Varieties of perception 41
processing. However, there are perfectly conceivable situations in which there
is an abundance of stimuli that does not convey any information, such as in a
brightly lit room filled with dense fog (Gibson 1979, 52–55; see the discussion in
Chemero 2003b) – or, to use a more commonplace example, occurrences of socalled whiteout conditions, which can be dangerous to pilots, motorists or mountaineers precisely for the combination of a strong stimulus with the utter lack of
visual information. Stimulus and information are in this sense detached, whereas
stimulus and perception are not.
First and foremost, perception, on the Gibsonian view, is to be considered
an activity that is intrinsically tied to other activities of an organism, and that
depends on his general constitution and abilities – his physiology, his body scale,
the behaviours he is capable and the resources he is in need of – on the one hand,
and on his current position and movements within his environment on the other.
Perception consists in the “pickup” of information from the “ambient energies”
surrounding the organism. In relation to his position and movements, these energies form the “optic array” for perception (Gibson 1979, Chapter 5). To accomplish the task of information pickup, visual and non-visual information about the
position, orientation and movement of the perceiving organism is included in the
act of perception (Gibson 1979, 115–120).
In this fashion, information is actively retrieved from the environment, so as
to detect patterns of persistence and change therein and track the “invariants”
of some object. Invariants are to be understood in analogy to the mathematical
meaning of the term (Gibson 1971, 30; 1973), as those properties of an object
which remain unchanged when a set of rule-governed transformations is applied
to it. For example the length ratios of a geometric figure remain unchanged when
it is scaled up and down proportionally. These ratios, but not the absolute measure
of the figure’s elements, are the object’s invariants. In the context of perception,
the transformations will encompass all naturally occurring changes in the conditions of perception, and the invariants will be what remains unchanged, as viewed
in relation to the transformations of these conditions. In Gibson (Gibson 1979,
45), we find a non-comprehensive list of candidate invariants, which comprises
“alignment or straightness [. . .] as against bentness or curvature; perpendicularity
or rectangularity; parallelity as against convergence; intersections; closures and
symmetries”. Citing Gibson’s own example (1979, 13), a solid substance is rather
persistent in shape, so shape is an invariant in the perception of all solid objects,
but not in the perception of any less-than-solid object, for which density or volume are likely to count as invariants.
The acts of retrieving information from the environment and hence tracking the
invariants of some object or event do not involve the “replication” or “copying”
of that object or event in the ambient light, as though some replica of the object
were picked up in perception (Gibson 1979, 102f ). The tracking of invariants is
much less concerned with detecting similarities between an image and an object
than with guiding the perceiving organism’s activities towards that object. That
guidance has to be accomplished throughout a multitude of transformations of
conditions within the environment.
processing. However, there are perfectly conceivable situations in which there
is an abundance of stimuli that does not convey any information, such as in a
brightly lit room filled with dense fog (Gibson 1979, 52–55; see the discussion in
Chemero 2003b) – or, to use a more commonplace example, occurrences of socalled whiteout conditions, which can be dangerous to pilots, motorists or mountaineers precisely for the combination of a strong stimulus with the utter lack of
visual information. Stimulus and information are in this sense detached, whereas
stimulus and perception are not.
First and foremost, perception, on the Gibsonian view, is to be considered
an activity that is intrinsically tied to other activities of an organism, and that
depends on his general constitution and abilities – his physiology, his body scale,
the behaviours he is capable and the resources he is in need of – on the one hand,
and on his current position and movements within his environment on the other.
Perception consists in the “pickup” of information from the “ambient energies”
surrounding the organism. In relation to his position and movements, these energies form the “optic array” for perception (Gibson 1979, Chapter 5). To accomplish the task of information pickup, visual and non-visual information about the
position, orientation and movement of the perceiving organism is included in the
act of perception (Gibson 1979, 115–120).
In this fashion, information is actively retrieved from the environment, so as
to detect patterns of persistence and change therein and track the “invariants”
of some object. Invariants are to be understood in analogy to the mathematical
meaning of the term (Gibson 1971, 30; 1973), as those properties of an object
which remain unchanged when a set of rule-governed transformations is applied
to it. For example the length ratios of a geometric figure remain unchanged when
it is scaled up and down proportionally. These ratios, but not the absolute measure
of the figure’s elements, are the object’s invariants. In the context of perception,
the transformations will encompass all naturally occurring changes in the conditions of perception, and the invariants will be what remains unchanged, as viewed
in relation to the transformations of these conditions. In Gibson (Gibson 1979,
45), we find a non-comprehensive list of candidate invariants, which comprises
“alignment or straightness [. . .] as against bentness or curvature; perpendicularity
or rectangularity; parallelity as against convergence; intersections; closures and
symmetries”. Citing Gibson’s own example (1979, 13), a solid substance is rather
persistent in shape, so shape is an invariant in the perception of all solid objects,
but not in the perception of any less-than-solid object, for which density or volume are likely to count as invariants.
The acts of retrieving information from the environment and hence tracking the
invariants of some object or event do not involve the “replication” or “copying”
of that object or event in the ambient light, as though some replica of the object
were picked up in perception (Gibson 1979, 102f ). The tracking of invariants is
much less concerned with detecting similarities between an image and an object
than with guiding the perceiving organism’s activities towards that object. That
guidance has to be accomplished throughout a multitude of transformations of
conditions within the environment.