It’s time for a positive proposal. I claimed that I would introduce suggestions for the building blocks of a new taxonomy of different kinds of reflections. As I argued, we need to further specify the kind of reflections required for Schooler’s “pure experience” and for his meta-awareness, and to get a better grasp on what is meant by “taking stock” and “re-representation”. I also argued that the difference between consciousness and meta-awareness should not be understood as a dichotomy. Rather, we should understand reflection itself as a hierarchical and multidimensional process. So, what exactly is the “taking stock” required for meta-awareness? According to Schooler, meta-awareness requires an explicit representation of the current contents of thought (2011, p. 321). But at least two of the terms involved in this characterization are used in several and distinct meanings: knowledge, and explicit representation.[20] Explicit representation might be interpreted as being itself conscious, or as having symbolic or conceptual content. The notion of knowledge is also problematic, simply because knowledge is a factive verb. It implies that we cannot be wrong.[21] As a result, Schooler built the impossibility of misrepresentation into his definition of meta-awareness. This might be consistent with his claim that the first-level perspective inhabits its own ontological realm. But it also creates a problem, because any view that understands introspection or reflection as an inner perception or re-representation automatically has to allow that this process can go wrong. In other words, it has to allow for misperception/misrepresentation. Moreover, Schooler himself want to allow for a certain kind of misrepresentation, in his terminology translational dissociations—cases in which at the personal level we misinterpret what we experience.
In my discussion of the distinction I claimed that we are able to rule out two kinds of reflections that are not helpful. First, categorization under concepts is one kind of meta-cognitive reflection that is itself unconscious, but necessary for conscious experience. Second, being able to distinguish between self and world is another dimension of reflection, at the highest level, that seems necessary for any conscious experience. Neither of these can be the kind of reflection that distinguishes Schooler’s experience from meta-awareness. In addition I claimed that both the self–other distinction and self-awareness can happen in a number of ways. Different kinds of meta-cognition, the general ability to monitor and control one’s owns cognition, and the ability to attribute mental states to oneself and others, can as a result be further specified and characterized along those dimensions.
But again, what is the kind of reflection or “taking stock” required for meta-awareness? At the end of the last section I suggested that the kind of reflexion involved in “taking stock” could be characterized as a case in which the content of these states (or the states) are accessed by other states, and maybe (unconsciously) evaluated. So at issue are complex processes rather than re-accessed individual states. Such a view is compatible with certain higher-order theories of consciousness. And this would allow that misrepresentation is in fact possible. However, not only do authors like Rosenthal build several meta-representational levels into their theories, the content of the higher level thought contains an element of self, a reference to “oneself”. Self-awareness is built in the analysis, not just any kind of reflection, access, or re-representation. This interpretation uses a certain reading of creature consciousness. It requires that an organism is not only aware but also self-aware. This is a notion of creature consciousness that at first seems to be in tension with Schooler’s main distinction. However, as I argued, this is not necessarily the case. Self-awareness itself comes in degrees and varies along multiple dimensions. Creature consciousness in mind-wandering can than be understood as an intentional relation between the organism and some object or item of which it is aware, in our case a train of thoughts (and/or the sensory states associated with it). This is where the contrast between content theories and functional theories comes into play. As I have argued, pure content or representationalist theories, which claim that conscious states have their mental properties due to their representational properties, are not a good strategic partner for Schooler. In contrast, a certain class of functional accounts, especially higher-order theories, turn out to be a nice fit for his account. These accounts analyze consciousness as a certain form of self-awareness. As a result, we can grant that for the experience of mind-wandering without meta-awareness there is some self-awareness required, and for meta-awareness another more demanding kind of self-awareness is necessary. Rosenthal’s higher-order account would give Schooler this kind of distinction: meta-awareness would include a third-order state, in his terminology a re-re-representation, whereas the experience of mind-wandering would involve only a second-order state, a re-representation (see p. here).
The literature on phenomenology offers more helpful distinctions of how we can further evaluate these different dimensions. Most of these distinctions are orthogonal. Several authors claim that “what is likeness” comes in different forms. For example, Carruthers distinguishes the “what it’s likeness” of the world (or worldly subjectivity—what the world is like for the subject—from experiential subjectivity—what the subject’s experience is like for the subject; Carruthers 1998, 2000). Rosenthal uses a similar distinction. He distinguishes thin from thick phenomenality, whereby thin phenomenality is the occurrence of a certain qualitative character. Thick phenomenality is richer: “[t]hick phenomenality is just thin phenomenality together with there being something it’s like for one to have that thin phenomenality” (Rosenthal 2002, p. 657, emphasis added). So thick phenomenality includes a certain kind of reflexion or extra level; it includes an awareness of a richer kind. For Rosenthal this is identical with the existence of an appropriate higher-order representation. But it is a specific kind of meta-cognitive process, one that contains a representation of “oneself”, or in other terminology, a selfmodel (Metzinger 2003). But the self-model itself, our understanding of ourselves and of the difference between oneself and others, might itself come in degrees and on different levels. So the issue is not meta-cognition or reflexion in general, but different levels and involvements of self-awareness.
Instead of focusing on the differing phenomenology, one might also try to specify the notion of access in further detail (Kouider et al. 2010), a suggestion that I think helps us to better understand what is meant by states referring to other states or accessing them. In the workspace model discussion a simple distinction between cognitive access and cognitive accessibility is introduced to defend the possibility of the abovementioned third category, unaccessed but conscious states. Instead of just access vs. accessibility, I suggest that we distinguish different kinds of access (see p. here). Rather than assuming a rich phenomenology and differing forms of consciousness, one could also propose that awareness itself might come in degrees and that something like partial awareness might exist (Kouider et al. 2010). Instead of distinguishing dissociable forms of consciousness or different kinds of personal level phenomenal character like the above accounts, Kouider et al. (2010) use sub-personal descriptions explaining what awareness might be. More exactly, dissociable levels of access are distinguished and differentiated by a hierarchy of representational levels. In case of partial awareness, we have informational access at some but not all representational levels. The crucial idea is that information at other levels can remain inaccessible. Or, in some situations, information at these levels could be accessed, but plausible content is filled, which than potentially results in misrepresentation.
I prefer this line of thinking, and I believe it gives us an improved understanding of the sub-personal processes involved in the different levels of reflection and “taking stock” we want to characterize. This framework is very suitable for a revised understanding of Schooler’s main distinction. However, Kouider et al. (2010) postulate partial access as an alternative explanation for conscious visual perception,[22] not for internal cases like mind-wandering. But I think the analyses might be useful for our purposes as well. According to this framework, accessible contents at each level of representation are seen as resulting from the integration of signals with contextual prior information, processes that are also influenced by other internal factors (for example attentional factors or vigilance); this integration is further assumed to be modulated by the degree of confidence of the subject. The result is a more fine-grained perspective on conscious experience; instead of simply conscious or unconscious, we can talk about different dimensions of experience. And this is done at the sub-personal level by a specification of access. This also avoids another problem. As I pointed out, Schooler’s account seems very close in spirit to modern versions of workspace accounts. However, these accounts typically assume all-or-nothing mechanisms for access. This is no problem for Schooler, who proposes his core distinction as a dichotomy. However, it is a potential problem for the revised view I suggest. But I think this can in fact be an advantage. We can indeed grant that representations within each level might be accessed in an all-or-nothing manner (as is assumed in workspace models), but none the less insist that the full set of all the representations associated with this process do not have to be conscious.
Different terminologies aside, I think this fits nicely with the spirit of Schooler’s general distinction, and his distinction between experience and meta-awareness. I admit that these are just first steps towards a better conceptual understanding. But interpreted this way, there is not just conscious experience of mind-wandering versus meta-awareness. The situation is more complex. Reflection comes in many forms and involves representations at many levels, as well as access at all these levels of representation. In addition, whether, and to what degree, self-awareness and a self-model is involved makes a difference as well.