The feeling that our body “belongs to us” and that we control its movements is so ubiquitous that most people take it for granted. Psychological and neuroscientific research has shown that this sense of bodily self is fundamentally based on processes in the brain. For example, we know that certain areas of the brain integrate information about the body and its movements from different senses, such as vision and touch, into a “body model”. The brain uses this body model to distinguish between our own body and other bodies, and to plan, execute and control actions. When these brain areas are damaged, for example as a result of severe strokes, the body model can be profoundly disrupted. Suddenly, the hand may no longer be experienced as “mine” – or its movements may seem to be controlled by someone else.

Body models in the brain are therefore fundamental to our sense of bodily self. On the one hand, they need to be stable; on the other hand, they need to be flexible and thus capable of being adapted through learning. After all, babies only a few months old have to learn a body model in the first place! But even in adulthood, we can adapt our body model to physical development and ageing, or to the use of tools or prostheses. New technological developments in the field of human-machine interaction also show that our body models can be “relearned”, for example by embodying and controlling virtual or robotic “alternative bodies”. Many of these scenarios raise the question of where the limits of our body models’ flexibility lie, and to what extent our “own” body can even be understood as a clearly distinguishable unit.

To investigate these questions, experimental work with the healthy brain is needed. The interactive exhibit on display offers a simplified version of an experiment we developed in our laboratory to investigate the “relearning” of body models. We have already conducted similar experiments several times in a brain scanner. The additional 3D-printed brains on display allow us to show which brain areas are involved in the different learning processes of the body model.

In the exhibit, a virtual hand model is controlled using a data glove, allowing you to track a target on the screen with simple grasping movements. The green markings on the 3D-printed brain show how motor control centres and the visual cortex are involved in controlling and visually processing the movement.

The problem is that the virtual hand does not always mirror the movements of the real hand in real time, but sometimes with a delay generated by the computer program. This delay “decouples” the two most important types of sensory information the brain receives about the body: the hand movement we see and the one we feel – that is, the movements of the virtual and the real hand – no longer correspond. The red markings on the 3D-printed brain show brain areas that are particularly sensitive to these delays. These areas signal that the movement we see does not match what the body model would normally predict – an important error signal for distinguishing between self-generated and externally generated body movements.

Due to the delay, tracking the target is initially not so straightforward. Nevertheless, it is possible to adjust the movement of the real hand so that the virtual hand can follow the target. In the process, the brain gradually updates its body model. Essentially, the brain learns a new relationship between the commands it sends to the muscles and the resulting hand movement. These learning processes take place in higher-order brain areas, primarily in the cerebellum, as shown by the yellow markings on the 3D-printed brain.
Parts of the visual cortex are also highly active during the task, reflecting a particular weighting of the hand movement we see. Fascinatingly, the opposite occurs with the felt hand movement, as shown by the blue markings on the 3D-printed brain: areas of the brain that process the felt hand movement are temporarily downregulated. This means that, when updating its body model, the brain temporarily gives greater weight to the hand position it sees than to the felt hand position.

In summary, this simple experiment demonstrates how quickly the brain can relearn its body model, in part through the temporary, targeted reweighting of different types of sensory information about our own body position.

Text by Jakub Limanowski
Jakub Limanowski is a psychologist and neuroscientist.