ contributions
All authors read and approved the final manuscript. Conceptualization:
PP and CAL; methodology: PP and CAL; validation: LCA and KMUT; formal
analysis: PP; investigation: CAL, PD and KMUT; resources: PP and KMUT; data
curation: CAL, PD and PP; writing—original draft preparation: PP, CAL; writ-
ing—review and editing: PP, CAL, KMUT, PD; visualization: PP; supervision: PP;
project administration: CAL and PD.
Funding
Péter Pongrácz and Petra Dobos were supported by the Hungarian National
Research, Development and Innovation Office (NKFIH, Grant # K143077).
Petra Dobos was supported by the New National Excellence Program of
the Ministry for Innovation and Technology (ÚNKP-23–2-II-ELTE-164) and
by the University Research Fellowships (EKÖP-24–2-I-ELTE-394). Kata Mária
Udvarhelyi-Tóth was supported by the Hungarian Academy of Sciences via
a grant to the MTA-ELTE ‘Lendület/Momentum’ Companion Animal Research
Group (grant no. PH1404/21).
Data availability
All data generated or analysed during this study are included in this published
article [and its supplementary information file Additional_file1.csv].
Declarations
Ethics approval and consent to participate
The applied procedure was fully reward-based and non-invasive. The study
was endorsed by the Institutional Animal Welfare Committee (Eötvös Loránd
University, Budapest), who checked and accepted the experimental method
(Ref. no.: ELTE-AWC-013/2023). The full procedure was performed in accord-
ance with the Hungarian regulations on animal experimentation and the
Guidelines for the use of animals in research described by the Association for
the Study of Animal Behaviour (ASAB). The research did not include human
experimentation or the collection of sensitive data from the owners, and
because of this, human ethical approval was not required in Hungary. We
obtained a written informed consent from every dog owner prior to testing,
and participation was voluntary. Before the test, dog owners were informed
about the use of data, the aim of the experiment, the procedure, and the
option to stop the test any time they thought their dog experienced any sort
of unacceptable stress (none of the owners opted to stop the test).
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Author details
1 Department of Ethology, ELTE Eötvös Loránd University, Budapest, Hungary.
2 MTA-ELTE Lendület “Momentum” Companion Animal Research Group, Buda-
pest, Hungary.
Received: 22 February 2024 Accepted: 16 October 2024
References
1. Bradshaw J, Rooney N, Serpell J. Dog social behavior and communica-
tion. In: Serpell J, editor. The domestic dog: Its evolution, behavior and
interactions with people. Cambridge: Cambridge University Press; 2017.
p. 133–59.
2. Scott JP. The evolution of social behavior in dogs and wolves. Am Zool.
1967;7(2):373–81.
3. Kotrschal K. How wolves turned into dogs and how dogs are valuable in
meeting human social needs. PAIJ. 2018;1(1):6.
4. Aria M, Alterisio A, Scandurra A, Pinelli C, D’Aniello B. The scholar’s best
friend: research trends in dog cognitive and behavioral studies. Anim
Cogn. 2021;24(3):541–53.
5. Miklósi Á, Topál J. What does it take to become ‘best friends’? Evo-
lutionary changes in canine social competence. Trends Cogn Sci.
2013;17(6):287–94.
6. Miklósi Á, Topál J, Csányi V. Comparative social cognition: what can dogs
teach us? Anim Behav. 2004;67(6):995–1004.
7. Topál J, Miklósi Á, Gácsi M, Dóka A, Pongrácz P, Kubinyi E, Virányi Z. The
dog as a model for understanding human social behavior. In: Brockmann
HJ, Roper TJ, Naguib M, Wynne-Edwards KE, Mitani JC, Simmons LW,
editors. Advances in the Study of Behavior, vol. 39. Burlington: Academic
Press; 2009. p. 71–116.
8. Pongrácz P. Social learning in dogs. In: In The Social dog (ed.J. Kaminski, S
Marshall-Pescini). Cambridge: MA: Academic Press; 2014. p. 249–93.
9. Silva K, De Sousa L. ‘Canis empathicus’? A proposal on dogs’ capacity to
empathize with humans. Biol Lett. 2011;7(4):489–92.
10. Topál J, Kubinyi E, Gácsi M, Miklósi Á. Obeying social rules: a comparative
study on dogs and humans. J Comp Psychol. 2005;3(3):223–43.
11. Topál J, Miklósi Á, Csányi V, Dóka A. Attachment behavior in dogs (Canis
familiaris): a new application of Ainsworth’s (1969) Strange Situation Test.
J Comp Psychol. 1998;112(3):219.
12. Pongrácz P, Dobos P. What is a companion animal? An ethological
approach based on Tinbergen’s four questions. Critical review Appl Anim
Behav Sci. 2023;267:106055.
13. Range F, Virányi Z. Tracking the evolutionary origins of dog-human coop-
eration: the “Canine Cooperation Hypothesis.” Front Psychol. 2015;5:1582.
14. Range F, Marshall-Pescini S, Kratz C, Virányi Z. Wolves lead and dogs fol-
low, but they both cooperate with humans. Sci Rep. 2019;9(1):3796.
15. Marshall-Pescini S, Schwarz JFL, Kostelnik I, Virányi Z, Range F. Importance
of a species’ socioecology: wolves outperform dogs in a conspecific
cooperation task. Proc Natl Acad Sci USA. 2017;114(44):11793–8.
16. Topál J, Gácsi M, Miklósi Á, Virányi Z, Kubinyi E, Csányi V. Attachment to
humans: a comparative study on hand-reared wolves and differently
socialized dog puppies. Anim Behav. 2005;70(6):1367–75.
17. Gácsi M, Győri B, Virányi Z, Kubinyi E, Range F, Belényi B, et al. Explaining
dog wolf differences in utilizing human pointing gestures: selection
for synergistic shifts in the development of some social skills. Allen C,
szerkesztő. PLoS ONE. 2009;4(8):e6584.
18. Pongrácz P. Modeling evolutionary changes in information transfer:
effects of domestication on the vocal communication of dogs (Canis
familiaris). Eur Psychol. 2017;22:219.
19. Udell MAR, Dorey NR, Wynne CDL. What did domestication do to
dogs? A new account of dogs’ sensitivity to human actions. Biol Rev.
2010;85(2):327–45.
20. Udell MAR, Dorey NR, Wynne CDL. Wolves outperform dogs in following
human social cues. Anim Behav. 2008;76(6):1767–73.
21. Arhant C, Bubna-Littitz H, Bartels A, Futschik A, Troxler J. Behaviour of
smaller and larger dogs: effects of training methods, inconsistency of
owner behaviour and level of engagement in activities with the dog.
Appl Anim Behav Sci. 2010;123(3–4):131–42.
22. Carballo F, Cavalli CM, Gácsi M, Miklósi Á, Kubinyi E. Assistance and
therapy dogs are better problem solvers than both trained and untrained
family dogs. Front Vet Sci. 2020;7:164.
23. Scandurra A, Mongillo P, Marinelli L, Aria M, D’Aniello B. Conspecific obser-
vational learning by adult dogs in a training context. Appl Anim Behav
Sci. 2016;174:116–20.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Page 15 of 15
Lugosietal. BMC Biology (2024) 22:245
24. Pongrácz P, Bánhegyi P, Miklósi Á. When rank counts — dominant dogs
learn better from a human demonstrator in a two-action test. Behaviour.
2012;149(1):111–32.
25. Howell T, King T, Bennett P. Puppy parties and beyond: the role of
early age socialization practices on adult dog behavior. Vet Med Res.
2015;6:143–53.
26. Morrill K, Hekman J, Li X, McClure J, Logan B, Goodman L, et al. Ancestry-
inclusive dog genomics challenges popular breed stereotypes. Science.
2022;376(6592):eabk0639.
27. Bognár Z, Szabó D, Deés A, Kubinyi E. Shorter headed dogs, visually coop-
erative breeds, younger and playful dogs form eye contact faster with an
unfamiliar human. Sci Rep. 2021;11(1):9293.
28. Gácsi M, McGreevy P, Kara E, Miklósi Á. Effects of selection for cooperation
and attention in dogs. Behav Brain Funct. 2009;5(1):1–8.
29. Dobos P, Pongrácz P. Would you detour with me? Association between
functional breed selection and social learning in dogs sheds light on ele-
ments of dog–human cooperation. Animals. 2023;13(12):2001.
30. Herberlein MTE, Turner DC, Manser MB. Dogs’(Canis familiaris) attention
to human perception: influence of breed groups and life experiences. J
Comp Psychol. 2017;131(1):19.
31. Pogány Á, Torda O, Marinelli L, Lenkei R, Junó V, Pongrácz P. The behaviour
of overweight dogs shows similarity with personality traits of overweight
humans. R Soc Open Sci. 2018;5(6): 172398.
32. McGetrick J, Brucks D, Marshall-Pescini S, Range F. No evidence for a
relationship between breed cooperativeness and inequity aversion in
dogs. Mappes T, szerkesztő. PLoS ONE. 2020;15(6):e0233067.
33. Slabbert JM, Rasa OAE. Observational learning of an acquired maternal
behaviour pattern by working dog pups: an alternative training method?
Appl Anim Behav Sci. 1997;53(4):309–16.
34. Fugazza C, Moesta A, Pogány Á, Miklósi Á. Social learning from conspecif-
ics and humans in dog puppies. Sci Rep. 2018;8(1):9257.
35. Pongrácz P, Miklósi Á, Timár-Geng K, Csányi V. Verbal attention getting as
a key factor in social learning between dog (Canis familiaris) and human.
J Comp Psychol. 2004;118(4):375–83.
36. Pongrácz P, Vida V, Bánhegyi P, Miklósi Á. How does dominance rank
status affect individual and social learning performance in the dog (Canis
familiaris)? Anim Cogn. 2007;11(1):75–82.
37. Kabadayi C, Bobrowicz K, Osvath M. The detour paradigm in animal
cognition. Anim Cogn. 2018;21(1):21–35.
38. Topál J, Miklósi Á, Csányi V. Dog-human relationship affects problem solv-
ing behavior in the dog. Anthrozoös. 1997;10(4):214–24.
39. Pongrácz P, Miklósi Á, Kubinyi E, Gurobi K, Topál J, Csányi V. Social learning
in dogs: the effect of a human demonstrator on the performance of dogs
in a detour task. Anim Behav. 2001;62(6):1109–17.
40. Pongrácz P, Miklósi Á, Kubinyi E, Topál J, Csányi V. Interaction between
individual experience and social learning in dogs. Anim Behav.
2003;65(3):595–603.
41. McGreevy PD, Georgevsky D, Carrasco J, Valenzuela M, Duffy DL,
Serpell JA. Dog behavior co-varies with height, bodyweight and skull
shape. Munn AJ, szerkesztő. PLoS ONE. 2013;8(12):e80529.
42. Junttila S, Valros A, Mäki K, Väätäjä H, Reunanen E, Tiira K. Breed differ-
ences in social cognition, inhibitory control, and spatial problem-solving
ability in the domestic dog (Canis familiaris). Sci Rep. 2022;12(1):22529.
43. Dutrow EV, Serpell JA, Ostrander EA. Domestic dog lineages reveal
genetic drivers of behavioral diversification. Cell. 2022;185(25):4737-4755.
e18.
44. Hansen Wheat C, Fitzpatrick JL, Rogell B, Temrin H. Behavioural correla-
tions of the domestication syndrome are decoupled in modern dog
breeds. Nat Commun. 2019;10(1):2422.
45. Lehoczki F, Andics A, Kershenbaum A, Kubinyi E, Passilongo D, Root-
Gutteridge H, et al. Genetic distance from wolves affects family dogs’
reactions towards howls. Commun Biol. 2023;6(1):129.
46. Lenkei R, Carreiro C, Gácsi M, Pongrácz P. The relationship between
functional breed selection and attachment pattern in family dogs (Canis
familiaris). Appl Anim Behav Sci. 2021;235:105231.
47. Csepregi M, Gácsi M. Factors contributing to successful spontaneous
dog–human cooperation. Animals. 2023;13(14):2390.
48. Ujfalussy DJ, Bognár Z, Molnár M, Miklósi Á, Kubinyi E. The difference
between two brachycephalic and one mesocephalic dog breeds’
problem-solving performance suggests evidence for paedomorphism in
behaviour. Sci Rep. 2023;13(1):14284.
49. Parker HG, Dreger DL, Rimbault M, Davis BW, Mullen AB, Carpintero-
Ramirez G, et al. Genomic analyses reveal the influence of geographic
origin, migration, and hybridization on modern dog breed development.
Cell Rep. 2017;19(4):697–708.
50. Rao A, Bernasconi L, Lazzaroni M, Marshall-Pescini S, Range F. Differences
in persistence between dogs and wolves in an unsolvable task in the
absence of humans. PeerJ. 2018;6:e5944.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.
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