Varieties of perception 53
perception of affordances and the notion of natural information that I suggested
was held by Gibson, there might be a good reason for believing otherwise: to the
extent that an external observer could apply point-to-point mappings, in terms
of a multidimensional mathematical function between, first, the variable as it is
being registered in any perceptual situation that falls within the range of what the
perceiving organism is adapted to, second, the variable as it can be measured by
that external observer, third, the concrete conditions under which the ability to
perceive that variable in a certain way is realised and, fourth, the adaptive function
of that ability under the given type of conditions, the information that is seemingly
misrendered is actually preserved.
Any apparent mismatch between perceived qualities and measurable variables
detectable in perception can be resolved by accounting for the very situation in
which perception occurs and the adaptive function of the mapping that applies
to this type of perceptual situation. If, under certain conditions, something looks
longer, brighter and so forth to the organism than measurement would suggest,
these conditions and the functions which that seeming departure serves override
correspondence with physical variables. More precisely, adding a “condition” and
a “function” variable to the equation will keep the informational mappings intact
and allow for an, albeit complex and indirect, correspondence to the values of the
physical variables – if and when perception works normally.
Under a genuinely ecological perspective, and in keeping with the strategy of
the Empirical Strategy, perceived qualities will be overridden by the informational relations on which successful activities are grounded. Genuine misperception will occur only if, returning to Warren’s aforementioned case, the subject
perceives a stair to be climbable when it is, in fact, not climbable, or vice versa, or
if he mistakes an exhaustingly steep stairway for a conveniently climbable one.
6
The subject will then misperceive the respective affordance – which does not even
need to imply that he or she is mistaken about the absolute physical measures of
the object. Conversely, perceptual illusion will occur if the subject turns out to be
mistaken about absolute, extrinsically measured riser heights or transformations
of heights in the course of the experiment – which, if the experimental set-up
matches normal conditions for perception, is unlikely to negatively affect perception of an affordance. Hence, perceptual illusion and misperception of affordances
may coincide but are independent issues. In some cases, perceptual illusions may
even contribute to the correct perception of an affordance.
Although the preceding discussion suggests that the Empirical Strategy in some
respects elegantly complements the ecological view, two notable limitations to
this approach should be mentioned here: first, almost all examples discussed by
the authors are concerned with cases in which identical targets look different to
the perceiver under different conditions, whereas their point of departure, the
inverse projection problem, was identified by Berkeley as the problem that different targets, when placed in a certain relation to the perceiver, will look identical. Berkeleyan ambiguity does not imply that perception is at variance with the
measurable properties but that it becomes ambiguous by one retinal image being
in accordance with different instances of measurable properties. The Empirical
perception of affordances and the notion of natural information that I suggested
was held by Gibson, there might be a good reason for believing otherwise: to the
extent that an external observer could apply point-to-point mappings, in terms
of a multidimensional mathematical function between, first, the variable as it is
being registered in any perceptual situation that falls within the range of what the
perceiving organism is adapted to, second, the variable as it can be measured by
that external observer, third, the concrete conditions under which the ability to
perceive that variable in a certain way is realised and, fourth, the adaptive function
of that ability under the given type of conditions, the information that is seemingly
misrendered is actually preserved.
Any apparent mismatch between perceived qualities and measurable variables
detectable in perception can be resolved by accounting for the very situation in
which perception occurs and the adaptive function of the mapping that applies
to this type of perceptual situation. If, under certain conditions, something looks
longer, brighter and so forth to the organism than measurement would suggest,
these conditions and the functions which that seeming departure serves override
correspondence with physical variables. More precisely, adding a “condition” and
a “function” variable to the equation will keep the informational mappings intact
and allow for an, albeit complex and indirect, correspondence to the values of the
physical variables – if and when perception works normally.
Under a genuinely ecological perspective, and in keeping with the strategy of
the Empirical Strategy, perceived qualities will be overridden by the informational relations on which successful activities are grounded. Genuine misperception will occur only if, returning to Warren’s aforementioned case, the subject
perceives a stair to be climbable when it is, in fact, not climbable, or vice versa, or
if he mistakes an exhaustingly steep stairway for a conveniently climbable one.
6
The subject will then misperceive the respective affordance – which does not even
need to imply that he or she is mistaken about the absolute physical measures of
the object. Conversely, perceptual illusion will occur if the subject turns out to be
mistaken about absolute, extrinsically measured riser heights or transformations
of heights in the course of the experiment – which, if the experimental set-up
matches normal conditions for perception, is unlikely to negatively affect perception of an affordance. Hence, perceptual illusion and misperception of affordances
may coincide but are independent issues. In some cases, perceptual illusions may
even contribute to the correct perception of an affordance.
Although the preceding discussion suggests that the Empirical Strategy in some
respects elegantly complements the ecological view, two notable limitations to
this approach should be mentioned here: first, almost all examples discussed by
the authors are concerned with cases in which identical targets look different to
the perceiver under different conditions, whereas their point of departure, the
inverse projection problem, was identified by Berkeley as the problem that different targets, when placed in a certain relation to the perceiver, will look identical. Berkeleyan ambiguity does not imply that perception is at variance with the
measurable properties but that it becomes ambiguous by one retinal image being
in accordance with different instances of measurable properties. The Empirical