Gautier Anselin – The Challenge of Attention for Direct Realism

The Challenge of Attention for Direct Realism – Gautier Anselin (Institut Jean Nicod (ENS/EHESS))


Should what we may learn about perceptual attention, from cognitive sciences andphilosophy of mind, prevent us from thinking that what we perceive are the things themselves? In this paper, I would like to address the idea that attentional facts threaten direct realism. Direct realism, in analytic philosophy of perception, is the view according to which we are perceptually acquainted with ordinary concrete particulars, or to put it another way, that the content of perceptual experience is not representational (but constituted by the worldly, external objects). One may construe direct realism as an attempt to phrase or flesh out, in the fashion of analytic philosophy of perception, the phenomenological key-idea that what we perceive are the things themselves, “in the flesh”, and not a merely representational ersatz.

Direct realism faces two objections derived either from empirically verified attentional phenomena, or from the logical consequences of common theories of attention in philosophy of mind and cognitive sciences.

First, one consequence analytic philosophers draw from direct realism is that thephenomenology of a given experience must be determined by its objects, and therefore that a change in its phenomenology must be explained by a corresponding change in its objects that causes or constitutes it. But experiments in psychology of perception prove that in visual experiences, when facing a fix set, the subject may change the focus of her attention without moving her eyes and then experience a phenomenological difference in perceived contrast, size, hue saturation, etc. This would be a phenomenological change without any difference among the worldly objects that the experience is supposed to depend upon, according to the direct realist. Ned Block (2010) sees this as a decisive objection against direct realism.

Second, whatever neuropsychological mechanisms underlie attentional phenomena, perceptual attention seems to imply that the subject becomes aware of previously unattended objects, some of which must already have been minimally affecting the subject before she focused on them. A common way to phrase it is to say that attention makes unconscious representations conscious. If we accept the legitimacy, for the neuroscientists studying perception, to use the notion of unconscious mental representations in their research, this leads to an objection. When properly formalized, it provides representationalism with the following argument: the subpersonal level of perception is composed of (unconscious) representations; attention is a perceptual process and its result is an element of perception; attention is a mental process that makes unconscious representations conscious; therefore the personal level of perception is composed of representations; therefore the content of perceptual experience is representational; i.e. direct realism is false.

My goal is to address these two objections. My response to the first is that the argument ultimately presupposes an actually inconsistent construal of the opposition between veridical perception and perceptual illusion. My response to the second is that the argument rests on a triple ambiguity about mental representations: conscious and unconscious representations are not representations in the same sense, cannot strictly share the same kind of content, and are not related to their objects by the same kind of intentionality.