Varieties of perception 39
Information specifies affordances: the ecological view
Animals and humans communicate with cries, gestures, speech, pictures, writing,
and television, but we cannot hope to understand perception in terms of these
channels; it is quite the other way round.
(Gibson 1979, 242)
If Marr concludes from the observation that “vision is the process of discovering from images what is present in the world” that vision is, “first and foremost,
an information-processing task” (Marr 2010, 3), and hence a task that involves
various levels of symbol manipulation and inference in the very process of discovering what is present in the world, we find very different conclusions being
drawn from the same basic observation, and from one seemingly shared premise,
in Gibson (1979): that an inquiry into visual perception will have to concern the
uptake of information and the channels through which it arrives at the visual system. However, where Marr focuses on the complexities of processing that information as the key to his analysis, Gibson quite flatly denies that perception is a
matter of information processing in the first place. Instead, he maintains, one will
understand vision only if and only when starting from the relation of the perceiving organism to his environment and the information provided by that environment. Informational relations are ubiquitous and objective, and can be directly
acted upon without stages of processing and inference intervening: “Information
as here conceived is not transmitted or conveyed, does not consist of signals and
messages, and does not entail a sender and a receiver” (Gibson 1979, 57).
Even though Gibson provides only a rudimentary positive definition of his concept of information when maintaining that it “refers to specification of the observer’s environment, not to specification of the observer’s receptors or sense organs”
(Gibson 1979, 242), it is obvious enough that no concept of information could
be further removed from the mathematical theory of information, and from its
presumption that intentional agents are present at the source and receiving sides
of information, and that they are in charge of the meaning of whatever is communicated.
1
If information is not symbolic in the first place, and if it is not processed
in perception (whatever else is done with it in the process of perception), and if it
is an objective affair that is present in the environment even in absence of sender
or receiver, it will not be affected by issues of its proper grounding. Natural information is grounded as such, in the very environments as they are inhabited by
organisms, and it can thus become the root of all perception and, by implication,
all cognition and empirical knowledge.
It is unsurprising that Gibson’s account of perception, unlike Marr’s, did not exert
much influence on cognitive science, let alone computational neuroscience, while
having laid the foundation to an entire school of ecological, non-computational
psychology of perception. Gibson’s approach has also become a source of inspiration for naturalistic philosophers such as Dretske and Millikan, and it has exerted a
strong, though mediated, influence on many domains of designing artefacts, from
Information specifies affordances: the ecological view
Animals and humans communicate with cries, gestures, speech, pictures, writing,
and television, but we cannot hope to understand perception in terms of these
channels; it is quite the other way round.
(Gibson 1979, 242)
If Marr concludes from the observation that “vision is the process of discovering from images what is present in the world” that vision is, “first and foremost,
an information-processing task” (Marr 2010, 3), and hence a task that involves
various levels of symbol manipulation and inference in the very process of discovering what is present in the world, we find very different conclusions being
drawn from the same basic observation, and from one seemingly shared premise,
in Gibson (1979): that an inquiry into visual perception will have to concern the
uptake of information and the channels through which it arrives at the visual system. However, where Marr focuses on the complexities of processing that information as the key to his analysis, Gibson quite flatly denies that perception is a
matter of information processing in the first place. Instead, he maintains, one will
understand vision only if and only when starting from the relation of the perceiving organism to his environment and the information provided by that environment. Informational relations are ubiquitous and objective, and can be directly
acted upon without stages of processing and inference intervening: “Information
as here conceived is not transmitted or conveyed, does not consist of signals and
messages, and does not entail a sender and a receiver” (Gibson 1979, 57).
Even though Gibson provides only a rudimentary positive definition of his concept of information when maintaining that it “refers to specification of the observer’s environment, not to specification of the observer’s receptors or sense organs”
(Gibson 1979, 242), it is obvious enough that no concept of information could
be further removed from the mathematical theory of information, and from its
presumption that intentional agents are present at the source and receiving sides
of information, and that they are in charge of the meaning of whatever is communicated.
1
If information is not symbolic in the first place, and if it is not processed
in perception (whatever else is done with it in the process of perception), and if it
is an objective affair that is present in the environment even in absence of sender
or receiver, it will not be affected by issues of its proper grounding. Natural information is grounded as such, in the very environments as they are inhabited by
organisms, and it can thus become the root of all perception and, by implication,
all cognition and empirical knowledge.
It is unsurprising that Gibson’s account of perception, unlike Marr’s, did not exert
much influence on cognitive science, let alone computational neuroscience, while
having laid the foundation to an entire school of ecological, non-computational
psychology of perception. Gibson’s approach has also become a source of inspiration for naturalistic philosophers such as Dretske and Millikan, and it has exerted a
strong, though mediated, influence on many domains of designing artefacts, from