Andrew Y. Lee has written an introductory piece of philosophy, intended for students, celebrating the metaphor of consciousness as a light, illuminating objects in a room. Indeed, this is a common way of speaking, and I have used it myself in published work: If an animal isn't conscious then, I've suggested, "all is dark inside, or rather, not even dark" (Schwitzgebel 2015, sec 7.2).
I've been asked to write a reaction to Lee's piece. On reflection, I've become convinced that the metaphor suggests five theses about consciousness, any or all of which might be doubted. I don't think Lee's piece is circulating yet [update: it is!], but I figured I'd share my thoughts anyway.
-----------------------
1. Metaphors Invite Ways of Thinking.
Metaphors invite ways of thinking. If philosophical disagreements are battles, only one side can win. If memory is a storehouse, recollection requires search and retrieval. If memory is instead a matter of shaping future responses, search and retrieval needn’t be involved. We can of course decline such cognitive invitations. We can describe north as “up” even if it’s lower elevation. But if too many implications are misleading, the metaphor or analogy is inapt.So, is consciousness like having “the lights turned on”? Let’s more thoroughly consider what patterns of thought are invited by this way of speaking.
2. Consciousness Is Determinately Present or Determinately Absent.
As Lee notes, lights are normally either determinately on or determinately off. A dim light is just as “on” as a bright light. Even a flickering light is determinately on or off at any particular moment. Photons are being emitted, or they are not. It requires some creative energy to imagine intermediate cases.If (as I've argued) some states or entities of interest are neither determinately conscious nor determinately nonconscious – states or entities somehow in an in-betweenish, intermediate condition (even if that’s difficult to imagine) – then the light metaphor becomes inapt.
3. Conscious and Nonconscious Cognition Are Similar.
When you flick the lights, the furniture does not change. The table, the rug, and the pile of laundry become easier to see, but apart from absorbing and reflecting more photons, they remain basically the same. If the illuminated objects are parts of your own mind (as Lee's usage suggests), then they too shouldn't radically change when the lights flick on.On some theories of consciousness, we should expect conscious and nonconscious cognition to be similar. Suppose that nonconscious cognition becomes conscious by being targeted by some higher-order self-representational process or by being broadcast across the mind. If so, it’s natural to suppose that elevation to consciousness doesn’t radically alter the contents of a previously nonconscious process. A representational content like red in this part of the visual field right now seemingly needn’t change when targeted by a higher-order representation (that is, a representation of the fact that I am representing red in this part of the visual field right now) or when shared across the mind (“hey, action-guiding and long-term memory centers, please notice that there’s red in this part of the visual field right now”).
An alternative family of views suggests that conscious and nonconscious processes are intrinsically dissimilar. On “recurrence” theories of consciousness, for example, conscious processes involve feedback loops of recurrent processing that differ structurally from non-conscious feed-forward processes. In a different vein, some psychologists distinguish nonconscious “System 1” cognition, which is fast, intuitive, and requires no attentional effort, from slower, step-by-step, effortful, and conscious “System 2” cognition. Consider: What is eight times seven? If “56” effortlessly pops to mind, that’s System 1. If you laboriously add eight seven times or consciously employ a mnemonic like “5-6-7-8, fifty-six is seven times eight”, that’s System 2. System 1 and System 2 processes or outputs might not convert seamlessly one into the other.
If, in general, the structure of nonconscious thought differs from that of conscious thought, the conscious light metaphor risks misleading us. Furniture doesn’t normally change shape when you flick the light-switch.
4. Consciousness Involves Knowledge.
Why do we care about lights? Mostly because they help us see. Illuminated objects are more readily known than those in the dark. Cross-culturally, light is associated with knowledge and understanding. The light metaphor connects consciousness and knowledge. If “the lights are on” in a dog, or a snail, or a comatose person, they know what’s going on. If the lights are off, they are, so to speak, mere reactive machines.If the contents of the room are the contents of your mind, illumination suggests self-knowledge. Your mental states, though being illuminated, become knowable within the perspective of the room. But I think I also hear a reading of “the lights being on” that doesn’t require the entity to understand its own mind. Maybe a “lights on” garden snail needn’t have anything as sophisticated as explicit self-representations of its own mental states. The illuminated furniture might then be analogous to external objects or events of which the snail is consciously aware. (This could also potentially help with the issue described in the previous section, since external objects don’t normally change by virtue of our thinking about them.)
5. Subjectivity and Phenomenal Character Are Distinct.
A light source is one thing, an illuminated object quite another. Analogously, perhaps, we should distinguish consciousness itself (Lee calls this “subjectivity”) from the objects or contents of consciousness (Lee calls this “phenomenal character”). The light metaphor suggests that the source of illumination and the object illuminated are distinct, and that the light source is causally responsible for the object's illumination. Maybe, for example, we can turn attention to our own thoughts, and this turning of attention is the distinct and separate cause of the illumination of the thoughts.Could the objects instead be self-luminous? Imagine not a light source amid reflective objects but rather a room of glowing objects. If there are processes that are intrinsically conscious by virtue of their own internal structure rather than by virtue of some relational feature like being a target of attention, then the most natural, vanilla interpretation of the light-and-room metaphor misleads, though adapting the metaphor to glowing objects might work.
6. Conscious Entities Come in Discrete, Unified Bundles.
Let’s keep playing with the metaphor. In the vanilla case, every conscious subject has one light, illuminating one room. But maybe we can imagine a series of linked caverns, progressively dimmer and less directly illuminated. Maybe we can imagine multiple lightbulbs in different recesses or partly shaded by room dividers, with partly overlapping spheres of illumination.Philosophers and consciousness scientists typically treat conscious subjects as unified and discrete. If you are (consciously) enjoying a sip of coffee, thinking about your dog, and hearing car horns in the distance, then one conscious subject is having those experiences conjointly, and no one else is having those very same experiences. Someone else could potentially have exactly similar experiences, just like someone could have a car exactly similar to your own – but they wouldn't thereby have your car or your experiences. Your light illuminates your experience of coffee with your experience of thinking about your dog with your experience of car horns. Someone else’s light illuminates a wholly disjoint set of (possibly very similar) mental furniture.
But maybe minds needn’t in general work that way. Could your coffee-sipping experience be unified with your dog thoughts, and your dog thoughts with your horn-hearing, with no unified experience of all three together? Could conjoined twins whose brains overlap (yes, there are real cases of this) share some experiences while not sharing others, illuminating from both sides some objects in a hallway between two rooms? Standard animal biology makes overlapping brains rare, but if it’s ever possible to create consciousness in artificial systems, overlap might become the norm. Efficiency might require systems to share some (conscious?) cognitive centers. Conscious subjects might then overlap, defying separation into discrete bundles. Creative effort would then be needed to adapt the metaphor of lights and rooms.
7. Working With or Against the Metaphor.
To treat consciousness metaphorically as a matter of “the lights being on” invites, but does not compel, a picture of consciousness on which (1.) entities are generally, at any particular moment, either determinately conscious or determinately nonconscious; (2.) conscious cognition and nonconscious cognition are fundamentally similar; (3.) consciousness is intimately linked to knowledge; (4.) consciousness involves a relationship between an object and some distinct thought or process that makes the object conscious; and (5.) conscious subjects are generally unified and discrete rather than disunified or overlapping. If 1-5 are all correct, the light and room metaphor works well. Employing it greases the path to correct thinking.If any of 1-5 are not correct, we can cancel that aspect of the metaphor. Just as we cancel the implication of higher elevation when we employ the “north is up” metaphor, we can cancel the implication of determinacy. We can specify that consciousness is like a light, except that it is commonly indeterminate whether it is off or on. Indeed, some cancelations are straightforward, if the conclusion would be obviously false. It hardly needs to be said that minds don’t contain literal sofas.
Alternatively, we can modify or enrich the metaphor. We might say that consciousness is like a light in a room, except that unlike in a typical room, every object glows with self-illumination. We can treat the illuminated objects as outward things instead of mental processes, altering theses 2-4. But too much modification destroys the metaphor. Incomprehensibility ensues if we attempt the idea that consciousness is like a light illuminating a room, except that the objects are self-illuminating, and often neither determinately off nor on, and there’s no relationship between illumination and knowledge, and there’s no discrete number of light sources or rooms, and objects are radically different when they are illuminated than when they are dark. Better, if so, to say that consciousness is not like a light illuminating a room.
Toying with the metaphor can open a flood of questions with a brainstorming character. If objects are normally illuminated on one side but not another, do conscious states likewise have an unconscious side? Can the room have dark corners? Could different types of light illuminate different types of object? Could there be mirrors or a transparent wall between two different rooms? Could some objects come and go while others are always present but only sometimes illuminated? Some questions will carry the metaphor too far, being nonsensical or misleading when extended to the case of consciousness, but the exact boundaries of fruitful extension can’t be known in advance of the attempt.
