Wiley, R. H. 1994. Errors, exaggeration, and deception
in animal communication. In L. Real (ed.), Behavioral
mechanisms in ecology. University of Chicago Press, Chicago.
Chapter 7, pp. 157-189.
My contention in this chapter is that animals make mistakes and that this
simple but neglected circumstance has deep implications for the evolution
of animal communication. The possibility of error is explicitly
incorporated into mathematical theories of communication and choice, in
particular information theory, signal detection theory, and decision
theory. Although the basic concepts of these theories are neither new nor
complex, they have yet to be integrated into evolutionary theories of
An objective of this chapter, by expanding themes introduced
earlier (Wiley 1983), is to show that these basic concepts provide
explanations for the evolution of some fundamental features of
communication. In particular, they can provide sufficient conditions for
the evolution of exaggerated displays and for the evolutionary stability
of deception. This approach also indicates that some of the parameters
needed to understand the evolution of communication have been overlooked.
Errors by receivers, in the sense of evolutionarily inappropriate
responses to signals, are likely to occur during communication at long
range, in dense aggregations, or in the presence of deception. Such
errors have major implications for the evolution of communication. For
signals that cannot be completely distinguished by a receiver, there are
four possible results of any decision to respond or not to respond to a
signal: correct detection, missed detection, false alarm, and correct
In general the probabilities of these results cannot be independently
adjusted by a receiver; in particular, criteria for response that increase
the probability of correct detection also inevitably increase the
probability of false alarm. The inherent discriminability of the signal,
together with the probability and net advantage of each outcome,
determines the expected utility of any criterion for response.
Maximizing the expected utility of a receiver's criterion can lead to
adaptive gullibility (evolutionarily stable susceptibility to
deception) or to adaptive fastidiousness (low responsiveness to
signals). In the latter case, signalers must contend with increased
uncertainty in the responses to a signal. Signalers in such situations
can improve signaling efficiency by increasing the inherent
discriminability of signals, often by exaggeration. Female choice of
mates, as an example of adaptive fastidiousness, can result in such
exaggeration of signals.
In general, explanations for the evolution of any receivers' performance,
such as examples of female choice or evolutionarily stable susceptibility
to deception, require evaluation of the expected utility of the receivers'
criterion as well as the probability and inherent detectability of the
[ENTIRE ARTICLE (.pdf file)]