The nature of cognitive artefacts 143
to those where correspondence between visible and thermal radiation holds, thus
ruling out LED lights and hot iron, or the condition of convergence will not be
met, for want of the required correspondence.
Even if correspondence is achieved, the sets of domains D a and D n do not have
to be fully coextensive, as some artefact may be trained on a subset or superset
of the information that is available for direct perception. Moreover, convergence
does not imply that the artificially created signal is a signal of the natural signal
and, in this sense, second-order information. Such second-order information constitutes a specific subcase of informational relations that will obtain only if there
is a device or a natural mechanism with the function of signalling the presence of
natural signals. Language is a mechanism that is often used to accomplish that.
If the natural information in a given situation is of the intrinsic kind discussed
in Chapter 3 (p. 45), that intrinsic character will typically also be preserved for the
receiver, although it is conceivable that it may turn extrinsic. Where in the former
case receiver-dependent changes in the character of the affordance – for example
when body-scaled information is involved – will have to be mapped by whatever
artefactual rendering that can be accomplished, that affordance will be replaced
by a display of measurements of physical variables in the latter case. Convergence
can be achieved either way, depending on the functions of the artefact in question and the purposes of the receiver. After all, he might have to rely on intrinsic
information in some kind of situation (Is that item too bulky for me to carry?), and
on extrinsic information in another (Is that cupboard too wide to fit into the niche
in our dining room?).
The requirement of convergence will also hold for devices designed to aid or
sharpen natural perception. Enlargement, amplification and other effects have to
preserve the informational mappings their user would pick up if his sense organs
were capable of the degree of discrimination available to the device. Similarly, the
requirement of convergence will hold for any device designed to collect and display information to me that, given my biological constitution, I could not obtain
even with an exceedingly discriminating perceptual apparatus. A Geiger counter
is definitely supposed to track the domain of signals of nuclear radiation, which
is natural information that might become relevant to me under certain circumstances, despite being inaccessible to the natural perceptual equipment with which
I am endowed – hence the condition of hypothetical observability in the definition
of convergence. This example also provides an additional illustration of the difference between ecological and informational environments: Human beings and
most other species are reliant, in a strictly biological sense, on nuclear radiation
in their surroundings constantly remaining below a certain threshold. This is an
ecologically relevant condition, and hence will be part of the ecological environments of all those species. However, none of these species (and, to my knowledge,
no extant or historical species on Earth) has any sensory equipment for picking up
information on radiation levels. Hence, the uptake of signals of radiation levels is
not normally part of their informational environments.
A markedly different kind of situation emerges when we consider a simulated
environment: if it is designed with a realistic representation of some natural
to those where correspondence between visible and thermal radiation holds, thus
ruling out LED lights and hot iron, or the condition of convergence will not be
met, for want of the required correspondence.
Even if correspondence is achieved, the sets of domains D a and D n do not have
to be fully coextensive, as some artefact may be trained on a subset or superset
of the information that is available for direct perception. Moreover, convergence
does not imply that the artificially created signal is a signal of the natural signal
and, in this sense, second-order information. Such second-order information constitutes a specific subcase of informational relations that will obtain only if there
is a device or a natural mechanism with the function of signalling the presence of
natural signals. Language is a mechanism that is often used to accomplish that.
If the natural information in a given situation is of the intrinsic kind discussed
in Chapter 3 (p. 45), that intrinsic character will typically also be preserved for the
receiver, although it is conceivable that it may turn extrinsic. Where in the former
case receiver-dependent changes in the character of the affordance – for example
when body-scaled information is involved – will have to be mapped by whatever
artefactual rendering that can be accomplished, that affordance will be replaced
by a display of measurements of physical variables in the latter case. Convergence
can be achieved either way, depending on the functions of the artefact in question and the purposes of the receiver. After all, he might have to rely on intrinsic
information in some kind of situation (Is that item too bulky for me to carry?), and
on extrinsic information in another (Is that cupboard too wide to fit into the niche
in our dining room?).
The requirement of convergence will also hold for devices designed to aid or
sharpen natural perception. Enlargement, amplification and other effects have to
preserve the informational mappings their user would pick up if his sense organs
were capable of the degree of discrimination available to the device. Similarly, the
requirement of convergence will hold for any device designed to collect and display information to me that, given my biological constitution, I could not obtain
even with an exceedingly discriminating perceptual apparatus. A Geiger counter
is definitely supposed to track the domain of signals of nuclear radiation, which
is natural information that might become relevant to me under certain circumstances, despite being inaccessible to the natural perceptual equipment with which
I am endowed – hence the condition of hypothetical observability in the definition
of convergence. This example also provides an additional illustration of the difference between ecological and informational environments: Human beings and
most other species are reliant, in a strictly biological sense, on nuclear radiation
in their surroundings constantly remaining below a certain threshold. This is an
ecologically relevant condition, and hence will be part of the ecological environments of all those species. However, none of these species (and, to my knowledge,
no extant or historical species on Earth) has any sensory equipment for picking up
information on radiation levels. Hence, the uptake of signals of radiation levels is
not normally part of their informational environments.
A markedly different kind of situation emerges when we consider a simulated
environment: if it is designed with a realistic representation of some natural