Accessibility and Consciousness
- Sciences, Normes, Décision (Paris-Sorbonne)
Talks at this conferenceAdd a talk
In the late 1980's and early 1990's, psychologists' renewed interest for consciousness has led to a fruitful discussion with philosophers, as empirically minded theorists imported experimental findings in the framework of the study of phenomenal consciousness. One important debate concerns the identification of the neural correlates of sensory conscious experience. Reacting against the cognitive approach promoted by such theories as Baars' global workspace or higher order reflexive theories of conscious experience, Ned Block famously distinguished between access consciousness and phenomenal consciousness, thereby drawing a sharp contrast between cognitive 'global' characterizations, and non-cognitive 'local' characterizations of the neural realization of conscious experience. Successive elaborations of this distinction have led to the 'overflow hypothesis', according to which, it is conceivable that some states, although eluding any kind of access, are nevertheless phenomenally conscious. This hypothesis has received support from psychologists and philosophers (for example : Burge, Lamme), but is also faces a strong barrage of criticisms, especially, as it casts doubt on the capacity of conscious report to adequately capture one's experience, making neural activity the sole marker of phenomenal consciousness, even in the absence of report as to the presence of any conscious state.
How to fully articulate the distinction between access and phenomenal consciousness ? How can these two aspects of consciousness be empirically investigated ? Whereas Block's proposal of a purely phenomenal, non-cognitive concept on consciousness has received support, some philosophers and psychologists have expressed a series of doubts as to the empirical tenability of such a concept. One might accept, for example, a conceptual distinction between access and phenomenology, while challenging that this distinction captures any real empirical distinction. A different, methodological worry is expressed by defenders of the global workspace : how can phenomenal consciousness be investigated independently of access ? A strong reading of the distinction has as a consequence that some phenomenal states will elude access and conscious report : however, once the link between phenomenology and reportability is severed, how can one ascertain the presence or absence of conscious sensory episodes (Kouider)? Defenders of the phenomenal/access divide will hold that there are good empirical evidence - stemming both from a first-person perspective and from considerations on the neural signature of sensory states - that there are indeed 'overflowing' phenomenal states, whereas defenders of the identification of phenomenology with access will deny any such overflowing sensory states, and eliminate them in favor of unconscious states (Dehaene & Naccache, Kouider). These deflationary considerations concerning phenomenal consciousness are supplemented with error theories (Dehaene, O'Regan) or alternative accounts of conscious contents (Stazicker, Grush, Philips), to explain away the idea of a 'rich' phenomenology which outstrips the ressources of access. How can the access/phenomenal distinction be empirically investigated ? What is the cost of this distinction for our ordinary conception of consciousness ? Can an anti-cognitive conception of consciousness, detached from any notion of access or accessibility, be fully articulated ?
Both the traditional 'hard problem' and the access/phenomenal distinction are concerned with the following issue : what distinguishes phenomenally conscious from unconscious sensory states ? The arguments concerning the nature of qualia are supposed to be independent from those concerning the phenomenal/access divide. Indeed, the latter debate concerns the mere identification of the correlates of phenomenal consciousness, with no further explanatory pretense as to the fundamental nature of phenomenal states and properties. However, (i) on the one hand, phenomenal consciousness is precisely what the explanatory gap is concerned with – the overall feel of sensory experience ; and, (ii) on the other hand, accepting the strong version of the phenomenal/access divide does undercut a functionalist proposal, according to which phenomenally conscious states are a subset of sensory states, which are poised or available for the central executive system. At least one reductive route is closed by the distinction. Accepting two distinct concepts of consciousness seems to strip phenomenal consciousness from any essential role ; phenomenology can indeed play a role by 'greasing the wheels' of cognition, but this falls short of any explanatory reduction. By contrast, access theories seem committed to a functional-relational understanding of phenomenal consciousness : either because they consider the availability of certain sensory states to be what makes them phenomenally conscious (Prinz, Dretske, Tye), or because they take consciousness to consist in the actual consumption of certain sensory inputs (Dennett, Graziano). Making sense, from a descriptive point of view of the metaphysics underlying theories of phenomenal consciousness, in the framework of the phenomenal/access debate, should also show which properties are shared by conscious and unconscious sensory representations, and which properties distinguish them.