Biology Of Dog Emotion, Cognition, Evolution & The Domestication Hypothesis


“Emotions are among our tools for adapting to the world-anticipating it, coloring it, so that objects and events are memorable in a meaningful way, for future use.”

-Dr. Huda Akil

“A cognitive approach is about celebrating different kinds of intelligence. Genius means that someone can be gifted with one type of cognition while being average or below average in another.”

-Dr Brian Hare

Cognition & Intelligence

Hare uses Howard Gardner’s Theory of Multiple Intelligences, (1983) to back up his exploration of the specific types of intelligence displayed in canines.

He looks at:

  • Empathy
  • Communication
  • Cunning
  • Memory
  • Reasoning

Everyone can test their dog’s ability in these areas (using his unique citizen science tool) by going to Dognition or, he has written a book called “The Genius Of Dogs” which covers all of these topics in great detail.

The more unidimensional concepts of intelligence, do not give us a clear picture. They do not factor in, the unique ways in which we can compare intra- & interspecies mental skills and how animals make inferences. Cognition tests are more robust than traditional IQ type testing. This is because they add more dimensions to the concept of intelligence, rather than the narrow measures used by the standardized (old style) tests.

There is a lot of debate as to how dogs evolved. It seems likely that they are distant descendants from wolves and via natural self-selection became part of a mutual domestication (convergent) process. The result was a change in allele frequency over time (evolution). All behaviour has its roots in the ancestry of the descendants, the function of some problem behaviours (particularly) among domestic dogs, stem from these behaviour traits. Around 50,000 years ago the human-dog relationship started to form. It has now evolved to be what it is today. It is quite perplexing as to how wolves stayed as wolves and some came to be dogs?

Morgan’s Canon or Occam’s razor (1903) states that we cannot exclude the interpretation of a particular activity in terms of the higher processes. Especially, if we have independent evidence of the occurrence of these higher order processes in the animal that is under observation. So, when we look at cognition, we are looking at the internal processes. Through physical experiments we can answer questions about how the mind works. When behaviours appear similar or different, it can often be the result of very different cognition.  We often ask simple questions about how animals solve problems and then we rule them out with experiments. Then we to use further experiments to test these and rule out parsimonious explanations. This is how we come to understand how animals process the complexities. Giving us indications of how sophisticated and unique their mechanisms are.

Dognition is like an online laboratory which you can use to test (your own) prediction of dog’s behaviour. You can then compare it with the actual evidence based on the tasks set. You will be able to ask questions about the way your dog scores on an array of tasks. There are some free games for you to try. You get a full report on the 5 types of intelligence that your dog displays. This means that everyone can be part of citizen science. The information is owned by Duke University and contributes towards data which will help us further understand our dogs.

Ecological Approach To Cognition

Evolutionary theory and cognitive theory combine together to give us the ecological approach to cognition. Tinbergen states that you need 4 levels of analysis.  The reason being, if you look at an organism such as the Lyrebird, you would say that they were amazing song artists? This ability for them, is totally unique. They can copy pretty much every sound (such as car alarms, mobile phones etc). What is the reason for this behaviour? As it turns out, they do not have any further comprehension of these sounds other, than just being able to copy them. This copying, is a trait that male Lyrebirds, use to attract females and the more diverse his portfolio of sounds, the more preferred he is among the female population; therefore increasing his mating success.

The Tinbergen 4 levels state that all levels of behavioural analysis serve to inform one another. All traits can be explained at all levels and all levels need to understand the evolution of that trait.

Tinbergen’s 4 Levels Of Analysis:

  • Phylogenetic
  • Ontogenetic
  • Functional
  • Proximate

Using this ecological approach to cognition, we can approximate that cognitive changes evolve as a consequence of unpredictability that cannot be solved by inflexible and innate strategies. No matter how complex. We need to view each species in his own context, look at the way these behaviours have helped the organism to solve problems and then how he has evolved to solve them. The function of this, can help guide us, with questions. We can then go on to explore the mechanisms of the behaviour. This approach, opposes the view that that cognition is a single one dimensional trait, which is a measure of the overall “intelligence” or “ability” of the individuals.

Theory Of Mind

The reverse engineering approach to problem solving, looks at functions of behavioural mechanisms. We look at what the problem was and how the animal evolved to solve it.  Problems include: food foraging and attracting mates.

Humans have theory of mind. They think about the thoughts of others. They understand that others have perceptions, intentions, emotions & knowledge. They also understand that these thoughts can be different from their own. Behne, Carpenter & Tomasello (05 & 06), showed that children as early as 14 months can become good at reading social cues with gaze and pointing. This is also demonstrated in other primates but to a lesser extent.

Darwin recognised that human cognition was a challenge for evolutionary theory. The ecological approach helps us to answer the question about how social problems were solved. This itself, drove the evolution of cognition, particularly in primates both human and non-human.  We look at theory of mind to accurately assess perceptions, intentions and knowledge of others. We have seen that humans emerge with this skill very early, are able to understand social cues and the intentions of others. This ability is executed through body language and expression. This ability is much more advanced in humans than other primates. Could this ability be explained in the learning potential during critical periods and Do non-human primates have a shorter critical period like wolves do, compared to our domestic dogs?

Do Dogs Understand That We Communicate To Show Intentions?

Hare, Call & Tomasello (98) conducted a series of experiments looking at how dogs read human social cues. They looked at pointing and gazing gestures. They found that domestic dogs were incredibly adept at reading these social cues. This was more so with heterospecifics (humans) than conspecifics (dogs). They were even able to rule out the obvious variable of olfaction. It was deduced, they were not just responding to motion created by gesture. It was actually found, that dogs were able to read the intentions of humans. Dogs are incredibly good at this when compared with other non-human primates, even better in fact.

Exposure & Ancestry Hypothesis

Is this ability (to understand human gestures, body language, intentions, emotions, facial expressions) developed through dogs exposure to humans? This was disproved by looking at domesticated dogs that had no exposure. So then, they looked at the “ancestry hypothesis”, is it inherited from their ancestors (wolves)? Studies show that wolves do not have the ability to understand human intentions through body language gestures. The ancestry hypothesis was rejected. The domestication hypothesis answered the questions but was difficult to test.

Domestication Hypothesis

Foxes- Trut 2003:

Wonderful experiment by Trut has been ongoing for the last 45 years. He has been looking at the domestication of foxes. There are two populations of foxes being bred in a longitudinal study. The results are remarkable and so interesting. This study gives us an overview of the changes that occur during domestication. Foxes are being bred in two groups, those selected for least fear and aggression towards humans and the control group.  Through just selecting on this criteria alone, these are the results:

Experimental Group Of Foxes- Notable Changes:

  • Floppy ears
  • curly tails
  • piebald coats (white and other colours)
  • feminised cranium (directly linked to testosterone levels)
  • Skeletal Gracilization
  • Serotonin levels higher
  • Corticosteroid levels lower (this could mean longer critical periods? Weiss, Kohler, Levine, Lundberg, Panksepp & Goldman)
  • Human  Approach (increase)
  • Tail Wagging (increase)
  • Barking/Crying Vocalisations (increase)

The fascinating thing about this study is that the foxes have exhibited a group of traits which are associated with domestication and no other criteria was selected for.  It was shown that the experimental group of foxes had similar skills to our domestic dogs and they were more skilled than the control group of foxes at reading human cues. Hare, Wobbler & Tomasello (09) supported this hypothesis with the singing dogs of New Guinea.

Overall, selecting on the basis of just friendliness, showed a lot of accidental changes. These came along with this selection process. The experimental foxes used human gestures more than control foxes and selecting on this criteria alone, changed a number of other factors. Those that relate to the physical and behavioural changes particularly reflect their growing relationship with humans.

Artificial Or Self Domestication?

  • Evidence shows that the friendly wolves likely self-selected due to friendliness
  • They could have been at an advantage due to humans being more sedentary around 12-15000 years ago
  • As we have seen with the foxes, this self-selection changes many aspects of behaviour, morphology and cognition
  • Other species may be self-domesticating today

Convergent Evolution:

This is where organisms who are distantly related are shaped by similar ecologies. Such as:

  • Thylacine (Tasmanian tiger)
  • Herbivorous Marsupials
  • Herbivorous Placentals
  • Lupus: North American Wolf

More interestingly, the remarkable differences seen in the Bonobo Monkey compared with the Chimpanzee have similarities that can be compared between wolves & dogs (the Bonobos being more friendly and domesticated than Chimpanzees). Most striking, is the Bonobo’s friendliness and pro-social behaviour towards humans and con-specifics. This does not seem to matter if they are either familiar or strange to them.

Bonobo’s Self-Domestication & Convergent Evolution:

  • Bonobo’s and chimpanzees developed at the same time (convergent) to solve a similar problem independently of each other
  • They are our closest living relatives along with Chimpanzees
  • They never kills (Chimpanzees do)
  • Less dominance hierarchical observable differences (not selected as not necessary to have dominant male for protection, survival)
  • They converged with other domesticated animals in the same physical, cognitive and behavioural way
  • evidence points toward self-domestication

Social Domestication, Evolution & Humans:

Socialisation development in humans came about due to them developing tolerance levels. These were required for living in proximity and being able to cooperate with one another. This tolerance would not have come about without evolution of cooperative traits. It could be that our early ancestors could have selected mates based on more tolerant personality traits. As a result the morphological changes such as those seen in the skull (ie brow prominence decreasing a direct effect of lower testosterone levels) and body shapes have changed dramatically through evolution. This evidence suggests that we may be self-domesticated.

Dog Cognition Studies

Questions that have been asked:

  • Do dogs imitate
  • Are they capable of intentional deception
  • Do dogs know what you can and cannot see
  • Do dogs understand symbols
  • How do dogs navigate
  • Do dogs take short cuts
  • Do dogs understand causal properties of the world
  • Do dog breeds differ

Human children have the ability to fast map. This allows them to learn thousands of words. Julia Kaminski, Josep Call & Julia Fischer looked at word learning in domestic dogs. They tested for the ability of dogs to recognize labels for 200 toys, learn novel labels in the presence of previous toys (ones he knew the names of, so he had to infer that a novel word, means a new toy). They then tested to see if he could remember these words a few weeks later and then this was repeated. They found Rico was able to fulfil all of these tasks. Was this unique to Rico? No, they found 3 other dogs capable of doing the same.

Do dogs know what we can and cannot see?

Kaminski (09) showed that dogs know what we can see. There was a transparent screen and an opaque one (which the human could not see through) it was found that the dog was able to infer what we could see.

Slide From Duke Cognition Course Coursera



Research looking at whether dogs could understand the concept of gravity, by putting food into tubes (apes performed at these tasks with learning) and attaching food to string. The food was tied to the string which was contained in a box. The food was at an angle to the string. The dogs could not understand how the string and food were connected.

When dogs look in the mirror, they do not understand that the reflection is them. Results show, that dogs were not very good at solving these kinds of problems on their own. These types of task were not obvious for the dog to solve.  Also, dogs do not make tools to use for tasks and (string food task where the string was connected to the food at an angle) they did not automatically understand connectivity. Dogs do understand their own reflection, so, in general, they do not solve problems on their own.

Dogs & Wolves Compared

They have very different social systems to one another. Wolves (apparently) have a more structured hierarchy of breeding and cooperation to raise their young. According to Hare, dogs have less structured hierarchy. Dogs living in a domestic setting are not part of a pack and the very nature of a domestic setting is “anti-pack”. The organisms within this group,  are not required to behave and acquire resources using their ancestral methodologies. They do not have the same risk of threat to their survival and so have adapted different behaviour strategies for this.

From a behaviour modification perspective, when things go wrong in the domestic setting, it is useful to have an understanding of where the function of that behaviour comes from. It does not come from dogs being alpha. Dogs are not alpha, they do not need to show dominance. An animal displaying characteristics which carry the label dominance are usually rooted in the ancestral need for that organism to adopt a suite of behaviours which improves his ability to survive (or his perceived threat to his survival). This is not required in a domestic setting (although instinctual drift has been mentioned in a few texts). More importantly, a dog that is displaying dominance is more likely to have the roots of that behaviour in fear. Using dominance techniques on fear (as some humans do) are not going to do anything to solve that problem, this article might be useful here:The Science Behind Punishment & Why It Does Not Teach Anything. (please see: Dogs Do Not View Us As Part Of Their Pack) Dogs display the same hormone response to situations as we do and they understand much of our communication. A lot of what is termed “pack behaviour” in a house can often be attributed to emotional contagion (a basic form of empathy). To find out about the mechanisms of this from a neurobiological and functional perspective, please see our week long course, which carries 9 CEU’s: Emotional Contagion & Empathy, The Neural Mechanisms & Evidence In Dog Behaviour.

Wolves are intolerant of strangers (termed xenophobics), often have intergroup aggression and do not allow genito-anal sniffing. Conversely, dogs are usually tolerant of strangers, feral dogs interestingly enough, do not show lethal aggression. They also, do not form packs or social groups in the ways that wolves do. Dogs allow strangers to sniff their genito-anal region.

Wolves are cooperative breeders. Pairs that mate will work together and raise the offspring. offspring will often remain in the pack to help raise the younger ones. This is the opposite in dogs. Dogs are promiscuous, non-cooperative, have no helpers and do not form packs. The exception to this is Dingos. Wolves have a higher survivorship compared to feral dogs. Female wolves have been shown to suppress reproduction in other wolves in the social group and male wolves only show aggression around mating time. Dogs do not display any of these behaviours and do not suppress reproduction.

When dogs look at us, the relationship is maintained through a mutual sender-receiver neurochemical function. This is kind of a feedback loop between both parties. The attachment is maintained by the secretion of psychobiological substrates relating to the somatosensory system (touch, pain etc and the other senses). Please see: Somatosensory Development, The Psychobiological Substrates Of Attachment & Learning In The Critical Period. Wolves have a shorter critical periods than domestic & feral dogs. Dogs essentially remain like young wolves their entire lives.

Breed Differences In Dogs:

Brian Hare’s Lab is collecting essential information relating to our domestic dogs. Some of the things he has found relating to our domestic dogs are:

  • Dogs bred from a breeder show more response to gesture and pointing signals compared to those with unknown origin.
  • Dogs who come from mixed backgrounds are more likely to rely on memory for completing and learning task
  • There is a size difference in dogs showing empathy favoured towards larger dogs

Dog Aggression:

Pitbulls have been given a bad rap due to the bite force of 1,800 pounds per inch. This was from a paper in 1984 by Baack which apparently mentions nothing of this breed or, their bite force.  Dogs cannot be identified as aggressive by their breed. It is often very difficult to tell what breed you are looking at, even experts can struggle with this.

Aggressive dog research lists Chows, Collies, Cocker Spaniels, Daschunds & Mixed breeds among the highest to be aggressive. When in actual fact, the most likely factor involved in a dog bite (in the US, according to Hare) is a Male Human Under the age of 10 years.  Overal, there is little scientific data to support evidence that dog breeds are aggressive. As aggression can be selected for and bred quite quickly, breed specific legislation (BSL) is doomed to fail.  Something does need to be done about this and BSL is not one of them.

Dog Cognition & Its Considerations

Call et al 2003 conducted an experiment on teaching a dog to leave its bowl. It was found that whilst the human had his attention on the dog he was happy to obey. If the human looked at a book, stepped away or turned their chair to face away, the dog would go to his bowl. In actual fact, you can train you dog to leave things if you understand how to effectively put a behaviour under stimulus control with the right value and rate of reinforcement.  A study was conducted on attractiveness and having a dog, it was found that having a dog greatly increases your chances of picking up a human mate too! This is a surprising benefit to owning a dog.  It is not all great though, unfortunately, in  some cultures, dogs are viewed as food or pests. This has lead to some pretty horrific treatment of them, probably because they were bred to be eaten. This attitude seems to be changing in developing economies.

Want to read more of our stuff?


Courses & Seminars:

Our Employers: Marmalade & Tango

positive reinforcement training in dogs

 Please Click On: Authors Biography, to find out more.


To sign up to our newsletter through Facebook click here: Simply Behaviours Facebook Page