How should we distinguish hallucination from perception? Traditionally, the distinction has been that in hallucination the apparent object is not present; or more precisely, that no closely resembling object is in clear view (the close resemblance condition means that we can count minor misperceptions as instances of seeing, rather than as instances of hallucination).
However, since veridical hallucination is at least a possibility, there is no principled objection to the idea that hallucination can present occurent objects in clear view. Grice (1961) famously used the possibility of veridical hallucination to motivate his causal theory of perception: our concept of perception must include causation, he claimed, since what distinguishes perception from veridical hallucination is a causal connection.
In order to rule out certain problem cases, Lewis (1980) proposed that causation should be described in terms of counterfactual dependence. He claimed that perception, unlike hallucination, displays a counterfactual dependence on the scene before us: if that scene changes, then our perception must change accordingly. Note that for this to be plausible, the dependence of perception on the scene can only be loose: the close resemblance condition must still apply.
Yet Lewis's account is challenged by the phenomenon of change blindness (e.g. Rensink et al 1997, Simons and Chabris 1999, Noë 2002), in which certain changes in a scene do not feed though into immediate conscious visual experience: it can often take some time to identify the change. The alteration in the scene will often be too great to be covered by a close resemblance condition. As a result, there seems to be no counterfactual dependence of experience on the actual scene, and yet we would presumably not want to say that these were cases in which, once the change occurred, we stopped seeing and started to hallucinate.
If we accept this conclusion, it seems that counterfactual dependence may be a sufficient, but not a necessary, condition for perception. Instances of counterfactual dependence will count as instances of perception, but there will also be instances of perception that do not satisfy this condition.
Change blindness cases are admittedly not veridical cases, but this seems unimportant: if we want a distinction between perception and hallucination, we want one that can apply in all cases, veridical or not. How should we respond to this challenge? I consider three possible responses:
I tentatively favour this last proposal.
References
Grice, P. (1961). 'The causal theory of perception'. Proceedings of the Aristotelian Society Supp. Vol. XXXV: 121-68.
Lewis, D. K. (1980) 'Veridical hallucination and prosthetic vision'. Australasian Journal of Philosophy 58(3): 239-49.
Noë, A. (ed. 2002). Is the Visual World a Grand Illusion? Thorverton: Imprint Academic
Rensink, R. A., O'Regan, J. K., and Clark, J. J. (1997). 'To see or not to see: the need for attention to perceive changes in scenes'. Psychological Science 8:368-373.
Simons, D.J. and Chabris, C.F. (1999). 'Gorillas in our midst'. Perception 28: 1059-74.