Paper review of: A neural correlate of sensory consciousness in a corvid bird (doi currently broken). Do crows have sensory / access consciousness?

ObservationsPermalink

Boils down a test for access consciousness to testing the memory representation of a stimulus in the bird - makes access consciousness seem quite weak

  • Memory of stimulus intensity - not access consciousness
  • Memory of whether or not the stimulus was seen - access consciousness

Test protocol makes sense to decouple crow’s response from input stimulus

  • Correct response depends on secondary cue presented after a delay from the stimulus
  • Crow can’t just associate action with seeing the stimulus and vice versa
    • Needs to “access” its memory of whether it saw the stimulus, when it receives the secondary cue after the delay period

Figure 4 b) wasn’t entirely convincing. We did some modelling here.

  • Wouldn’t a similar result be expected from a single, noisy neuron encoding the stimulus intensity?
    • Noise would cause the neuron to become more predictive of the outcome closer to the decision time
    • Noise would also cause the neuron to become less predictive of the stimulus intensity over time - the representation would get corrupted
  • While this could seemingly work for a single neuron, unclear if it would for a whole population
  • Would need to perform more modelling to definitively rule out a simpler explanation than a binary encoding of whether the stimulus was seen i.e. access consciousness

ConclusionsPermalink

  • The findings seemed a bit less “grandiose” than we expected overall
    • Could be simplified to: crows form and access an abstract binary representation of whether a stimulus was seen or not, rather than a representation of stimulus intensity
  • Access consciousness seems quite “weak” - need to investigate more, particularly with regards to global workspace theory
  • It would have been super interesting (but maybe not possible) to as the crow to make two different choices based on the stimulus visibility.
    • If the two decisions were consistent, the crow’s memory is of whether it saw the stimulus. If not, the memory is of the stimulus intensity.
    • Making the first choice might anchor the crow’s second choice, then there wouldn’t be any interesting results.