Carruthers’s First-Order Representationalism

Philosopher Peter Carruthers's theory of consciousness is considered a first-order theory because it holds that conscious experience arises directly from non-meta-representational, first-order perceptual states—that is, from representations of the world, rather than from representations of those representations (which are central to higher-order theories and differentiate them from first order theories). First-order theories assert that consciousness depends only on the presence and characteristics of mental states that represent the world (e.g., seeing colors, hearing sounds). Carruthers explains phenomenal consciousness in terms of global broadcasting of perceptual states, not meta-representations, and he emphasizes availability to cognition and behavior as the hallmark of consciousness. "Phenomenal consciousness does not consist in, or require, any kind of inner monitoring" (Carruthers, 2000).

Global Broadcasting and Dispositionalism

Carruthers promotes his "dispositionalist" global workspace model, in which a perceptual or cognitive state becomes conscious if and only if it is globally broadcast to a wide array of consumer systems (e.g., verbal report, decision-making, motor control), even if the subject is not aware of it as a mental state. "Phenomenally conscious states are those which are globally broadcast, and which thereby become available to the suite of conceptual and planning systems that guide action and thought." Moreover, "A perceptual state is conscious if it is suitably poised—if it is disposed to be taken as input by a range of consumer systems—whether or not it is actually taken as input by any of them" (Carruthers, 2000).

A decade later, he remained "inclined to think that phenomenal consciousness consists in globally broadcast first-order perceptual content, not in any kind of higher-order representation" (Carruthers, 2011).

Carruthers focuses on access rather than on introspection: Conscious states are those that are poised for cognitive access and control, not those about which the subject is introspectively aware. Thus, he argues for a simpler, computationally parsimonious architecture that doesn't require a distinct and complex metacognitive system to generate phenomenal consciousness.

He explains phenomenal consciousness in terms of "dispositionalist higher-order thought theory," which he characterizes as "a certain sort of intentional content ('analog', or fine-grained), held in a special-purpose short-term memory store in such a way as to be available to higher-order thoughts … all of those contents are at the same time higher-order ones, acquiring a dimension of seeming or subjectivity" (Carruthers, 2000).

Shifts Toward Higher-Order and Back

Over time, Carruthers moved to a more nuanced, targeted approach, tilting toward a higher-order thought theory of self-knowledge, while still keeping phenomenal consciousness grounded in first-order representations. So, even when embracing higher-order theories of introspection and metacognition, as his theory of self-awareness becomes more meta-representational, Carruthers separates these from the foundation of phenomenal consciousness itself, which remains first-order.

Reflecting years later, Carruthers recounts his unusual approach to the first-order-higher-order debate. While he always provided "a general defence of reductive representationalism about phenomenal consciousness," he also critiqued the "first-order theories of the sort proposed by Baars, Tye, Dennett, and others (thereby motivating a form of higher-order account)." He then proceeds to defend first-order theories against his own earlier attack^[1] (Carruthers, 2017).

Carruthers describes, in retrospect, his three earlier goals. The first was "to critique mysterian and property-dualist accounts of phenomenal consciousness (Chalmers, 1996), defending the view that consciousness can be reductively explained in terms of active non-conceptual representations of a certain sort that occupy a specific kind of functional role. An important part in the account involved appealing to a set of purely recognitional phenomenal concepts, since these were deployed to disarm (and explain away the appeal of) the various 'hard problem' thought experiments (zombies, explanatory gaps, and the rest)."

A second goal was to show that "dual-content theory" (according to which conscious experiences have higher-order non-conceptual contents that mirror their first-order counterparts) is preferable to other forms of higher-order theory (Carruthers, 2017).

Significantly, Carruthers says, his third goal at the time (Carruthers, 2000) was to demonstrate that "first-order representationalist accounts of phenomenal consciousness (as proposed by Baars, 1988; Dretske, 1995; Tye, 1995; Dennett, 2001; and now Dehaene, 2014) were inadequate." At the time, Carruthers gave two main arguments. "One was that first-order theories cannot account for our possession of purely-recognitional phenomenal concepts. The latter need to be grounded, rather, in higher-order non-conceptual content (thus requiring Carruthers's dual-content account)."

The other main argument, he says was that "first-order theories can in a sense accommodate, but not really explain, the distinction between conscious and unconscious experience" (Carruthers, 2017).

Carruthers-2017 assessing Carruthers-2000 concludes: "Neither argument now seems to me to have been successful," and he proceeds to critique them both. The later Carruthers concludes that the earlier Carruthers had "rejected first-order representational theories of consciousness on inadequate grounds." As a result, "since there is extensive evidence that conscious experience co-occurs with the global broadcasting of first-order non-conceptual contents in the brain, and since this evidence is most easily accommodated by first-order representationalism, the latter is preferable to any form of higher-order account" (Carruthers, 2017).

Footnote

[1]. Again, I [RLK] love when a philosopher changes their mind.