Evidencing sentience in low arousal by probing vestibular processing with caloric irrigations

Consciousness is typically linked to coordinated brain function, thereby precluding mental activity in non waking states. The existing medico-social debates about these conditions, though, speak to the intuition that being unconscious might be more complex than we think. The project suggests that we can evidence sentience in noncommunicating states if we embrace brain-body interactions. This is based on the theoretical stance of active inference which holds that we constantly sample our environment through our senses, in order to create and update internal models of the world with the aim to supervene over environmental uncertainty. Here, environmental uncertainty will be prompted in wakefulness and deep anesthesia by perturbing the vestibular interceptive system with caloric irrigations. Caloric irrigations refer to the injection of water at 44°C for 30 seconds into the inner ear to activate the semi. circular canals of the vestibular system, eliciting the nystagmus reflex and occasionally the sensation of vertigo. Twenty participants will undergo propofol-induced deep anaesthesia during which they receive six caloric irrigations, and additional six after recovery. A matched control group of 20 participants who will receive irrigations (two sets of six) but only during wakefulness will be used. With this design the aim is to show that correction of this induced imbalance a) will be mediated by the brain's ability to predict the consequences, beyond sensory modalities (measured with EEG), b) will be translated into increased physiological arousal as increased heart rate (measured with ECG), and c) lead to the adaptation of peripheral reflexes, ie nystagmus (measured with EOG), all these measures jointly accounting as minimization of prediction errors. Similarly for anesthesia, if participants hold an internal body-world model this will be evident on the brain's predictive processing at high cortical hierarchies and on the adaptation of peripheral body signals as well. Positive evidence for minimization of prediction errors will imply that sentience needs to accommodate implicit mental action and hence lead to a paradigm shift in what is commonly thought about mentation in noncommunicating states. Negative evidence of this process will ensue that being unconscious entails no active mental life. In both cases, the generation of specific data will determine an explainable model of mental function in seemingly unconscious states and help the resolution of neuroethical debates.