42 Informational environments
We may find Gibson barking up a largely uninhabited tree of naïve metaphysics when he thus insinuates that other psychologists and philosophers of perception actually believe in a “copy” theory of perception, but his underlying point
is quite remarkable: that perception does not involve representation. It neither
builds on image-like representations, nor does it rely on generating an internal
model that would represent the relevant variables that determine the appearance of
an object, or on classifying objects or on otherwise abstracting from their concrete
properties – all the things the computational approach to visual perception cannot
do without.
Above all, the organism’s acts of perception imply the direct uptake of information on what can be done with a perceived object. Throughout all instances of
being perceived, objects of all kinds, including other organisms, to use Gibson’s
most used (and often misused) concept, provide affordances to the perceiving
organism. An affordance is what some object offers to be done with it, in the particular way in which it is related to the perceiving organism at a given time, under
a given set of conditions. Accordingly, an object’s affordance is always described
in terms of possible actions of, or interactions with, the perceiving organism, such
as standing, sitting, climbing up, jumping over, falling off. For example a fruit
may afford activities such as being eaten, burrowing, being picked and used as a
missile or poisoning – or simply nothing at all. Affordances will vary among different animals for one and the same object or even the same organism on different
occasions. The activities an object affords to an organism constitute its specific
“values” or “meanings”, which can be directly perceived by the organism (Gibson
1979, 127).
It would be unfortunate though to simply equate affordances with values and
meanings. These are normatively highly charged terms that are likely to result in
philosophical entanglements – which to avoid was one of the purposes behind
the introduction of the concept of affordances (see Gibson 1966, 285). However,
there are at least two related points at which that concept appears so vaguely
circumscribed in Gibson’s famous definition (1979, 127, 129) as to invite such
entanglements: there, he says that “affordances of the environment are what it
offers the animal” (emphasis in original), but he also says that an affordance is
“neither an objective property nor a subjective property; or it is both if you like”
and that it is “equally a fact of the environment and a fact of behaviour”. So, first,
are affordances properties of objects, or are they better defined in terms of relations and situations? On an ecological view, objects and their properties might,
and should, not be considered in detachment from the environments in which
they are encountered, so Gibson’s reference to objects and their properties should
be read in the light of this basic commitment. Second, are affordances entirely
provided by the environment, or are they (partly) subjectively defined? If they are
something that can be detected by an organism, they are supposed to be contained
in the environment, but what is actually afforded in the act of perception can be
determined only by reference to the constitution and abilities of the perceiver.
Although the second apparent vagueness can be demonstrated to have a systematic purpose behind it, the first might be more difficult to parse.
We may find Gibson barking up a largely uninhabited tree of naïve metaphysics when he thus insinuates that other psychologists and philosophers of perception actually believe in a “copy” theory of perception, but his underlying point
is quite remarkable: that perception does not involve representation. It neither
builds on image-like representations, nor does it rely on generating an internal
model that would represent the relevant variables that determine the appearance of
an object, or on classifying objects or on otherwise abstracting from their concrete
properties – all the things the computational approach to visual perception cannot
do without.
Above all, the organism’s acts of perception imply the direct uptake of information on what can be done with a perceived object. Throughout all instances of
being perceived, objects of all kinds, including other organisms, to use Gibson’s
most used (and often misused) concept, provide affordances to the perceiving
organism. An affordance is what some object offers to be done with it, in the particular way in which it is related to the perceiving organism at a given time, under
a given set of conditions. Accordingly, an object’s affordance is always described
in terms of possible actions of, or interactions with, the perceiving organism, such
as standing, sitting, climbing up, jumping over, falling off. For example a fruit
may afford activities such as being eaten, burrowing, being picked and used as a
missile or poisoning – or simply nothing at all. Affordances will vary among different animals for one and the same object or even the same organism on different
occasions. The activities an object affords to an organism constitute its specific
“values” or “meanings”, which can be directly perceived by the organism (Gibson
1979, 127).
It would be unfortunate though to simply equate affordances with values and
meanings. These are normatively highly charged terms that are likely to result in
philosophical entanglements – which to avoid was one of the purposes behind
the introduction of the concept of affordances (see Gibson 1966, 285). However,
there are at least two related points at which that concept appears so vaguely
circumscribed in Gibson’s famous definition (1979, 127, 129) as to invite such
entanglements: there, he says that “affordances of the environment are what it
offers the animal” (emphasis in original), but he also says that an affordance is
“neither an objective property nor a subjective property; or it is both if you like”
and that it is “equally a fact of the environment and a fact of behaviour”. So, first,
are affordances properties of objects, or are they better defined in terms of relations and situations? On an ecological view, objects and their properties might,
and should, not be considered in detachment from the environments in which
they are encountered, so Gibson’s reference to objects and their properties should
be read in the light of this basic commitment. Second, are affordances entirely
provided by the environment, or are they (partly) subjectively defined? If they are
something that can be detected by an organism, they are supposed to be contained
in the environment, but what is actually afforded in the act of perception can be
determined only by reference to the constitution and abilities of the perceiver.
Although the second apparent vagueness can be demonstrated to have a systematic purpose behind it, the first might be more difficult to parse.