Title: Self-awareness - an emerging field in neurobiology
Abstract: One of the current hot topics in the scientific community is the burgeoning interest in consciousness and, in particular, self-awareness, which until quite recently was considered off-limits for the natural sciences. However, advances in the characterisation of self-awareness have made it possible to design experiments that adequately describe self-awareness, not only in terms of its neural correlation, but also how and why it emerges in the brain. The aim of this international workshop was to bring together philosophers and neuroscientists to discuss the progress that has been made so far and the eventual aim is to open up new avenues for research. Philosophical discussions on the subject date from antiquity, and, as Andreas Roepstorff (Denmark) pointed out, the biological interest in self-awareness has only just taken off. As a result, the main focus of the workshop and this report is on this new development. The difficulties encountered in investigating the biology of self-awareness were discussed by Dan Zahavi (Denmark). He drew attention to the discrepancies between the effort applied to setting up self-awareness experiments and discussing the experimental results, while disregarding the very notion of the concept of self-awareness. Morten Overgaard (Denmark) stressed that there are explanatory challenges related to consciousness per se, citing the philosopher Fred Dretske, who put forward the theory that we never have experiences of ourselves, but only on the contents of consciousness. These difficulties have led to indirect ways of assessing self-awareness, either by detecting neural changes as a result of self-perspective versus other perspective, as stated by Kai Vogeley (Germany), or the consequences of placebo manipulations, as stated by Martin Ingvar (Sweden). In contrast, Shaun Gallagher (USA), following Wittgenstein and the phenomenologists, claimed that we are self-aware whenever we are conscious. He said that self-awareness is an indispensible part of our experience of the world, whether it is minimal self-awareness (prereflective, with an automatic sense of ownership of the experience) or narrative (extended and reflective, based on retrieval of episodic memory). At the workshop, he distinguished between a conservative view, in which the immunity to error through misidentification in minimal self-awareness is due to the mode of access (introspection or proprioception) and the liberal view, in which immunity to error depends on the first-person perspective of the experience. Gallagher argued in favour of the latter. Gallagher's comments led to a discussion among the participants on the homogeneity versus the heterogeneity of biological results of self-awareness experiments. The audience agreed that there is an impressive homogeneity of results from diverse experimental set-ups measuring diverse aspects of self-awareness. These range from those that strictly adhere to Gallagher's definition, as maintained by Troels W Kjaer (Denmark) and the author of this report, Hans Lou (Denmark), to self-perspectives versus other perspectives, as stated by Vogeley, to self-recognition as well as indirect, not verbally reported measures as stated by Vogeley and Martin Ingvar (Sweden). Gitte Moos Knudsen from Denmark also mentioned pharmacological intervention by psychedelics. All these diverse aspects of self-awareness have a close link with activity in a paralimbic core network, including medial prefrontal, as stated by Steve Fleming (UK), with Andrea Eugenio Cavanna (UK) adding that these were usually together with medial parietal cortices and also with the angular gyri, insula and subcortical regions. This led Vogeley to propose that the paralimbic network is a common signature to all forms of self-awareness. This network has generally been termed the default network, referring to the fact that while it is exceedingly metabolically active in idleness and tasks relevant to the self, it is often decreased in activity during goal-directed action in the outside world. It is somewhat curious that this network has been called a default network, with 2,000 entries in PubMed, rather than a network of self-awareness, as defaulting is a general feature of a region or network when demand on its activity is reduced. One well-known example is cross-modal sensory inhibition. The results described above are essentially based on imaging data, and, as such, are only correlative. The significance of these findings for self-awareness has been disputed. To test whether the paralimbic default network has a causal function in the generation of self-awareness, manipulation of the network has been carried out. Hans Lou (Denmark) told colleagues that direct electromagnetic intervention has unequivocally shown that the medial parietal cortex (precuneus) is, in fact, instrumental in the retrieval of self-related episodic memory, an essential part of extended (narrative) self-awareness, a process dependent on the emotional content as demonstrated by Anthony Jack (USA), while medial prefrontal cortex is instrumental in self-evaluation, another component of self-awareness. Patrick Haggard (UK) pointed out that the latter is influenced by the experience of self-generated action. Lou (Denmark) said that magneto-encephalography and granger causality computation had been used to show that the interaction between the anterior and posterior hub occurs with oscillations mainly in the gamma range, that it is recurrent and causal and that it is regulated by dopamine. Joachim Gross (UK) said that gamma oscillations are generally linked to cognition and Peter Uhlhaas (UK) added that its disturbance is a characteristic feature of schizophrenia. Generation of gamma frequency oscillations depends on fast-spiking gaba-ergic interneurons. Vikaas S Sohal (USA) reported that experiments with transgenic mice deficient in these interneurons have shown deficient gamma oscillations and behavioural defects which may be restored by optogenetically stimulating prefrontal cortical interneurons, and Lou added that this also established causality at the cellular level. Knudsen and Shihul Han (China) said that causality is also inferred molecularly by the finding that dopamine regulates conscious experience and that it is influenced by psychedelic drugs and oxytocin. There is therefore now evidence for the conclusion that not only is the paralimbic default mode network generally active in biological studies of self-awareness, but also instrumental in self-awareness and its regulation. Axel Cleeremans (Belgium) stated that self-awareness is tightly linked to conscious metacognition, not only as a component of metacognition, but possibly also as a result. Metacognition is a tool for monitoring and controlling behaviour and adjusting our beliefs of the world, not only within ourselves, but, importantly, between individuals. Chris Frith (London) added that this faculty has been shown to provide an advantage when occurring between people with comparable perceptual capabilities. The evolutionary benefits of self-awareness were stressed by Morten L Kringelbach (UK) and Julian P Keenan (USA). Keenan argued impressively for the benefits of deception in phylogenetic and ontogenetic development. Kjaer and Jonathan Smallwood reported that with continuous measurements of cerebral blood flow patterns in self-generated thought during yoga nidra and the normal idle brain have shown that in spite of the great variability of the elicited flow patterns with changing contents of consciousness, there is activity in the paralimbic network in all patterns. Lou stated that this finding is consistent with the assumption that self-awareness is an integral function of all conscious experiences, binding conscious experiences together in a single construct with a sense of unity of consciousness. As well as the phylogenetic aspects, the ontogenetic aspects of self-awareness and metacognition were addressed by Hugo Lagercrantz (Sweden) and Josef Perner (Salzburg). At birth, following a cascade of dopaminergic stimulation, the newborn clearly has a proprioceptive sense of its own body and readily imitates gestures and facial expression essential for social bonding and anticipation of emotional reactions of other people, a process mediated by mirror neurons. The infant is probably only aware of events in the present time, while the adult capacity to put these sensations into a time perspective, with ability to plan for the future has to await later development. Similarly, the development of metacognition does not seem to occur until after three years of age. This development occurs concomitantly with the development of resting state activity in the insula, medial prefrontal and medial parietal hubs, that is the paralimbic default mode network Due to the fundamental biological importance for the individual and society of self-awareness and metacognition, their disturbance is likely to be linked to severe pathology. In fact, a recent study reported by Kristine Römer Thomsen (Denmark) showed that in a monosymptomatic condition such as behavioural addiction, gamma synchrony between medial prefrontal and medial parietal cortices is dysfunctional. That is also the case for more complex disorders involving poor self-awareness and self-monitoring, such as attention deficit hyperactivity disorder and autism spectrum disorder. Even in a pervasive dysfunctional state such as the vegetative state, recovery of the paralimbic network is tightly linked to clinical recovery. Lou pointed out that the exceedingly high oxygen demand of the paralimbic default network makes its functions in self-awareness and metacognition highly vulnerable to deficient oxygen supply. Therefore, proper oxygenation is mandatory. Leif Østergaard (Denmark) presented a new theory of the need to ensure not only adequate blood supply to the brain, but also an optimal distribution of capillary flow in neural tissue. As one of the organisers of the workshop, this reviewer would like to thank all the participants for their enticing contributions to the meeting. The workshop was supported by the Danish Ministry of Science, Innovation and Education's UNIK initiative (MINDLab). A draft of this review was sent to all contributors before publication and Shaun Gallagher, Andrea Cavanna, Chris Frith, Gitte Moos Knudsen and Leif Østergaard all responded by comments and, or, suggestions on how to improve manuscript.