I have been asked to write an entry on higher-order theories of consciousness for the Routledge Encyclopedia of Philosophy, which apparently has not ever had an entry on this! Below is a very (very) rough draft of the entry so far. It really isn’t much more than a first-draft and obviously needs a lot of work but it will give you some idea of the direction I am heading. Any feedback/criticism would be most welcome!
Higher-order theories of consciousness take a variety of forms but they are united by the claim that consciousness crucially involves some kind of inner awareness of one’s own mind. Though there are clear historical precedents and inspirations in the work of Aristotle, Descartes, Locke, and Kant it is not clear which version (if any) of higher-order theory these historical figures had. There is among these thinkers seemingly a commitment to the idea that consciousness requires some kind of inner awareness but higher-order theories were most clearly formulated in contemporary philosophy of mind. This entry will focus on contemporary developments.
- The Higher-Order Approach to Consciousness
When giving a theory of consciousness one must first delineate what the target phenomenon is supposed to be, especially when pursuing something as ambiguous as consciousness.
We say of creatures that they are conscious or unconscious, that they are awake or asleep, etc. This has been called creature consciousness(Rosenthal). This can be contrasted with what is often called state consciousness, which marks the contrast between a particular mental state being conscious versus unconscious (as in subliminal perception).
Phenomenal consciousnesscaptures the subjective ‘what it is like’ component of consciousness. When we taste chocolate, see red, experience pain, hunger, or anger, there is something that it is like for us to have those experience. The specific way that it is for us to have those experience consist in various phenomenal properties (Chalmers).
Higher-order theories are often cast as theories of state consciousness. That is, higher-order theories are often aimed at explaining what the difference is between a state which is conscious and a state which is unconsciousness. The higher-order strategy is to appeal to the inner awareness that we have of our own mental lives. A conscious state, on this approach, consists in my being aware of myself as being in that state.
Some higher-order theorists go so far as to deny that phenomenal consciousness exists (Rosenthal). However, there is a natural way to connect these two notions of consciousness. When one is in a mental state that one is in no way at all aware of being in, there is nothing that it is like for one. For example, when subliminally presented with a red strawberry, so that one denies seeing it, it is natural to say that it is not like seeing red for one. It is also natural to say that the state which represents the strawberry and its redness is unconscious. The converse of this is that when there is something that it is like for one to see the red strawberry one is in some way aware of oneself as being in the state that represents it. Thus when a state is conscious there is something that it is like for one to be in that state. This is the way in which these terms will be used in this entry.
Construed in this way higher-order theories of consciousness aim to explain phenomenal consciousness which is the same as trying to explain state consciousness. Traditionally we recognize two ways in which we can become aware of things in our environment, which are by perceiving and by thinking. First-order theories argue that phenomenal consciousness can be understood by appeal to the awareness of the world. Higher-order theories argue that these first-order states are not enough and in addition to an awareness of things, properties, and facts about the world we must also have an awareness of our outer-directed awareness. This inner awareness is higher-order in that it is an awareness of something that is mental rather than in the environment or the animal’s body.
- Higher-Order Thought Theories
Classical higher-order theories often appealed to inner sense or inner perception as a way to capture inner awareness (Armstrong; Lycan). But this kind of view has faced difficulties which have rendered it all but obsolete. First, we do not have any reason to posit higher-order mental qualities (Rosenthal). In addition, we have not discovered any kind of inner sense (Lycan and Suret).
Since we can also be aware of things with the appropriate thought higher-order thought theories appeal to intentional thought-like states to explain the way in which we are aware of our mental lives.
Perhaps the earliest explicit version of this kind of theory is that of David Rosenthal. On his view we become conscious of our first-order mental states via having a thought to the effect that we are occurently in those states. This thought must have assertoric force and indicate that the relevant mental qualities are currently present.
Higher-order thought theories themselves come in many different varieties, each with a different structure posited. What units them is the postulation that there are two levels of content in the mind. The first level of content represents the environment, the second, higher-order level, represents the first level.
One model of the relation between these, which I will call the Relational Model (RM), is as follows. One starts with an unconscious mental state and then one adds a higher-order representation of that state which results in the first-order state becoming conscious. The consciousness of the first-order state is explained, on this model, by the relation-the awareness relation- that holds between the first order state and the higher-order state. The first-order state is conscious because you are aware of it. On this way of thinking the higher-order state is a distinct mental representation.
Some have felt that this is unsatisfactory because my awareness of non-mental items like rocks does not result in the rocs becoming conscious (Goldman). RM theorists have responded that theirs is a theory of mental state consciousness and so does not include rocks. To be made conscious, on RM, we require a mental state to become conscious in the first place. Whatever the merits of this response there is an additional a well-known objection based on the possibility of misrepresentation. Since RM claims that there are two distinct states one may misrepresent the other. So, if one is representing that there is a red tomato in the environment but then has a higher-order state that represents one as seeing a green tomato, what is it like for the individual in question? (Lycan) According to RM it is the first-order representation of red that is conscious but it is also the case that the higher-order state determines what it is like for you. This suggests that there are deep problems with RM (Block).
Because of this some have moved to what I will call the Joint-Determination Model (JM).On this model the first-order state is postulated not to be a distinct mental state but rather to be part of the conscious state itself. JDM posits that there is one state with two contents. Part of the content is first-order and part of the content is higher-order. JM comes in different varieties (Kriegel, Gennarro, Lau). One major difference between these models is that of whether the higher-order state itself employs conceptual content (Kriegel, Lau). Some versions, which I will call Same-Order Models(SOM) claim that the higher-order content is itself conceptual and then seek to rule out misrepresentation worries by putting restrictions on the kind of higher-order content that results in a conscious mental state. Gennaro is the most vigorous defender of this kind of view. On his account a conscious mental state results only when there is a (full or partial) conceptual match between first-order and higher-order states, or when the first-order content is more specific than the higher-order content, or when the higher-order content is more specific than the first-order content, or when the higher-order concepts can combine to match the first-order representations (2012 p 179). All of the provisos are arrived at so as to block the claim that there can be a conscious mental state in cases of mismatched content between higher-order and first-order states. However, they seem ad hoc. When examining the cases presented in detail it seems straightforwardly the case that the higher-order content determines what it is like for one. Why wouldn’t it be that way for case of radical misrepresentation as well?
Other versions of JDM that I will call Split-level Models (SLM) deny that the higher-order state is itself conceptual in this way (Lau, Lau and Brown). On these versions the higher-order state is some kind of ‘mere’ pointer, which points to the relevant first-order state. The content of the conscious state is given by the content of the first-order state, but that it is a conscious experience at all is given by the higher-order state. In its most recent iteration the higher-order state ‘toggles’ between three states indicating that the first-order state is veridical, held in working memory, or just noise. SLM is distinct from the other versions of JDM because of what the theory claims happens in radical misrepresentation. On SOM, when one just has the higher-order representation and no first-order target at all there is no conscious experience at all. On Model SLM one will have some kind of conscious experience but it will not be specific. That is to say on SLM the higher-order state will indicate that one is verdically perceiving something but if one has no relevant first-order state then there will be no content to experience other than that one is veridically perceiving something. When one goes to report what it is one will fail.
This extravagant disjunctive theory has been resisted by those who endorse what I will call the Non-Relational Model (NRM).NRM rejects the claim that the first-order state is made conscious by the higher-order state (Rosenthal, Brown). On NRM it is the higher-order state itself that accounts for conscious experience. There is some disagreement among those who endorse this model as to which state is the conscious state. Rosenthal has suggested that it is the notional state that becomes conscious (Rosenthal, Weisberg). Berger has suggested that it is the individual that becomes conscious and not the state at all (Berger). Brown has suggested that it is the higher-order state itself that is phenomenally conscious (Brown).
- Still Further Varieties of Higher-Order Theory
In addition to these kinds of theories there are non-traditional ways to account for the inner awareness that many think is a crucial part of phenomenal consciousness.
On the one hand are those theories that explicitly seek to find some non-traditional form of inner awareness. On the other hand, are those that deny this and yet end up appealing to something like inner awareness.
Lycan has recently argued that his version of higher-order perception really is a version of the attention hypothesis. In his paper with Wesley Sauret he argues that attention is one of the ways in which we can become aware of things. On this view attention makes us aware of our mental states but it does so not by representing the states in question. They appeal to analogies like a funnel or sieve. A funnel directs something, a fluid say, towards a target but not by representing what is being directed. As recognized by these authors work remains to be done to explain what exactly the relation is, they suggest that it may be some kind of acquaintance.
In a similar vein other theorist have adopted some kind of ‘inner acquaintance’ view (Hellie). Hellie presents a version of higher-order acquaintance as a non-intentional relation of awareness to one’s first-order qualitative states. Chalmers has also endorsed a non-reductive, non-physical version of higher-order acquaintance. On Chalmers view to be aware of x is also, by the very nature of phenomenal awareness, to be acquainted with one’s awareness of x (Chalmers). This may be a (non-reductive, non-physical) version of SOM above.
There also have been philosophers who have sought to implement inner awareness via a quotational model (Coleman,Picciuto). On Coleman’s model one quotes a quality and thereby becomes conscious of it. This view requires that the mental quality is already primitively red and is fundamental (Coleman endorses pan-qualityism). The quotation of that red quality makes it a phenomenally conscious experience. On Picciuto’s view one forms a phenomenal concept of the relevant mental quality. As Picciuto formulates it, the mental quality does not have an intrinsic redness to it but becomes qualitatively red once one quotes it.
Finally there are those who seek radically non-traditional ways. For example Ned Block has agreed that some kind of inner awareness is necessary for phenomenally conscious experiences (Block). He denies that this kind of inner awareness is any kind of cognitive awareness. He has suggested that it may be a deflationary kind of awareness. Much as I walk my own walk or smile my own smiles, so to I am aware of my own phenomenally conscious states. This kind of deflationary move seems to include every mental state as phenomenally conscious. On the other hand Block has suggested that some kind of same-order awareness may do the trick (i.e. a version of SOM). However it is unclear how this notion of non-cognitive awareness differs from any of the models canvased above. Perhaps Block will ultimately settle on something like JDM but if so the relevant notion of awareness will seem to be cognitive after all. Or perhaps he will ultimately settle on something like acquaintance but then that needs to be spelled out.