Animal Watch: Family Dog Project: New scientific insights into the evolutionary and ethological foundations of dog-human relationship

Family Dog Project

Eötvös Loránd University, Department of Ethology
Budapest, Hungary

Dog Behaviour Research

The science behind a friendship

The Family Dog Project was established in 1994 as the first research group dedicated to investigate the evolutionary and ethological foundations of dog-human relationship.

The project was initiated by Professor Emeritus Vilmos Csányi together with Antal Dóka, Ádám Miklósi, and József Topál at the Department of Ethology at the Eötvös Loránd University, Budapest, Hungary.

We hypothesised that dogs have evolved to survive in the anthropogenic environment, and our investigations aim at revealing the contribution of humans and dogs to this long-standing partnership. Thus we are not interested solely in the mental abilities of dogs but in all aspects of human and dog behaviour that have strengthened this bond, and may even expand it further. Surprisingly, in our experience this research does not only reveal important insights on dogs but also on us, people.

New scientific insights are fed into practical applications through our Dogs for Humans charity, which educates dogs for the disabled and dog assisted therapy.

The Family Dog Project

Why the dog?

A unique process: domestication for social competence

Due to their domestication the dog became one of the most successful mammals in the last 20-40.000 years of biological evolution. Compared to its living ancestor the wolf, dogs are now more wide-spread on the Earth and live in far greater number. This achievement can be very likely attributed to the fact that the dog has joined to live in the human niche which allowed him access to new resources of food and protection. However this change in the evolution of the dog could have not been achieved without changes in the behaviour that made it able to adapt to the human social environment. Sharing their environment dogs interact with the humans in many ways and living in such a complex social environment is cognitively challenging. It is widely accepted that the adaptational demands of the highly organized social life have led to special socio-cognitive abilities in dogs.

Many assume that studying dog-human communication offers a unique opportunity for our understanding the evolution of human communicational skills. This claim is based on the view that the dog can be regarded as unique among domesticates. In fact dogs were not only the firstly domesticated animal, but had from very early on a “special” relationship with humans. The transition from the wild state to the domesticated one changed the selective forces radically leading to the adaptive specialization of dogs to the human environment. It is increasingly assumed that many aspects of dog behaviour can be functionally analogue to the corresponding human trait. Since human environment is challenging for dogs by virtue of its complex social and cognitive nature, dogs had to develop human-compatible social behaviour traits including functional analogues of human communicational skills.

Wolf-dog comparisons

Dog-human attachment

Attachment between dog and owner is analogous to that of a human infant and his/her caregiver. The development of attachment is not restricted to a “sensitive period”.
More...

Social learning

Dogs are able to learn through observation both from humans and other dogs, also in cases when the goal of the activity is not evident.
More...

Social cognition

Dogs are very sensitive to human social cues which often mediate their learning about the environment.
More...

Physical cognition

Visual communication

Dogs are able to utilize a wide range of human communicative gestures.
More...

Acoustic communication

Various acustic parameters of barks correlate with assumed emotional content.
More...

Personality

At the behavioural level dogs share many aspects of human personality traits. Some of these are associated with highly polymorphic genes.
More...

Artificial companions

Ethology, in particular the study of dog behaviour, could provide important insights in the development of synthetic companions or embodied robots.
More...

Acoustic communication

Although it is one of the most conspicuous features of dog behavior, barking has received very little attention from animal behaviorists, ethologists or from an applied perspective. Emerging new research has indicated that in the repertoire of dog vocalizations barking has unique features by showing wide ranges of acoustic parameters like frequency, tonality and rhythmicity. According to the new experimental data, barking is strongly context dependent, and is informative at least for humans. At the same time, there are still only a few indications for intra-specific communication with barking in the dog.

We assume that dog barking emerged through selective processes in which human preference for certain acoustic aspects of the vocalization might have been of paramount importance. We call for a more experiment-oriented approach in the study of dog vocalization that could shed light on the possible communicative function of these acoustic signals.

Dogs, just like their wild relatives, have a rich vocal repertoire, including not only barking, but also other types of vocalizations. We have recently started to investigate the role of growls in dog-dog communication, especially in respect of the possible referential content of these signals.

Referentiality means that a signal contains information about not only the signaller’s inner state, but some part of the outer environment too. Growls are suitable subjects for this kind of research, because dogs emit them in various social contexts, both in agonistic and non-agonistic situations. The acoustic communication of animals can be also interesting from the aspect of researching the unique features of human language. In a new series of playback experiments we use artificially constructed sequences of dog barks, for testing the sensitivity of dogs to particular language-features, like recursivity.

Artificial companions

Ethology, in particular the study of dog behaviour, could provide important insights in the development of synthetic companions or embodied robots. Although Reto U. Schneider listed our work with AIBO robot among the maddest experiments in science, those observations have clearly established the limitation of that robotic pet when it comes to interactions with conspecifics (Kubinyi et al 2004). Interestingly, both dogs and AIBO can be taught by clicker training (Kaplan et al 2002), however, both adults and children seem to find the AIBO less interesting as they disrupt playing with this toy earlier in comparison to a dog puppy (Kerepesi et al 2005). Actually, building robotic pets that mimic real animals may not be the way forward.

We aim to develop a new interdisciplinary science, ethorobotics, that is interested in whether and how the control and dynamics of animal behaviour can be utilized to build better robots which however are not copycats of real animals. We hope that detailed modelling of the social behaviour of the dog could advance also the field of robotics. But we do not want to replace pet dogs by building robots, these two creatures should play different functions in our lives.

Dog-human attachment

Attachment is a very important concept for animals that live in closed groups since group activities depend on the actual relationships among the members. Some forms of social learning or cooperation take place only if there is a close relationship between two animals.

Psychologists have found that there is a special bond between the mother and her infant, and not only the quality of this relationship varies from one individual mother to the next but later children's performance in the school, for example cooperative willingness etc, seems to depend on the quality of this relationship.

The most striking feature of the social life of dogs is that they seem to prefer joining human groups. The dog-human relationship has a long evolutionary history and this could be based both on dogs’ evolutionary heritage, being the descendants of wolves, and on changes which took place during their adaptation to living with humans. However, the notion that the dog is just a tamed version of the wolf and the affiliative behaviour of dogs towards human is simply the manifestation of a wolf-like behaviour in an interspecific context is not really supported by current knowledge. Recent evidence suggests that domestication led to significant changes in the social-affiliative behaviour system of dogs and these changes served as the basis of the evolutionary development of dog-human relationship.

In order to test dog-human attachment we have utilized the experimental developed by Mary Ainsworth (Strange Situation Test, 1969) for studying behavioural criteria of human infant-parent bond. This experimental approach offers direct comparisons between the behavioural manifestation of attachment behaviour in dogs and human children. We have found that dogs (both puppies and adult ones) displayed a specific reaction towards their owners, but not towards a stranger, by looking for them in their absence and making rapid and enduring contact upon their return. They also preferred to play with the owner, and decreased play activity in the absence of the owner.

Follow-up work provided evidence that this pattern of attachment is stable over at least one year and is independent of the peculiarities of the testing location. An important further analogy to the human case has been revealed by observing the emergence of attachment behaviour in shelter dogs. These observations suggest that dogs that have been deprived of human contact (adult shelter dogs) are able and motivated to initiate a new relationship rapidly after a short duration of social contact with humans.

More recent results show, that in contrast to 4-month-old dog pups, grey wolf cubs of the same age did not fulfil the criteria for attachment to human. Despite being hand-raised and socialized to an extreme level (in contact with their owners 20-24h per day for the first 3-4 months of their life), these hand-reared grey wolf pups did not seem to discriminate between their caregiver and a stranger greeting them when left alone in an unfamiliar enclosure. It seems that unlike dogs, the human caregiver does not act as a ‘secure base’ for wolves in stressful situations. These observed differences between wolves and dogs show that the emergence of ‘infant-like’ attachment in dogs is not (only) due to social experience during early exposure to humans.

Attachment has also another interesting effect: The attached individuals seem to develop some kind of dependence toward the attachment figure. This means that in a problem situation their first strategy is to seek the help of the attachment figure before attending the problem itself. This finding is often interpreted (in popular literature) as the 'pet dogs' being 'stupid' but this is not true! In contrast, it means that dogs living in close relationship with their owner prefer to wait for the other to do the job for them, and only if this 'strategy' fails are they willing to solve the problem themselves. In other words dogs are very flexible in using social strategies.

In sum our findings on the attachment behaviour point to a characteristic selective responsiveness to the human caregiver (owner) in dogs, and this supports the view that attachment is a functionally distinct component of the social behaviour of the dog showing striking functional behavioural similarities to that of described in human infants.

Personality

Personality is often defined as an individual's distinctive pattern of behaviour (besides feeling and thinking) that is consistent across time and situations. Personality studies in dogs have become very popular in the last decade. Dog personality is a matter of great public concern, and besides theoretical interest, it has a wide range of practical applications, including significant influence on the dog-human bond.

Despite the increased interest, at present there is neither standard methodology nor standard terminology in dog personality studies, therefore we had to work out our own methodology. In the following we present our dog personality studies based on questionnaires and behaviour tests.

1. Dog and owner demographic characteristics and dog personality trait associations The aim of this study was to analyze the relationships between four personality traits of dogs (calmness, trainability, dog sociability and boldness) and dog and owner demographics on a large sample size with 14,004 individuals. German speaking dog owners filled in an online questionnaire in German which was advertised in the “Dogs” magazine (www.dogs-magazin.de). We found multiple associations between four traits (calmness/emotinal stability, trainability/openness, dog sociality and boldness) and demographic variables such as age, sex, neutered status or the gender of the owners.

2. Cross-Cultural Comparisons Cross-cultural comparisons of dog behavior are limited. We compared the questionnaire responses of German shepherd owners in Hungary and the United States (Wan, 2009, in press). We found for example, that American owners were more likely to keep their dogs indoors during the day and at night, to report that their dogs were kept as pets, and to engage their dogs in a greater number of training types (e.g. conformation training, agility training). Concerning the behavior, American owners reported higher scores than Hungarian owners on the confidence and aggressiveness scales of our survey (diffrent from the above one).

3. Personality trait and gene polymorphism associations At the behavioural level dogs share many aspects of human personality traits. Some of these are associated with highly polymorphic genes. Studies suggest that the dog could be a very valuable natural model for behaviour-gene associations in humans. In recent years molecular genetics has begun to identify certain neurotransmitter-associated genes, called candidate genes, for quantitative behavioural traits. According to this model, complex behavioural traits such as activity, impulsivity, and aggression are determined by various genes which interact additively or nonadditively. Recent research has revealed a number of allele polimorphism in candidate gene in dogs. Based on the human parallels intensive search for behavioural correlates has been started. Dopamine D4 receptor (DRD4) gene polymorphism was found to be significantly associated with the “activity” and "social impulsivity" trait in the dog in a population of German shepherds. Subsequent analysis in in vitro assays showed that alleles differ in functional properties.

This was revealed both by questionnaire-based evaluations and behavioural tests. Behavioural tests are independent largely from the perception biases of the observer, and provide the behavioural elements that are necessary for the implementation of dogs' behaviour in robots. We developed a test battery with 14 subtests, named Family Dog Test, for measuring individual variability in dogs. We found that the Liveliness trait obtaind by the test battery is associated with the tyrozine-hydroxilase intron 4 polymorphism in German shepherds.

4. Environmental effects on the personality of dog Laboratory dogs are frequently used in scientific studies. The contrast between the quality of life of laboratory and family dogs is salient. Laboratory dogs are kept in a highly restricted environment, in a limited area (4 nm2 for 1-2 dogs). Usually they do not have human contact except with their caretakers, once a day in feeding and cleaning time. We tested laboratory beagles and family begales in the Family Dog Test (see above), and compared their behaviour. We found that there was no difference in the Liveliness of the two populations, but Playfulness, Sociality, and Neuroticism significantly differed.

Social cognition

Some argue that dogs’ complex social skills have been selected for during their co-habitation with humans. Indeed, dogs may have to rely on sophisticated social abilities in order to get along in their relationships. However, instead of reporting a list of anecdotes researchers need to find well controlled methods to reveal the cognitive mechanisms that may operate in the dogs’ mind. We have not yet convincing evidence to say that dogs are able to read our minds but nevertheless they seem to be very skilful readers of our behaviour. In most cases this “trick” also does the job.
The attention of humans (the direction of looking – head orientation) has an important effect on the dogs’ behaviour. Dogs take human attention into account when they retrieve objects or respond to commands. In the latter, they can also figure out whether a command was intended to them or somebody else.

Dogs are also very good in displaying rituals and following simple rules. Not surprisingly they seem to get engaged easily a “hide & search” game, and will look for the object if actually nothing was hidden. Human behaviour often becomes a sample to which dogs match their behaviour. Thus sometimes a pointless short detour by the owner during the daily walk gets also included in the routine of the dog.

In particular situations dogs can take into account what the their owner know or does not know, however they observe some minute visible differences in the owners’ behaviour rather than being able to think about thoughts of the person. Our results show for example, that adult pet dogs, similarly to 3-year-old children, were able to complement information that was missing in the human. However whether dogs (and apes/children) used ‘mindreading’ or solved the task by some ‘insightful’ learning about observable behavioural cues of the human (behaviour-reading) is not clear. These results support the claim that dogs are able to provide humans with information that helps them in obtaining a goal; or, alternatively, dogs can rely on human behavioural cues indicating the lack of certain ‘knowledge’.

Physical cognition

In order to find food, catch a prey, defend a territory or make a burrow wolves and dogs need to know something about the physical lows of their environment. Often such skills are also referred to as ecological cognition because they may differ among species.

Dogs are able to trace and follow objects both visually and on the basis of smell. The former ability is often referred to in cognitive terms as “object permanence” because it is assumed that such tasks can only be accomplished if the dog is able to build a “mental image” of the object. Such skills could be useful both for the wolf and the dog when they hunt for rapidly moving prey on a complex terrain.
Many other experiments are aimed to find out about the navigation skills of dogs and wolves. There are many interesting anecdotes how wolves find their way around on their territory, how they optimise seemingly their way of approaching the prey, and make short cuts or detours depending on the obstacles they face. It is however more difficult to find the cognitive basis for such skills in the laboratory or at least under controlled condition.

Pet dogs living in the city seem not to be very professional, for example, in relatively simple detour tasks. They need about 5-6 trials to learn how to get a piece of food from behind a 3 m long fence. They seem to have also problems with transferring their knowledge about the detours to other similar situations and it is also not clear how early (“puppyhood”) or general experience may enhance the skills of dogs getting around.

Recent work has also shown that humans’ social and communicative influence also modifies the ability of dogs to process environmental cues. It seems that the selection for enhanced preference for humans and human actions may inhibit the dogs to react optimally to challenges in the environment if manipulated by humans.

Social learning

There are many cases and definitions for social learning – in general we consider an act social learning, when there is a demonstrator, who does something, while at least one observer witnesses this action, and later the observer will do something similar to the demonstrator’s action, while without demonstration this similarity would be less likely.

Dogs are excellent subjects for examining both interspecific and within-species social learning. As dogs form well-working social units with humans, it is easy to use human demonstrators in particular tasks, where dogs should learn by observation from the humans’ actions. At the same time trained dog demonstrators can be used also for tasks, where we expect dogs to learn from each others exemplar.

In several experiments we found that dogs learn easily from humans in many tasks, like detouring a V-shaped fence, opening a problem box, or operating a two-action device. Dogs follow human actions also in the ‘Do as I do’ paradigm, where they show ability for generalization.

We have demonstrated that dogs are able to use social information provided by humans in problem solving situations. Naive dogs learn a detour task faster if they can observe a demonstrator making the detour. Moreover they can use socially provided information to overcome previously learned but now maladaptive behavioural routines. The results show that other factors, like social rank can influence the dogs’ performance in social learning tasks: high ranked (dominant) dogs learned more easily from a human demonstrator, while subordinate dogs learned much more effective from a dog demonstrator.

Finally, it seems that dogs have special kind of receptivity to behaviour cues of human „teaching” which process ensures efficient transfer of information, even when its content is cognitively ’non-transparent’, arbitrary and actually does not have any perceivable adaptive value at all.

Visual communication

In humans the pointing gesture can take many forms in everyday life. Many independent studies have established that dogs comprehend the human pointing gesture and we showed that dogs are able to rely on more subtle human visual cues like head turning, nodding or bowing. Moreover they can generalize to a certain degree from familiar pointing gestures to unfamiliar ones, and thereby they can use also novel pointing gestures as a cue. In some cases dogs seem to regard the pointing gesture indeed as being a communicative act, as in the experimental setting they tended to choose the bowl pointed at by the human even when it was contradicted by direct olfactory or visual information.

We showed that the comprehension of the human pointing in dogs may require only very limited and rapid early learning to fully develop. In a more recent study we found that although dogs perform well in the case of many different pointing gestures they perform poorly in the case of the gestures in which from the observer’s point of view the pointing arm and hand stays within the silhouette of the body. This result made us to conclude that for the dogs the protrusion of a body part of the body torso provides the key feature of the signal. Examining this question more closely in a very recent study we found that making the gesture visually more conspicuous could have an enhancing effect in cases where the gesture does not stick out from the body torso. On the basis of these results it seems that the most informative sign for the dogs is not even the line of the pointing arm but a clearly visible patch, which appears conspicuously and asymmetrically at one side of the body torso. We have provided evidence that during the domestication process, selection for two factors under genetic influence (visual cooperation and focused attention) have led independently to increased comprehension of human communicational cues in dogs.

In a comparative study we revealed that 3-year-old children are able to rely on the direction of the index finger, and show the strongest ability to generalize to unfamiliar gestures. Although some capacity to generalize is also evident in younger children and dogs, especially the latter appear biased in the use of protruding body parts as directional signals. There are also some indications that dogs have a strong propensity to initialize communicative interactions with humans by using visual and sometimes also acoustic signals functionally similar to the ones used by humans. Several studies have demonstrated that also dogs “point” to humans, e.g. when dogs when facing an unsolvable situation use attention-getting behaviour. For example, after looking at the owner, dogs display gaze alternation between the location of the target object and the owner. A similar phenomenon was observed in a separate experiment, in which dogs, after having learnt how to solve a task, were prevented to get the target object the same way. Characteristically, after a few attempts most dogs stopped trying and looked at their owner. According to the results of these studies gaze alternation is a typical sign of dogs’ “pointing behaviour”, and it proved to be also very effective in the sense that humans can extract the information of it about the actual location of something that the dog wants to obtain.

Wolf-dog comparisons

Differences between hand-raised wolves and dogs indicate that social attraction, synchronizing behaviour and communicative abilities of dogs changed markedly during the process of domestication.

Raising wolf cubs and dog puppies in an identical way re¬vealed many specific social behavioural differences between the two species, especially with regard to their interactions with humans. Even at an early age (3-5 weeks), dogs dis¬played more communicative signals (e.g. vocalization, tail wagging, gazing at the human’s face) and were less aggres¬sive and avoidant than wolves, although the general activity level did not differ between the two species.

Due to human fostering, 5-week-old wolves showed a clear preference for their caregiver in a preference test, if the other stimulus was another human. However, in contrast to dogs, wolves’ preferences for the caregiver did not develop into a behavioural pattern that could be categorized as attachment. In contrast to hand-reared dogs and pet dogs, individually socialized, hand-reared wolves did not show highly different responsive¬ness to their caregiver compared to an unfamiliar human. While wolves did not display characteristic patterns of attachment toward their caregiver, their preference for her remained strong at the age of 1 or 2 years.

Our young hand–reared dogs, but not wolves, were able to use more difficult human pointing gestures (e.g., momentary distal pointing) spontaneously. Young wolves needed massive training to reach the same level of success that dogs reached instantly. The reason for this difference might be that in contrast to dogs it was very difficult to establish gaze-to-gaze contact with the wolves; therefore, wolves were less able to attend to an experimenter’s gestures for an extended duration. Dogs are inclined to look at our faces, and this inclination provides them with a broadened opportunity for learning about human gestures. However, socialized adult wolves can utilize human communicative signals. Thus the dog-wolf difference should be interpreted as a developmental change in the timing of some social behaviours rather than an overall difference in the ability.

Animal Watch

FAIR USE NOTICE

FAIR USE NOTICE

All Blogs licensed under Creative Commons Attribution 3.0

Thursday, July 4, 2013

Family Dog Project: New scientific insights into the evolutionary and ethological foundations of dog-human relationship

Family Dog Project

The science behind a friendship

The Family Dog Project

Why the dog?

Further reading

Wolf-dog comparisons

Dog-human attachment

Social learning

Social cognition

Visual communication

Personality

Acoustic communication

Further reading

Artificial companions

Further reading

Dog-human attachment

Further reading

Personality

Further reading

Social cognition

Further reading

Physical cognition

Further Reading

Social learning

Further reading

Visual communication

Further reading

Wolf-dog comparisons

Further reading

No comments:

Post a Comment