Introduction to the experimental study of consciousness

The question of the nature of consciousness, long banished from experimental psychology, has for centuries been the subject of numerous speculations, each stranger than the last. Let's briefly recall that Descartes, while explicitly asserting that "God made our body like a machine" and first conceiving "the idea of a self-sufficient nervous mechanism capable of complex actions and having all the appearance of intelligence" (William James), was unable to rid his philosophy of the concept of an immaterial soul, which he curiously attached to the pineal gland due to its central position in the brain. Closer to home, the physiologist John Eccles imagined that the Cartesian interaction between body and soul was produced by the linking of "dendrons" (fundamental units of the cortex) and "psychons" (units of the conscious mind) by quantum effect. Following in his footsteps, physicists Roger Penrose and Stuart Hameroff saw in Gödel's theorem proof that the human mind possesses intuitions that go beyond the calculations accessible to Turing machines, and sought in unknown properties of matter (quantum gravitation) a new form of conscious calculation. For contemporary philosopher Dave Chalmers, it's not incoherent to imagine two exactly identical brains, atom by atom, only one of which is conscious... These contemporary crypto-dualisms are hardly more coherent than Julian Jaynes' own position, according to which human consciousness is a cultural invention that did not exist at the time of theIliad, when men were no more than automatons obeying the voice of the gods!

The conditions for an experimental science of consciousness

Faced with such bizarre and disparate propositions, we can legitimately ask whether a true science of consciousness is conceivable. Does the subjective nature of conscious experience preclude any objective, third-person scientific analysis, such as can be carried out using the methods of cognitive psychology? Is conscious experience, by its very nature subjective, beyond the reach of experimentation? Of course, our lecture takes the opposite view. It aims to show that consciousness is a legitimate object of study, "a real, natural, biological phenomenon, literally localized in the brain" (Antti Revonsuo), and that we have many experimental paradigms at our disposal to study it.

In connection with last year's lecture, let's first point out that the problem of consciousness is well-defined. It is not a question of analyzing the totality of cerebral operations, but only those that go beyond non-conscious processing, which is nonetheless widespread in the nervous system. "The cognitive neuroscience of consciousness aims to determine whether there is a systematic form of information processing, and a class of states of brain activity, that systematically distinguishes states that subjects identify as "conscious" from other states" (Dehaene & Naccache, 2001).

The search for the cerebral mechanisms of consciousness thus involves correlating a conscious mental state with its computational and cerebral mechanisms, a classic approach shared with many other fields of cognitive science. Only one specific condition distinguishes it: the need to gather and take seriously the introspection of the participants, which defines the very phenomenon we seek to study. Thus, "the first and crucial step is to take seriously what subjects report from their introspection and phenomenology. These subjective reports are the key phenomena that a cognitive neuroscience of consciousness aims to study. As such, they constitute the primary data that must be measured and recorded alongside all other physiological data". (Dehaene & Naccache, 2001).

This primacy of the subjective in no way implies that introspection is a privileged mode of access to the intimate mechanisms of the mind. Participants' introspection may be limited, or even false, as in the case of the subjective experience of "leaving the body" (autoscopy). Nonetheless, it is a primary datum of consciousness that needs to be accounted for (e.g., through the abnormal discharge of integrating neurons at the parieto-temporal junction, involved in the multimodal representation of one's own body in space, see Blanke, Landis, Spinelli, & Seeck, 2004).

In Abstract, our research program consists in identifying objective neural correlates of subjective processes accessible to conscious introspection. With Francis Crick and Christoph Koch, we can speak of "neural correlates of consciousness"( NCC). However, our aim goes beyond mere psycho-neural correlation. What we're really looking for is a list of necessary and sufficient conditions for the establishment of a conscious subjective state, in other words, the causal mechanisms of consciousness.

Some theoretical distinctions

The word "consciousness" is polysemous, and the introduction of a precise, restricted vocabulary is an important step in theorizing. Most researchers agree to distinguish, at the very least, between the state of consciousness and the content of consciousness. The first, intransitive, use of the word "consciousness" refers to gradual variations in the state of vigilance: wakefulness, sleep, anesthesia, coma, vegetative state... The second, intransitive usage refers to awareness of a particular piece of information. This is called "access consciousness", or the access of information to consciousness. According to Larry Weiskrantz (1997), the essential experimental criterion is the "reportability" of stimuli: when we are aware of information, it is always possible to report it, either verbally or through communicative gestures. Conscious information is accessible: practically all available processors can exploit it (verbal reporting, evaluation, memorization, intention, control, strategies...).

Other researchers focus on different aspects of the word "consciousness". For some (Antonio Damasio, Gerald Edelman), reference to the self plays an essential role: to become conscious, each percept must be related to a subjective point of view (the self), a representation of the organism in the very act of perceiving. Others emphasize more advanced functions of consciousness: the faculty of recursivity that leads to "higher-order" thoughts, the fact of representing that we think what we think (David Rosenthal, Axel Cleeremans); the development of an explicit and sometimes fictitious narrative of our acts (Michael Gazzaniga, Lionel Naccache, Dan Wegner); finally the ability to share our mental states with others, which suggests that the emergence of human consciousness in the course of evolution could be linked to the social skills specific to our species (Chris Frith).

For others (Ned Block, Victor Lamme), all conscious perception is accompanied by a rich, personal, unique subjective experience that may seem ineffable (the unique color of a sunset, the thrill of music, etc.). These phenomena are referred to by Lewis (1929) as qualia - the particular qualities of sensible or "phenomenal" experience. Phenomenal awareness would precede and exceed what we are able to report (i.e., access awareness). Philosopher David Chalmers thus distinguishes the "easy" problem of consciousness (which is, after all, to explain the entirety of behavior, including voluntary decisions, verbal reports and introspective descriptions!) from the "hard" problem of explaining why these operations are accompanied by conscious experience(what it is like). An "explanatorygap" would separate any attempt at a materialist, externalist explanation of mental states ("third-person") from the intimate experience of being conscious ("first-person").

Although all these facets of consciousness have a certain legitimacy, the lecture was largely limited to an analysis of the mechanisms of access to consciousness. There are two reasons for this choice. Firstly, most recent scientific findings have focused on this well-defined question. Secondly, in our current state of knowledge, the concepts of self and qualia are based more on intimate intuition than on genuine analytical modelling. And yet, like Daniel Dennett (2001), research often leads to a radical revision of such beliefs. Even if introspection seems to validate the idea of a phenomenal awareness of the totality of a visual scene, it may only reflect an "illusion of seeing". Perhaps we really only see a tiny fraction of the external world at any given moment - our sense of global awareness would result from the fact that, as soon as we pay attention, each of them becomes instantly accessible (Dehaene, Changeux, Naccache, Sackur, & Sergent, 2006). Several recent experiments conducted on change blindness and iconic memory paradigms support this view (de Gardelle, Sackur, & Kouider, 2009; Simons & Ambinder, 2005).