honest signals in birds and mammals. In Section 3, we turn to perceptual
and cognitive mechanisms in communication, showing how perceptual constraints can leave perceivers vulnerable to “sensory exploitation” and how
learning mechanisms can allow perceivers to short-circuit such manipulation. We conclude that an understanding of constraints in signal evolution,
together with a dynamic view of evolution, can provide deeper insights into
animal communication systems.We hope to show that, when integrated with
the comparative evolutionary perspective advocated above, the study of
physical and phylogenetic constraints can provide rich insights into the evolution of acoustic communication systems.
1.2. Distinctions and Definitions
There are many important distinctions to be made in discussions of the evolution of honest signals. Here, we will focus on three: arbitrary versus direct
signals, lies of omission versus commission, and external- versus selfreference. These distinctions do not represent hard and fast categories that
exist in nature but rather lay out a continuum on which any given signal
can be located. A signal might be intermediate or extreme along any of
these continua. Thus, these distinctions must be seen as aids to communication and understanding and not claims about dichotomies in the world.
1.2.1. Arbitrary and Direct Signals
One continuum along which we can analyze acoustic signals concerns the
degree to which the mapping between signal and referent is arbitrary. In
human language, word meaning is generally not tied in any direct way to
the acoustic structure of that sound. Thus, except for a small number of onomatopoetic or sound-symbolic words, one cannot guess the meaning of a
word in a foreign language based only on its sound. It is often assumed that
the information in animal calls is similarly arbitrary. There is, however, significant evidence that at least some of the information in animal calls is
“direct,” with a nonarbitrary mapping between sound and meaning. The
most obvious example is the association between low frequency and large
size. Because only large objects can produce and radiate low-frequency
signals effectively, there is a natural mapping between size and frequency,
particularly large body size and low-frequency calls (e.g., elephants and
infrasound). We will argue that physical facts such as the size/frequency
relationship have played a pervasive role in the evolution of communication systems and that a solid understanding of production mechanisms is
necessary to uncover and understand such relationships. In contrast, the difference between the vervet monkey’s leopard and snake alarm calls probably has little or nothing to do with the sounds themselves (Cheney and
Seyfarth 1990). Rather, these calls are more like human words, providing
an example of an arbitrary sound/meaning pairing that has been conventionalized over evolutionary time.
70
W.T. Fitch and M.D. Hauser
and cognitive mechanisms in communication, showing how perceptual constraints can leave perceivers vulnerable to “sensory exploitation” and how
learning mechanisms can allow perceivers to short-circuit such manipulation. We conclude that an understanding of constraints in signal evolution,
together with a dynamic view of evolution, can provide deeper insights into
animal communication systems.We hope to show that, when integrated with
the comparative evolutionary perspective advocated above, the study of
physical and phylogenetic constraints can provide rich insights into the evolution of acoustic communication systems.
1.2. Distinctions and Definitions
There are many important distinctions to be made in discussions of the evolution of honest signals. Here, we will focus on three: arbitrary versus direct
signals, lies of omission versus commission, and external- versus selfreference. These distinctions do not represent hard and fast categories that
exist in nature but rather lay out a continuum on which any given signal
can be located. A signal might be intermediate or extreme along any of
these continua. Thus, these distinctions must be seen as aids to communication and understanding and not claims about dichotomies in the world.
1.2.1. Arbitrary and Direct Signals
One continuum along which we can analyze acoustic signals concerns the
degree to which the mapping between signal and referent is arbitrary. In
human language, word meaning is generally not tied in any direct way to
the acoustic structure of that sound. Thus, except for a small number of onomatopoetic or sound-symbolic words, one cannot guess the meaning of a
word in a foreign language based only on its sound. It is often assumed that
the information in animal calls is similarly arbitrary. There is, however, significant evidence that at least some of the information in animal calls is
“direct,” with a nonarbitrary mapping between sound and meaning. The
most obvious example is the association between low frequency and large
size. Because only large objects can produce and radiate low-frequency
signals effectively, there is a natural mapping between size and frequency,
particularly large body size and low-frequency calls (e.g., elephants and
infrasound). We will argue that physical facts such as the size/frequency
relationship have played a pervasive role in the evolution of communication systems and that a solid understanding of production mechanisms is
necessary to uncover and understand such relationships. In contrast, the difference between the vervet monkey’s leopard and snake alarm calls probably has little or nothing to do with the sounds themselves (Cheney and
Seyfarth 1990). Rather, these calls are more like human words, providing
an example of an arbitrary sound/meaning pairing that has been conventionalized over evolutionary time.
70
W.T. Fitch and M.D. Hauser