Causation with a Human Face: Normative Theory and Descriptive Psychology by James Woodward

Author:James Woodward
Language: eng
Format: epub, pdf
Publisher: Oxford University Press
Published: 2021-02-15T00:00:00+00:00

Nonhuman animals display an impressive range of abilities in various learning tasks, some of which probably reflect associative procedures and some of which may have other bases.69 Despite this, it is a striking fact that nonhuman animals, including primates, are greatly inferior to humans, including small children, at many tasks involving causal learning, especially those involving tool use, object manipulation, and an understanding of “folk physics.” This is so despite the fact that nonhuman primates and many other mammals not only have capacities for associative learning but also abilities on object permanence tasks and sensitivity to nonphysical behavior (as when one solid object appears to pass through another and so on) that are apparently not so very different from those possessed by human children and adults.70 This suggests that while these various abilities may well be necessary for the acquisition of the causal learning abilities and understanding possessed by human beings, they are not sufficient. Can we say something about what more might be involved?

In approaching this question, let me begin by briefly describing some representative experimental results involving nonhuman primates. (These results underscore my earlier point that although humans make plenty of mistakes in causal learning and judgment, they are nonetheless strikingly more successful at these tasks than other animals. Again I stress that this is something that needs explanation.) Most of the experiments I will describe involve tool use because these are the abilities on which primatologists have tended to focus. There have been fewer experimental investigations of other sorts of learning and reasoning tasks such as those involving contingency information. Obviously this focus on tool use privileges sensitivity to “mechanical” properties that are important to successful tool use and has other limitations as well.71

In experiments conducted by Kohler and subsequently repeated by others, apes (including chimps, orangutans, and gorillas) were presented with problems that required stacking several boxes on top of each other in order to reach a food reward. In comparison with humans, including relatively young children, the apes had great difficulty. They behaved as though they had no understanding of the physical principles underlying the balancing of the boxes and the achievement of structures capable of providing stable support—as Kohler put it, they had “practically no statics” (1927, 149, quoted in Povinelli 2000, 79). The structures they succeeded in building, after considerable trial and error, were highly unstable, and completely neglected center-of-gravity considerations, with boxes at an upper level extending in a haphazard way far over the edges of lower-level boxes. Subjects even on occasion removed lower-level boxes from beneath boxes they supported. Errors of this sort were made repeatedly, suggesting what from a human perspective would be described as complete lack of insight into the causal principles governing the construction of stable structures. When stable structures were achieved, this appeared to be the result of trial-and-error learning (rather than “insight”). There was little evidence that the apes were able to reason hypothetically or counterfactually about what would happen if they were to create this

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.