124 Environments of intelligence
processes. It is the additional criterion of the involvement of “non-derived content” recruited by the authors that shall suffice to identify the subset of intracranial processes that are properly cognitive. In introducing that second criterion,
the authors follow an argument in John Searle’s critique of AI (1980): although
linguistic items or other artefacts have the meanings they have and serve the functions they do because they are endowed, by stipulation or convention, with those
meanings and functions, our thoughts have their meanings and serve their functions in at least partial independence from any such external fiat. The nature of the
human mind bearing those thoughts is supposedly at least partly sufficient for the
presence of that content. That nature is not further explicated though.
On the aetiological view, in contrast, both artefacts of all sorts and our thoughts
have the meanings and functions they have because individual artefacts and mental tokens have been reproduced by some mechanism with sufficient frequency
because they mapped onto some world affair, in terms of being connected to an
appropriate set of behaviours towards it, reliably enough and according to a welldefined set of regularities. On this view, cognition basically is a specific activity
of relating to one’s environment, in which such mapping relations are used to produce adaptive behaviours. No principled difference exists between artefactual and
internal mechanisms that accomplish this feat, namely producing the appropriate tokens on the appropriate occasions, where the conditions of appropriateness
essentially depend on what in the environment the organism relies on. As we will
see in more detail in the next section, the human faculty of language is the primary
candidate for a properly cognitive trait that may both count as environmentally
extended and fit the constitutive S condition, and hence militates against either of
Adams and Aizawa’s criteria: its function and contents depend on the operation of
mechanisms that partly operate outside the human brain or body.
Entities and processes in the environment may contribute to cognitive abilities in constitutive S fashion in either of two ways, both of which include evolved
mechanisms of coupling: on the one hand, they might be part of the necessary
conditions for the realisation of the direct proper functions of a mechanism that
is otherwise based within the organism. The entities and processes in the environment need not be reproducible items with their own direct proper functions,
but their presence and use will be essential to the realisation of the functions
of the organism-based production mechanisms. On the other hand, there might
be mechanisms for the production of tokens of some type that are based in the
environment while being coordinated with organism-based mechanisms, so as
to jointly help him to produce behaviours appropriate to a given world affair in a
given situation. It is in the latter, more complex type of case where reproducible
artefacts will come to play a central role.
The art of coupling, basic and advanced
Clark and Chalmers only briefly and tentatively refer to relations that would count
as constitutive S on the present account (1998, 11). However, these references are
part of a brief evolutionary excursus and include two suggestions that match the
processes. It is the additional criterion of the involvement of “non-derived content” recruited by the authors that shall suffice to identify the subset of intracranial processes that are properly cognitive. In introducing that second criterion,
the authors follow an argument in John Searle’s critique of AI (1980): although
linguistic items or other artefacts have the meanings they have and serve the functions they do because they are endowed, by stipulation or convention, with those
meanings and functions, our thoughts have their meanings and serve their functions in at least partial independence from any such external fiat. The nature of the
human mind bearing those thoughts is supposedly at least partly sufficient for the
presence of that content. That nature is not further explicated though.
On the aetiological view, in contrast, both artefacts of all sorts and our thoughts
have the meanings and functions they have because individual artefacts and mental tokens have been reproduced by some mechanism with sufficient frequency
because they mapped onto some world affair, in terms of being connected to an
appropriate set of behaviours towards it, reliably enough and according to a welldefined set of regularities. On this view, cognition basically is a specific activity
of relating to one’s environment, in which such mapping relations are used to produce adaptive behaviours. No principled difference exists between artefactual and
internal mechanisms that accomplish this feat, namely producing the appropriate tokens on the appropriate occasions, where the conditions of appropriateness
essentially depend on what in the environment the organism relies on. As we will
see in more detail in the next section, the human faculty of language is the primary
candidate for a properly cognitive trait that may both count as environmentally
extended and fit the constitutive S condition, and hence militates against either of
Adams and Aizawa’s criteria: its function and contents depend on the operation of
mechanisms that partly operate outside the human brain or body.
Entities and processes in the environment may contribute to cognitive abilities in constitutive S fashion in either of two ways, both of which include evolved
mechanisms of coupling: on the one hand, they might be part of the necessary
conditions for the realisation of the direct proper functions of a mechanism that
is otherwise based within the organism. The entities and processes in the environment need not be reproducible items with their own direct proper functions,
but their presence and use will be essential to the realisation of the functions
of the organism-based production mechanisms. On the other hand, there might
be mechanisms for the production of tokens of some type that are based in the
environment while being coordinated with organism-based mechanisms, so as
to jointly help him to produce behaviours appropriate to a given world affair in a
given situation. It is in the latter, more complex type of case where reproducible
artefacts will come to play a central role.
The art of coupling, basic and advanced
Clark and Chalmers only briefly and tentatively refer to relations that would count
as constitutive S on the present account (1998, 11). However, these references are
part of a brief evolutionary excursus and include two suggestions that match the