Enaction and the Intentionality Debate — Epoche C2
The distinction, as Block drew it Ned Block's 'On a Confusion about a Function of Consciousness' (1995) separates two things the word 'conscious' runs together. A state is access-conscious when its content is poised for free use in reasoning and for the direct rational control of action and speech: the information is there to be drawn on, whether or not it is being drawn on now. A state is phenomenally conscious when there is something it is like to be in it — when it has experiential character. Block's purpose was hygienic rather than metaphysical. Arguments purporting to show that consciousness has some function, he observed, typically establish the claim for the access notion, where function is definitionally in play, and then quietly conclude it for the phenomenal one, where it is not. The distinction exists to make that slide visible. Two cases carry the paper. The first is superblindsight, a deliberately hypothetical extension of the clinical phenomenon in which patients with damage to primary visual cortex, who report seeing nothing in part of the field, nevertheless perform above chance when forced to guess about stimuli presented there. Block asks us to imagine a patient who has been trained to volunteer such information spontaneously and use it in reasoning. That would be content poised for use with no experiential character: access without phenomenality. The second runs the other way. You become aware of a pneumatic drill outside the window and, in the same moment, judge that you had been hearing it for some time. The natural description is that there was something it was like to hear the drill before the noise was available for report or reasoning: phenomenality without access. Block is careful about the status of these cases. He does not claim the two come apart routinely in human beings; he claims the two concepts are distinct, so that evidence for one is not automatically evidence for the other. The dispute the distinction is meant to referee, and a correction to its labels The distinction is standardly deployed as a tool for adjudicating between the two main families of theory about what makes a mental state conscious, and the first version of this essay named one of those families incorrectly. The correction is worth making at the outset, because the mistaken label points to an unrelated doctrine. The first family is higher-order thought theory, associated above all with David Rosenthal (2005). Its claim is relational: a mental state is conscious in virtue of being the object of a suitable further thought — roughly simultaneous with it, arrived at non-inferentially, and to the effect that one is oneself in that state. Nothing about the first-order state itself makes it conscious; the accompanying thought does. The second family is first-order representationalism, of which Michael Tye's (1995) is the most developed version: phenomenal character just is a certain kind of first-order intentional content, on Tye's account content that is poised, abstract and nonconceptual. No higher-order state is required; a state is conscious because of the content it carries and the functional position it occupies. The earlier version of this essay called the second family 'Lower-Order Thought (LOT) theory'. There is no school of that name, and the abbreviation is already spoken for: in philosophy of mind 'LOT' means the language of thought, the hypothesis that thinking consists in computational operations over sentence-like internal representations, which is a claim about the format of cognition and is neutral between the two theories at issue. The description the essay gave — that a state is conscious simply by being a first-order state of a certain kind, without an additional higher-order state — is an accurate description of first-order representationalism, and that is the name used from here on. The substantive contrast is unaffected: the dispute is whether being conscious is a relational property of a state or a matter of the state's own content and role. The premise the two sides share Block's distinction is well suited to refereeing that dispute, and the reason is also the reason it is the wrong tool for the question this essay raises. Both families, and the distinction that referees them, share a commitment that neither states because neither doubts it: the content of a conscious state is something the state has , individuated prior to and independently of the activity in which it gets deployed. The word 'poised' is the tell. A content is poised when it is available for use in reasoning and the control of action — and availability presupposes an item that already exists, already determinate, sitting there to be drawn upon. Rosenthal's higher-order thought must be a thought about a first-order state, hence about something whose identity is settled independently of the thought that targets it. Tye's content is nonconceptual, but it is still a content the state bears, specifiable by what the state represents the world as being like. On all three views, the story has the same shape: first an item with a content, then a relation to it — access, or a higher-order thought, or a functional role — and the theoretical work goes into the relation. That shared shape is what the enactive tradition attacks. Naming it as a specific commitment, rather than as a lingering Cartesian atmosphere, is the first thing this essay adds to its original version, because a commitment can be assessed and an atmosphere cannot. Enaction, and the experiment behind it Varela, Thompson and Rosch's The Embodied Mind (1991) proposes that cognition is embodied action: perceptual capacities are not a channel through which a pre-given world is registered, but grow out of recurrent patterns of sensorimotor engagement, so that organism and world specify each other rather than one representing the other. Their image for it is laying down a path in walking. That is a large claim, and the book's authors are clear that it n