Executive Summary for General Audience


"Like me" addressed a crucial problem in cognitive science: what is the relationship between cognition and emotion? Two recent discoveries – that imitation and empathy are tightly linked, and that peoples’ tendency to inadvertently copy each other’s gestures facilitates social interaction – have fueled the discussion. The overall scope of the project was to understand the evolutionary, cognitive and neuronal bases of the human potential for imitation, empathy, and prosocial behaviour, and of the way these phenomena are related.

From an evolutionary point of view, a big question is how the neural networks for movement and emotion mirroring came about. It is very likely that some basic forms of empathy, the ability to understand and vicariously experience the emotional states of others, has been inherited from our ancestors. But so far the empirical evidence for this hypothesis was weak. We decided to test dogs, because earlier studies in our lab showed that this species has the capacity for imitation, and are yawning contagiously with other dogs or even humans.

Before directly addressing empathy and the link to imitation in dogs, we needed to find ways to objectively determine emotional states in dogs. In the first two years we have made substantial progress in developing novel, highly controlled experimental settings and physiological approaches for assessing dog emotions, including measuring of heart rate, tail-wagging laterality, changes of skin temperature, release of stress hormones, and changes of postures after exposure to emotionally relevant stimuli. The most significant methodological success was in our measuring of looking behavior of attentive and naturally behaving (e.g. standing and tail wagging) dogs when confronted with emotional visual stimuli. By applying human eye-tracking methodology we could show that pet and lab differ in how they look at human faces, indicating an influence of the amount of exposure to humans. However, both showed a strong left gaze bias, like humans, and they looked slightly different to faces showing different emotional expressions (happy, sad, angry, neutral).

The indication of heterospecific emotion discrimination is especially interesting, as dogs are highly domesticated animals. Another experiment using the touch-screen methodology achieved a major break-through in this respect. Using an innovative experimental design (showing only half faces in training) we provided the first solid evidence for the ability of a non-human animal to discriminate emotions of another species. With this evidence we then sought to investigate empathic-like responses in dogs. In two experiments we examined the dog's responding to emotional stimuli from either conspecifics or heterospecifics (humans). The subjects reacted in the expected ways (increase of cortisol levels, comfort-seeking behavior toward their familiar partners) to vocalizations of conspecifics (whines) recorded during a distressful event. These findings are the first to suggest that dogs can experience and demonstrate “empathic-like” responses to conspecifics’ distress-calls. However, first studies on the relationship of empathy and imitation, pro-social priming of contagious behaviour and changes of affiliative behaviour towards an imitator (the so-called Chameleon effect), did not reveal such bi-directional effects in dogs.

To study the mechanisms underlying the Chameleon effect in humans in a laboratory setting, we developed and validated several new experimental paradigms. We used this approach in two work packages, which yielded several remarkable and highly novel insights. In a series of experiments we assessed the effects of imitation on empathy and prosocial behavior, thereby systematically disentangling several potential mechanisms. We found that imitation has several socio-affective effects: it induces selective conformity with the imitator, a preference in prosocial behaviour for the imitator, and increases empathic arousal when the imitator is in pain. Crucially, however, imitation only has these effects when the imitated movements are spatially congruent (“go in the same direction”), but not when they are merely temporally contingent (“happen at the same time”) with one’s own movements. This suggests that that the Chameleon effect is rooted in actual motor processes, not in generalized positive affect induced by the appreciation of perceptual predictabilities.

We furthermore found that the prosocial effects of imitation are predicted by increases in affective resonance (autonomic arousal), suggesting that they are rooted in a process of “bodily tuning in” with the imitator. Lastly, we showed that the effects of imitation are selectively targeted towards a specific interaction partner (the imitator). This challenges the generally held view that being imitated fosters a general prosocial attitude. Instead, it suggests that it may also trigger or reinforce tendencies towards selective conformity, parochial empathy and ingroup favoritism in prosocial behavior. We also carried out an fMRI study to study more in-depth how representing other’s bodily, cognitive and affective internal states ultimately drive enhancements in empathy and prosocial behaviour. Our findings indicate that these effects depend on the core ability to overcome one’s own perspective, rather than the ability to put oneself in others’ shoes, suggesting that actively disengaging from self-representations might be more crucial to prosocial behaviours than previously thought.

In a second line of studies in humans, we examined how socio-affective processes affect the tendency to imitate, again using a newly developed task. In a series of behavioural, fMRI and EEG experiments, we showed that the tendency to automatically imitate others can be modified by task-irrelevant social stimuli that signal different levels of affiliative intent. Interestingly, our behavioural findings showed that mimicry is not only boosted by smooth and save social interactions. Interacting with persons that are seen as threatening can also enhance automatic imitation; here, mimicry seems to serve a need to appease. An fMRI study has substantiated and extended these findings by showing that the regulation of mimicry in response to happy and to out-group faces is mediated by partly distinct neuro-cognitive processes. Lastly, we performed two EEG studies, being the first to investigate the temporal neural dynamics of automatic imitation and its modulation by ethnically diverse stimuli. We showed several novel ERP components linked to action execution in automatic imitation paradigms, with a parietal ERP component predicting the extent of automatic imitation.

Taken together, the human laboratory studies have resulted in several major methodological and theoretical contributions, and have considerably increased our understanding of the mechanisms underlying the Chameleon effect in humans. In order to bring the two strains of the project, the relationship between empathy and imitation in dogs and in humans, closer together, we launched a human-canine fMRI initiative. From investigating the brain activities of (awake) dogs and humans in a fully comparative manner we expect quite soon pioneering insights into how the brain of another mammal enables its advanced socio-cognitive skills, and whether these are related to similar socio-cognitive functions in both species.