Why can’t we tickle ourselves? Why do we perceive our own steps differently than those of a stranger walking behind us? An interdisciplinary research team from the fields of psychology, psychiatry and neuroscience has developed a mathematical model that explains why our perception is attenuated when we generate movements ourselves. This so-called sensory attenuation (SA) plays a crucial role in our understanding of self-perception and external perception. ‘In the long term, the findings could help to better understand mental disorders such as schizophrenia,’ reports the lead author of the study, Dr Anna-Lena Eckert. The study appears in the current issue of the journal ‘PLOS ONE’.

Insight into sensory attenuation

The mathematical model is based on so-called Bayesian causal inference and graph theory and shows that the brain is constantly weighing up whether sensory information comes from an internal or external source. In other words: am I hearing myself or something else? Is the touch coming from myself or from my boyfriend or a stranger? The brain is constantly trying to figure out the causes of sensory information.

If the stimulus is the predictable result of our own movement, it is categorised as ‘internal’ and downregulated in further processing. ‘This explains why we can’t tickle ourselves or why we pay more attention to the steps of a stranger than to our own on a dark street,’ says psychologist Eckert.

Experimental confirmation through two studies

The model was tested using two independent experimental data sets collected by the research groups of Prof. Katja Fiehler and Elena Führer (University of Gießen) and Prof. Benjamin Straube and Christina Schmitter (University of Marburg).

In a first experiment on tactile perception, subjects stroked their finger over corrugated 3D-printed objects. Just before the subject’s finger reached the grooved surface, a small vibration was triggered on the finger. This showed that the perception of the vibration was influenced by predictions from the brain – ‘a strong indication of sensory attenuation,’ comments Eckert.

The second experiment focused on visual delay: Participants saw their hand movements on a screen, whereby in some cases they actively performed the hand movement themselves, and in other cases their hand was passively moved by a lever. In some cases, a time delay was inserted between the movement and the video of the movement. The perception of this delay was lower when the movement was actively performed.

The data sets from both experiments were compared with simulations and model optimisations from the working group led by theoretical psychologist Anna-Lena Eckert. The calculations were carried out on the Marburg high-performance computing cluster MaRC3. ‘The results show a high degree of agreement between the experimental data and the model’s predictions,’ reports the research team from the universities of Marburg and Giessen.

Significance for understanding mental illnesses

Sensory attenuation (SA) not only plays a role in everyday perception, but could also explain why some people with mental illnesses have difficulties correctly attributing their own actions. In particular, in the case of schizophrenia, an altered SA could contribute to the fact that those affected have the feeling of being controlled by external forces or do not perceive their own movements as self-generated. ‘The model developed could help to develop new diagnostic and therapeutic approaches in the future,’ concludes Marburg psychology professor Dominik Endres from the results. Endres heads the research group “Theoretical Cognitive Science” at the University of Marburg.

An interdisciplinary joint project

The project was carried out as part of the cluster project The Adaptive Mind and is an example of successful collaboration between different disciplines. The experimental data sets come from the working groups of Prof. Katja Fiehler (University of Gießen) and Prof. Benjamin Straube (University of Marburg), while Prof. Dominik Endres (University of Marburg) and Dr. Anna-Lena Eckert developed the mathematical model and optimised it for the data.