Current Research

The Hamilton lab aims to uncover the cognitive and brain mechanisms which allow us to make sense of other people's everyday actions and respond appropriately. We also test if these systems are compromised in people with autism.

Action understanding is important because it is a fundamental building block for human social interaction. A better scientific theory of how we understand and respond to other people will have benefits for many fields, including

• helping children who struggle with social interaction (e.g. children with autism)
• building better social robots to interact with people
• enhancing learning by imitation

Topics we are currently studying include -

Representing actions and goals hierarchically

When you see someone tie a shoelace or turn on a light, you understand the sequence of movements and are able to link them to your own experience. We are studying how this action knowledge is organised in the brain. In particular, we are testing a hierarchical model of action knowledge, as shown in the diagram. This means that simple hand movements can contribute to many different goals, and a single goal might draw on a sequence of hand movements. This makes intuitive sense, so we are trying to find evidence for this hierarchy in the brain. Are different levels of the hierarchy localised to different brain regions? Or are different types of action localised to different brain regions?

Further reading: Hamilton & Grafton 2007

Actors and their actions

If you see someone fire a gun, your reaction depends on whether the person with the gun is a friend or enemy. Previous studies of action observation have normally looked at the action in isolation, without considering the identity of the actor. Neuroscientists also know that different brain systems respond to seeing faces compared to seeing hand actions. We are now looking at action understanding in a broader context, including different types of actors, both human and non-human. We want to know how information about different actors is integrated with information about actions.

Further reading: Ramsey & Hamilton 2010

Social signals for mimicry

Mimicry is the tendency to unconsciously copy other people's actions. People mimic more in some situations than others, for example, people tend to mimic more when they are with someone they like. We are studying the cognitive and brain systems when control when and why you mimic. We have found that direct eye gaze rapidly enhances mimicry - you are faster to mimic the hand action of the girl on the right when she is looking at you. Brain imaging shows this effect is implemented in medial prefrontal cortex

Further reading: Wang et al 2011

Goal understanding in autism

Autism is a neurodevelopmental condition affecting social and communication skills. There is some debate about whether autism affects the ability to understand other people's simple, non verbal actions, and how this might relate to difficulties in other social domains. We are currently using behavioural and neuroimaging methods to explore these ideas.

Recently, we have shown that when adults with autistic spectrum disorders watch movies of simple actions, their brains show just the same pattern of activation in parietal cortex as typical adults. This suggests that basic action understanding mechanisms are intact in autism. We are using further eyetracking methods to test how typical and autistic people understanding irrational actions. We also plan to examine mimicry in autism and test if is is modulated by social signals in the same way as it is in typical individuals.

Further reading: Marsh & Hamilton 2011

Perspective taking

Understanding other people and their movements in space involves elements of perspective taking. You must be able to link together actions from different viewpoints, and consider what another person can or cannot see, in order to predict his actions. We have used neuroimaging and behavioural methods to test if adults and children with autism can understand body postures from different viewpoints.

Data sharing

In accordance with funding guidelines, published anonymised data collected in the Hamilton lab is available to other researchers for secondary analysis. Available datasets are described in our publications. To access data, please email me with details of the dataset you want and why. Access to some data (e.g. video) may be limited by issues of partipant confidentiality.