The extension of the extended mind 125
distinction introduced at the end of the preceding section: the evolution of vision
as exploiting the structure of local environments, and the evolution of language
as a key structuring feature of human interaction.
10
Although Clark and Chalmers
proceed to using language as their paradigm extension of the mind, it might be
worthwhile to also pay attention to more basic modes of environmental coupling.
The function of vision and language is certainly not one of instruments that could
be used or discarded by human beings at will. In the context of human evolution,
both are indispensable components of, and hence constitutive S to, human cognition. If an individual is deprived of either faculty, his or her cognitive functions
will be impaired to some extent, and some functions could not be accomplished
at all.
Still, there is a difference between vision and language, with their respective
modes of environmental coupling, in terms of the depth of their evolutionary roots
and the levels of cognition they feed into. Vision is not only the older and more
common phenomenon in the animal kingdom. It can also be plausibly argued to
be more directly tied to objects and structures in the environment, serving as a prerequisite for spatial orientation and object recognition for most animals, including humans. Language, in contrast, serves the coupling between human agents in
the first place and presupposes a number of pre-linguistic cognitive capabilities,
including mechanisms of orientation and object recognition. Moreover, whereas
the environmental coupling in visual perception, in terms of necessary constituents of the visual process, occurs between mechanisms of perception and invariant properties of objects in the environment, language will rightfully count as a
production mechanism in its own right.
On this view, the cognitively most foundational set of external entities to be coupled with is to be found in natural objects or processes in the environment. According to the ecological view of visual perception, as developed by James Jerome
Gibson (1979) and discussed in Chapter 3, features of an organism’s natural environment will provide the information to guide his activities, and will be available
to him reliably, directly, implicitly and automatically, fulfilling all four of Clark
and Chalmers’s earlier cited criteria. Visual perception, as we saw, crucially relies
on the perceiver’s interaction with his environment and the objects he encounters
therein. On the ecological view, inference or inner representations are largely irrelevant to perception, whereas none of these higher-order cognitive accomplishments
could emerge without the prior establishment of mechanisms of perceptual coupling. Alternatively, between such elementary direct guidance by perception and
fully fledged internal representations, one might find representations that detect
patterns in one’s environment and direct appropriate behaviours towards them
at the same instance. These elementary representations have been introduced as
“pushmi-pullyu representations” by Millikan (1995), and they figure as “actionoriented representations” in Clark (1997; see also Clark and Toribio 1994).
This image of directness has to be extended when moving from perception to
higher-order kinds of cognition. On the one hand, some of the guidance that has
hitherto been directly provided by perceptual couplings with the environment is
now delegated to internal resources, some but not all of which will assume the
distinction introduced at the end of the preceding section: the evolution of vision
as exploiting the structure of local environments, and the evolution of language
as a key structuring feature of human interaction.
10
Although Clark and Chalmers
proceed to using language as their paradigm extension of the mind, it might be
worthwhile to also pay attention to more basic modes of environmental coupling.
The function of vision and language is certainly not one of instruments that could
be used or discarded by human beings at will. In the context of human evolution,
both are indispensable components of, and hence constitutive S to, human cognition. If an individual is deprived of either faculty, his or her cognitive functions
will be impaired to some extent, and some functions could not be accomplished
at all.
Still, there is a difference between vision and language, with their respective
modes of environmental coupling, in terms of the depth of their evolutionary roots
and the levels of cognition they feed into. Vision is not only the older and more
common phenomenon in the animal kingdom. It can also be plausibly argued to
be more directly tied to objects and structures in the environment, serving as a prerequisite for spatial orientation and object recognition for most animals, including humans. Language, in contrast, serves the coupling between human agents in
the first place and presupposes a number of pre-linguistic cognitive capabilities,
including mechanisms of orientation and object recognition. Moreover, whereas
the environmental coupling in visual perception, in terms of necessary constituents of the visual process, occurs between mechanisms of perception and invariant properties of objects in the environment, language will rightfully count as a
production mechanism in its own right.
On this view, the cognitively most foundational set of external entities to be coupled with is to be found in natural objects or processes in the environment. According to the ecological view of visual perception, as developed by James Jerome
Gibson (1979) and discussed in Chapter 3, features of an organism’s natural environment will provide the information to guide his activities, and will be available
to him reliably, directly, implicitly and automatically, fulfilling all four of Clark
and Chalmers’s earlier cited criteria. Visual perception, as we saw, crucially relies
on the perceiver’s interaction with his environment and the objects he encounters
therein. On the ecological view, inference or inner representations are largely irrelevant to perception, whereas none of these higher-order cognitive accomplishments
could emerge without the prior establishment of mechanisms of perceptual coupling. Alternatively, between such elementary direct guidance by perception and
fully fledged internal representations, one might find representations that detect
patterns in one’s environment and direct appropriate behaviours towards them
at the same instance. These elementary representations have been introduced as
“pushmi-pullyu representations” by Millikan (1995), and they figure as “actionoriented representations” in Clark (1997; see also Clark and Toribio 1994).
This image of directness has to be extended when moving from perception to
higher-order kinds of cognition. On the one hand, some of the guidance that has
hitherto been directly provided by perceptual couplings with the environment is
now delegated to internal resources, some but not all of which will assume the