What is an informational environment? 101
It is in light of these considerations on the normative nature of natural and
artificial signals that some of the clauses in the earlier definitions should be read –
which may have the additional effect of making the notion of the receiverspecificity of informational environments appear at least a bit less trivial than
one might first believe. To begin with, the condition in IE-1a that, in order to
be a constituent of an informational environment, information has to serve an
adaptive function for its receivers and, by analogy, serve a function not only to
an individual but also, potentially and in principle, for all members of his population, is designed to keep out arbitrarily formed signalling relations. Without this
double specification, signals placed in an individual’s environment would have to
be counted as part of his informational environment that are designed and used to
serve contingent, idiosyncratic and ad hoc purposes either for that same individual
while not possibly doing so for members of his population, or that serve adaptive
functions only for members of other populations.
If such signals were admitted, the first type of case would represent one aspect
of Lewontian environmental construction that has been demonstrated to obscure
the very distinction between just any effect that an organism’s behaviour has on
his environment and those effects which have a history of being useful to members of a population and hence acquired a proper function in Millikan’s terms:
there might be signals that are useful for one or a few individuals but have the
intended or unintended effect of undermining conditions for the population of
which these individuals are members. Speaking in game-theoretical terms, an
informational free-rider may maximise his individual benefits, but his selfish
ways cannot provide a basis for a dominant strategy of acting within a population,
as the reproductive chances of its members in general would be jeopardised at
some point – especially if coordinated activities are relevant to that population.
Even where competitive or otherwise uncooperative relations between individuals within one population are concerned, universal deception and unreliability
of information could not be the norm. There would be no use in even trying to
pick up information then. For this reason, I refer to cooperation and coordination
in IE-2: displaying signs of putative dominance in a group’s pecking order or in
fending off another group is not precisely cooperative but still serves coordinated
actions within a population, nor is it deceptive. Coordination may occasionally
be served by deception too, unless it becomes self-defeating on the group level.
In the second type of case, however, there is a perfectly adaptive function for
signals that are emitted by some individual or individuals within a population in
order to deceive a potential prey or predator, host or parasite, as indicated in IE-2c.
The individuals to be thus deceived typically are not members of the signalling
individuals’ own population, and the signals in question are produced precisely in
order not to provide information. Here, deception will be the norm when viewed
from the sender population’s side, whereas there certainly is no adaptive function
of those deceptive signals when viewed from the receiver population’s side.
Consequently, the informational environment of a receiver R does, strictly
speaking, not comprise signals that are intended to deceive R in the same way as
it does properly mapping signals, nor does it equally comprise signals that are too
It is in light of these considerations on the normative nature of natural and
artificial signals that some of the clauses in the earlier definitions should be read –
which may have the additional effect of making the notion of the receiverspecificity of informational environments appear at least a bit less trivial than
one might first believe. To begin with, the condition in IE-1a that, in order to
be a constituent of an informational environment, information has to serve an
adaptive function for its receivers and, by analogy, serve a function not only to
an individual but also, potentially and in principle, for all members of his population, is designed to keep out arbitrarily formed signalling relations. Without this
double specification, signals placed in an individual’s environment would have to
be counted as part of his informational environment that are designed and used to
serve contingent, idiosyncratic and ad hoc purposes either for that same individual
while not possibly doing so for members of his population, or that serve adaptive
functions only for members of other populations.
If such signals were admitted, the first type of case would represent one aspect
of Lewontian environmental construction that has been demonstrated to obscure
the very distinction between just any effect that an organism’s behaviour has on
his environment and those effects which have a history of being useful to members of a population and hence acquired a proper function in Millikan’s terms:
there might be signals that are useful for one or a few individuals but have the
intended or unintended effect of undermining conditions for the population of
which these individuals are members. Speaking in game-theoretical terms, an
informational free-rider may maximise his individual benefits, but his selfish
ways cannot provide a basis for a dominant strategy of acting within a population,
as the reproductive chances of its members in general would be jeopardised at
some point – especially if coordinated activities are relevant to that population.
Even where competitive or otherwise uncooperative relations between individuals within one population are concerned, universal deception and unreliability
of information could not be the norm. There would be no use in even trying to
pick up information then. For this reason, I refer to cooperation and coordination
in IE-2: displaying signs of putative dominance in a group’s pecking order or in
fending off another group is not precisely cooperative but still serves coordinated
actions within a population, nor is it deceptive. Coordination may occasionally
be served by deception too, unless it becomes self-defeating on the group level.
In the second type of case, however, there is a perfectly adaptive function for
signals that are emitted by some individual or individuals within a population in
order to deceive a potential prey or predator, host or parasite, as indicated in IE-2c.
The individuals to be thus deceived typically are not members of the signalling
individuals’ own population, and the signals in question are produced precisely in
order not to provide information. Here, deception will be the norm when viewed
from the sender population’s side, whereas there certainly is no adaptive function
of those deceptive signals when viewed from the receiver population’s side.
Consequently, the informational environment of a receiver R does, strictly
speaking, not comprise signals that are intended to deceive R in the same way as
it does properly mapping signals, nor does it equally comprise signals that are too