Once, when my son and I were walking our jolly puppy, we found ourselves behind another dog walker whose terrier was seized by aggressive fury. It was snarling and snapping at us.
The terrier's owner turned and half snarled herself: ‘Will you keep your dog under control!’ (Our puppy, incidentally, was walking along with us quite happily). As we overtook the unhappy duo — terrier and owner — I just couldn't resist the opportunity of saying to my son slightly louder than needed: ‘Well, Luke, you know what they say. Like owner like dog’.
Who better than a veterinary nurse to know just how much like an owner a pet can be? Owners and pets resembling each other physically is an old chestnut and can be easily explained — people, somewhat narcissistically, often choose pets that look like them. But what about behaviour? Do suspicious, withdrawn people end up with nervy, unsociable dogs while incurably sociable optimists with invasively friendly dogs? And if so how on earth does this come to be?
That dogs do indeed learn behaviour from their owners was recently confirmed by Cecilia Heyes from Oxford University, England, and Friederike Range and Ludwig Huber of the University of Vienna, Austria. Dogs were trained to open a sliding door to get a treat using the same method as their owner, by pressing a lever with either their heads or their hands. Hands translated to paws and heads were heads. So far so good. A second group of dogs were trained to do the opposite of their owners' actions — these dogs were rewarded for using their paws if their owner used their heads, or vice versa. But this group of dogs took almost three times as long to train than the dogs who simply had to copy their owners, suggesting that they were having to overcome an automatic imitation response to learn how to get the treat.
At first glance these results seem to state the obvious — that dogs copy their owners. What the researchers teased out here was the basic drive in dogs to imitate. In a second round of experiments they trained both sets of dogs to do a similar task for which they were only rewarded if they did copy their owners. The dogs that had previously learned to do the opposite of their owners' actions were much slower to learn this second task. Here the authors argue that this shows that imitation responses are learned and not in the genes — it takes the dogs longer to learn the second task if they have to reverse the automatic response to do the opposite which they learned earlier.
It is important for the development of knowledge to have what we already knew confirmed by experimentation. Of course they go further in their experiment by teasing out whether this learning is nature or nurture, are the dogs born to copy exactly or can they learn what to learn? But the really exciting thing about this research is the implications for how imitation works. It suggests that dogs have neural pathways in place that are similar to those thought to underlie some pretty sophisticated human and primate behaviour such as the cultural transmission of behaviours, social interactions and empathy. One of the key players in these neural systems are thought to be mirror neurons.
Ever since mirror neurons were discovered in primates in 1996 by Giacomo Rizzolati and his colleagues, the study of socially learned behaviours has changed gear. This singular class of neurons are active when someone moves or senses another move in a similar way. I use on purpose that rather mysterious phrase ‘sensing another move’. Take opening a pack of biscuits; one doesn't actually have to see and hear the whole movement to know it has happened or for the mirror neurons involved in that action to fire. In these cells at least there appears to be no distinction between seeing/hearing/smelling/feeling something done and doing it yourself. These intriguing nerve cells appear to simulate the actions of others inside our brains. They are the link between sensing and doing.
Scientists and philosophers have got very excited by these cells as they also suggest a neural basis of empathy, thinking about others' states of minds. And guess what? They were discovered in primates and have since been found in humans. The type of imitative learning that Ceclia Heyes and her colleagues found in dogs — in which observing an act automatically activates a corresponding motor programme — fits very nicely into the type of neural processing suggested by mirror neurons. Imitative learning has also now been seen in budgerigars, dolphins, marmosets and parrots. And neurons with properties like mirror neurons have been found in songbirds. And so, with these bodies of research coming together the old barrier between humans and animals has taken another battering. Perhaps it's little wonder dogs end up behaving like their owners after all.