- This topic has 2 replies, 3 voices, and was last updated 1 year, 3 months ago by
.
-
Posts
-
Below is part quote and part paraphrase from the section “The Perceptual Level as the Given.”
Some philosophers, most prominently David Hume, have denied the existence of the perceptual level entirely. They start at the level of sensations and try to infer the perceptual level. This is a dead end, and they fail. But this is a self-imposed failure.
Question: Is this necessarily a dead end? Imagine working on object-recognition or edge-detection algorithms for Computer Vision or Robotics. Wouldn’t that enterprise depend on getting sensory input and detecting entities from it? (Perhaps there’s a bootstrapping concern here that I’d benefit from discussing.)
Hume failed to infer perceptions from sensations because perceptions are automatically integrated by our brains. Sensations are not open to inference which is a conceptual process. Having failed, Hume is stuck at the sensations level which makes it impossible for him to get to concepts. He should have realized that he had made a mistake because, by writing about all this in a book on philosophy, he obviously was working with concepts.
Yash, the reason we can’t start at the level of sensations is that, from sensations, we could not grasp entities, identity, or causality. There would be no way to go from “splotch of blue,” “cool pressure,” “blaring noise,” and so on to “a police car just zoomed past.” Sensations simply do not register. In regard to computers, sensations would be totally useless to a computer, because sensations are not integrated or retained. What engineers have to figure out, and are already doing from what I understand, is to register, let’s say, pixels and patterns of pixels, as certain things, thereby enabling the assignment of identity. This is the rough computer equivalent of perception, not of sensation.
- You must be logged in to reply to this topic.