Operationalising What? (Poster)
Document Type:ASSC Conference Item
Topics:Neural Correlates of Consciousness
Deposited by:Liz Irvine
Date of Issue:2010
Event Title:ASSC Toronto
Event Type:ASSC Conference
Abstract:Operationalising what? The problem of whether reports or behaviours are the best way of operationalising consciousness is particularly clear when they conflict, such as in change and inattentional blindness (see e.g. Block, 2007, Mack & Rock, 2000). Based on the idea of perception as hypothesis testing (see e.g. Friston, 2005), it is argued that conflicting reports and behaviours are a natural product of a multi-stream perceptual system. None of these streams better reflect the real ‘contents of experience’, but the existence of ‘default’ reported levels of hypothesis generation may explain our intuitions about consciousness. Using the case study of partial report superiority (Sperling, 1960, Landman et al., 2003, Sligte et al., 2008), the different roles, informational input and time-frames of gist and item-specific processing are explored. This shows how hypotheses determining reports of scene gist (rich content) are produced in parallel to those that drive behavioural capacities (sparse content). In predictable environments, gist level hypotheses typically provide correct descriptions of scenes and constitute our ‘default’ description of experience, but in unpredictable environments gist level hypotheses may be incorrect (de Gardelle et al., 2009). Conflicts between rich reports and sparse behavioural capacity should not be seen in terms of the amount of content in experience, but how accurate hypotheses are. Gist hypotheses are not indicative of the ‘illusion’ of richness, but they can be wrong in certain contexts. This model of report generation and behavioural capacity in terms of multi-stream hypothesis generation suggests that there is no ‘right’ way to operationalise consciousness. Different operationalisations (including introspective techniques) access the products of different processing streams. In this case, consciousness is not a single stage or type of processing that can be better or worse accessed using reports or behavioural tasks, but refers to a shifting range of more or less detailed, and more or less accurate, hypotheses about the contents of an environment. Research into hypothesis generation in different perceptual streams can provide explanations of reports, behaviours, and our intuitions about consciousness, and avoids the problems of identifying the ‘correct’ operationalisation of consciousness.