Letter to David Kelley: Whole Entities and Floating Percepts

Actually written AUGUST 22, 2016

Dear Dr. Kelley,

I hope you have the time and interest to answer some questions, the answers to which I see as crucial to my understanding of integration in consciousness.

Situation:

An entity looks like a boulder partly buried in the ground.

This entity structures my sensory experience such that my mind perceives it.

Since I have no conceptual level knowledge of the entity, for all I know it could be

(1) A replica of a boulder made of styrofoam and paint.

(2) An actual boulder.

And if it is an actual boulder, I don’t know how deep it is embedded in the ground. It could be, for example, be embedded

(2A) 3 feet deep.

(2B) 3 miles deep.

Given that perception is the awareness of entities as wholes, I perceive the entity as a whole.

Question:

Does my perceiving the entity as a whole mean I am perceiving the whole entity?

For decades now my own tentative answer to this question has been yes, even if it seems particularly counter-intuitive in the case of the boulder extending 3 miles into the ground while my best guess might be that it only goes about three feet. Yes, I am aware of the whole boulder, no matter how deep it goes. Yes, I am aware of the whole thing, even if it is not really a boulder, but instead a replica.

Perceiving the whole thing does not require getting the size or shape or even the category right. Which, for me, means that perceiving the whole entity does not require sensory awareness of the whole entity.

The core experience of perception is to experience the existence of a unified whole. And since that unified whole includes all of it, we experience the existence of all of it.

Perception “goes beyond” sensation. This “going beyond” is what I take to be the nature of integration in consciousness.

So now I wonder about the relationship between the experiences that get integrated (the input to the mysterious integration process) and the experience that results from integration (the output of the mysterious integration process).

The output, when contrasted against the input, seems “open ended.” Perhaps it more precise to call it “open boundaried” in the case of perceiving the boulder. After perceptual integration, I experience my sensory array that got structured by the boulder as only a part of the boundary of the boulder. The rest of the boundary is not sensorally present to me. I get a sense that it must have a complete boundary, but where that boundary actually lay is an “open” issue. It is open boundaried. The boundary must exist in some contour, but may exist in any contour.

So there’s an open ended-ness in perception similar to the open ended-ness Rand pointed out regarding concepts. (Could it be said that perception involves “boundary contour omission”?)

In considering this issue, I have noticed that integration includes a necessarily available awareness of the relationship between input experiences and output experiences. This input-output relationship awareness enables us to have … 

(1) the boundary contour omission experience I just discussed.

(2) the ability to explore entities and navigate among them.

(3) the ability to do a derivative, reduced focus perception of an entity’s attributes.

(4) the ability to make the form vs object distinction in perception, since the form too requires us to connect our awareness of it with the sensory input that got integrated to produce it.

But this has me wondering whether this input-output relationship awareness really is necessary for integration. I can’t find any necessary reason for this awareness. The three abilities I just listed that come from having this awareness are just that: derivative abilities. And I don’t see why these abilities are necessary for all conceivable perceivers.

Without input-output relation awareness, those abilities are lost and the perceiver would then have “floating percepts” that appear totally unrelated to their sensory input. Such a perceiver would have awareness of entities as wholes, but could not locate them or know the contour of their boundaries. If integration already gives us boundary contour omission for the part of an entity we don’t currently sense, perhaps integration could give total boundary contour omission for those perceivers who can’t connect their floating percepts to their sensory experience.

Of course such float perceivers can’t reduce focus to perceive attributes. So they can’t experience whole entities as subjects of their attributes. And experiencing entities as subjects of their attributes is part of your definition for perception, per The Evidence of the Senses. So float perception violates your definition of perception. But perhaps this is a time to add a clause to the definition of perception to accommodate a broader context of possibility. Perhaps the requirement that we experience entities as subject to their attributes only applies to all the instances of perception we currently know about, but would exclude float perceivers, if any exist.

As for float perceivers failing to explore and navigate entities, I don’t see this failure as relevant to the validity of their means of perception. Maybe they can’t survive more than a minute, let alone reproduce. This does not alter the fact that they perceived, even if it killed them to have done so in a floating manner. (Maybe float perceivers did exist, but died out.)

Finally, the fact that float perceivers cannot experience a distinction between form and object – because they can’t experience the sensory-based form of their percepts qua sensory-based form of their percepts – is useful to me. I find it useful precisely because it provides a radical example of the form-object distinction to use in explaining it to other folks. It’s like a thought experiment to help people grasp the real object of perception (the entity as a whole) in isolation against the forms of perception – by mentally eliminating the forms.

And this, in turn, seems useful to me in explaining the flaw with certain approaches to “artificial intelligence” that model perception on sensing the “form” and getting the “artificial neuro-network” computer to detect the form over a wide range of sensory array contexts. I think the float perceiver example helps explain why such approaches completely miss the real object of perception.

So ya, I’m curious what you think about this stuff.

< Previous Knowledge Theory writing

Next Knowledge Theory writing >