What is an informational environment? 103
to very different sources by different regularities, and are indistinguishable from
each other only to the extent that R has no other cues available that would reveal
the difference to him.
Keeping cases of deceptive signalling apart from the genuine constituents of
informational environments has a normative aspect: it honours the environmental
grounding of the principled possibility for R to distinguish between things that
are different in kind even if and when signals emanating from them are indistinguishable at first sight. If signals that are indistinguishable for R when using his
normal means of information uptake indeed have very different aetiologies, there
will be a point at which the variance in aetiologies is detectable in principle. In
all realistically conceived cases, sparing Descartes’s malicious demon, brains in
a vat and their kin, there will be at least a chance for that variance to also become
detectable in practice – however remote that chance might be. It will be difficult to
conceive of a situation in which the variant aetiologies will informationally manifest themselves as being identical on all levels and in all possible ways. (Could
Twin Water, or the designer of Twin Water, not only pretend that it has all the phenomenal properties of water but also maintain a pretence of that substance being
H 2 O in the most sophisticated of laboratories, except that it is not H 2 O?) Hence,
although deception is common in many animals’ environments, and although the
world may be hostile to its inhabitants in many ways, such hostility should not
be considered the norm. In a perfectly treacherous world, there would neither be
a possibility nor a use for getting things right. Conversely, in a perfectly benign
world, there would be no need for an ability to tell right from wrong, because,
apart from internally caused malfunction, there would be no risk of getting things
wrong to begin with. The possibility and the usefulness of getting things right
with some reliability should provide the normative grounding for steering through
a world that often tends to wrong-foot us.
How informational environments change
The possibility of an informational environment to become impoverished in the
way previously described in IE-1c is a condition that should not be confused with
either of two other issues concerning the shape of informational environments: on
the one hand, an impoverished informational environment is not equivalent to an
informationally more primitive informational environment, as it would pertain to
an organism who relies on fewer or less interconnected cues. We would then be
comparing the informational environments of differently constituted organisms,
not different states of one organism’s informational environment. On the other
hand, an impoverished informational environment should not be equated with an
environment in which the regularities that connect source and signal have altered.
In such cases, as I argued in the preceding section, something more fundamental
in the organism’s environment has changed. We will then have to consider the
possibility that his ecological or external environment have been transformed in
a significant way, with secondary effects on the informational conditions relevant
to him.
to very different sources by different regularities, and are indistinguishable from
each other only to the extent that R has no other cues available that would reveal
the difference to him.
Keeping cases of deceptive signalling apart from the genuine constituents of
informational environments has a normative aspect: it honours the environmental
grounding of the principled possibility for R to distinguish between things that
are different in kind even if and when signals emanating from them are indistinguishable at first sight. If signals that are indistinguishable for R when using his
normal means of information uptake indeed have very different aetiologies, there
will be a point at which the variance in aetiologies is detectable in principle. In
all realistically conceived cases, sparing Descartes’s malicious demon, brains in
a vat and their kin, there will be at least a chance for that variance to also become
detectable in practice – however remote that chance might be. It will be difficult to
conceive of a situation in which the variant aetiologies will informationally manifest themselves as being identical on all levels and in all possible ways. (Could
Twin Water, or the designer of Twin Water, not only pretend that it has all the phenomenal properties of water but also maintain a pretence of that substance being
H 2 O in the most sophisticated of laboratories, except that it is not H 2 O?) Hence,
although deception is common in many animals’ environments, and although the
world may be hostile to its inhabitants in many ways, such hostility should not
be considered the norm. In a perfectly treacherous world, there would neither be
a possibility nor a use for getting things right. Conversely, in a perfectly benign
world, there would be no need for an ability to tell right from wrong, because,
apart from internally caused malfunction, there would be no risk of getting things
wrong to begin with. The possibility and the usefulness of getting things right
with some reliability should provide the normative grounding for steering through
a world that often tends to wrong-foot us.
How informational environments change
The possibility of an informational environment to become impoverished in the
way previously described in IE-1c is a condition that should not be confused with
either of two other issues concerning the shape of informational environments: on
the one hand, an impoverished informational environment is not equivalent to an
informationally more primitive informational environment, as it would pertain to
an organism who relies on fewer or less interconnected cues. We would then be
comparing the informational environments of differently constituted organisms,
not different states of one organism’s informational environment. On the other
hand, an impoverished informational environment should not be equated with an
environment in which the regularities that connect source and signal have altered.
In such cases, as I argued in the preceding section, something more fundamental
in the organism’s environment has changed. We will then have to consider the
possibility that his ecological or external environment have been transformed in
a significant way, with secondary effects on the informational conditions relevant
to him.