Skip to main content

Kilteni Laboratory

Try to tickle yourself. No matter how hard you try, the resulting sensation will always feel less ticklish and less intense compared to the sensation produced by somebody else tickling you. In our lab, we study how the human brain predicts the sensory stimuli that are generated by our actions and how this affects our perception.

Konstantina Kilteni on stage in Aula Medica, presenting her research and pointing at image on screen.

How the brain formulates predictions based on our motor commands is a long-standing issue in cognitive neuroscience and more specifically in motor control, but very little is known about it.

Using a combination of psychophysics and state-of-the-art neuroimaging techniques, we study how the brain predicts the sensory consequences of our actions and the contribution of cerebellum and corticocerebellar connectivity in forming and updating these predictions. Given earlier evidence that predictive mechanisms are disturbed in patients with schizophrenia, we are particularly interested in the relationship between predictions and schizotypy personality traits in healthy subjects.

Poster of tickling machines for exhibition at Tekniska Museet.


The neurobiological basis of self-generated tactile predictions

Why can’t we tickle ourselves when we can be tickled by other people? One reason is because our brain can efficiently distinguish the touch that is produced by our movements from the touch that is produced by external causes. This distinction is fundamental for our survival: we must recognize the touch of a spider on our neck as a potential threat while we ignore the touch of our own fingers at the same skin area.

Computational motor control theories posit that the cerebellum makes this distinction by predicting the self-generated touches and attenuating their perception relative to external touches. However, how this mechanism is neurobiologically implemented in the brain remains unknown.

The project tackles four key questions about self-generated tactile predictions:

  1. what aspects of touch they encode
  2. where they are implemented in the cerebellum
  3. whether they can flexibly change during learning
  4. how they are related to emotional and affective features of touch that are vital for social interaction, like ticklishness and pleasantness.

Using state-of-the-art behavioral, neurophysiological and neuroimaging methods, the project will reveal the neurocomputational principles of tactile predictions and their relation to social cognition and schizotypal traits in healthy participants. Therefore, the proposed research is expected to break new ground in the fields of motor control and social neuroscience and have important implications for schizophrenia research.

Research support

  • Swedish Research Council (Vetenskapsrådet)
  • Karolinska Institutet

Selected publications

Efference Copy Is Necessary for the Attenuation of Self-Generated Touch.
Kilteni K, Engeler P, Ehrsson HH
iScience 2020 Jan;23(2):100843

Functional Connectivity between the Cerebellum and Somatosensory Areas Implements the Attenuation of Self-Generated Touch.
Kilteni K, Ehrsson HH
J. Neurosci. 2020 Jan;40(4):894-906

Rapid learning and unlearning of predicted sensory delays in self-generated touch.
Kilteni K, Houborg C, Ehrsson HH
Elife 2019 11;8():

Motor imagery involves predicting the sensory consequences of the imagined movement.
Kilteni K, Andersson BJ, Houborg C, Ehrsson HH
Nat Commun 2018 04;9(1):1617

Body ownership determines the attenuation of self-generated tactile sensations.
Kilteni K, Ehrsson HH
Proc. Natl. Acad. Sci. U.S.A. 2017 08;114(31):8426-8431

Sensorimotor predictions and tool use: Hand-held tools attenuate self-touch.
Kilteni K, Ehrsson HH
Cognition 2017 08;165():1-9


Konstantina Kilteni

Assistant Professor, research group leader