ABSTRACT
On a superficial level the behavior of SCRAPS appears to model human cognitive behavior: M arker-passing seems to resemble neurophysical models of activation, the paths found seem to resemble actual inferences, and the overall behavior of the system seems to resemble the behavior of some sort of “subconscious process” th a t screens information prior to conscious processing. It should be clear, however, th a t this is far from the tru th of the m atter: Neurons don’t correspond to the frames in FRAIL, human inferences are far more complex than the simple assertions SCRAPS generates, the information passed in marks is too complex to be neurologically plausible, and there’s no clear dividing line in my system th a t would let one distinguish between w hat is subconscious and w hat isn’t. Why then a chapter entitled “ cognitive aspects”?
