discussed and, where possible, lay out a series of experiments that might
take our understanding further.
Lies of commission appear relatively often during interspecific interactions. Thus, in the nonvocal domain, we know of plovers that dupe their
predators by performing the injury-feigning display, predatory Photuris fireflies that mimic the mating-flash patterns of their congeners, thereby providing the mimic with a meal, snakes that play dead in order to avoid being
eaten, fish, birds, and mammals that enlarge some portion of their body in
order to look bigger, and insects and frogs that evolve coloration patterns
that resemble a sympatric but poisonous species (Lloyd 1984; Mitchell and
Thompson 1986; Burghardt 1991; Ristau 1991; Hauser 1996, 2000). In the
vocal domain, there are fewer examples, but Charles Munn’s (1986a, 1986b)
study of a mixed-species flock in Peru is perhaps one of the more compelling examples.
In a Peruvian rainforest, Munn noted that among the members of a
mixed-species flock of birds, some species appeared to be responsible for
finding food, whereas other species appeared to be responsible for alerting
the flock to danger.The fluidity with which these species interacted was spectacular, but perhaps more intriguing was the fact that the alarm-calling
species—the bluish-slate antshrike and the white-winged shrike tanager—
sometimes produced alarm calls when there were no predators in view.
These were not mistakes. Rather, the alarm calls were given almost exclusively when the antshrike or tanager was in direct competition with the foodfinding species over insect prey.As they approached the insect, the antshrike
and tanager produced an alarm call, causing the food-finding species to look
up and thereby forfeit its access to the insect. Surprisingly perhaps, this was
not a rare event. Out of 104 alarm calls recorded from the tanager, Munn
found that 55% were false alarms. Although this rate is high, we cannot conclude that it accurately represents the rate of deceptive alarm calls. It is
entirely possible that in some cases the bird detected an animal it considered
to be (or confused with) a predator when there was no predator at all. The
alarm call is a false alarm in the signal-detection sense—a perceptual error
that is likely to occur when sensitivities are set high, as are the costs of a miss
(i.e., of failing to detect the predator when it is present).
Is there any evidence that antshrikes and tanagers are capable of creating lies of commission? Let us return to our three conditions. First, the
antshrike and tanager alarm calls are often given during encounters with
predatory birds. Thus, there is a correlation between the signal and a specific context. Further, playback experiments of alarm calls given to actual
predators as opposed to virtual predators revealed no differences with
respect to the food-finding species’ responses; in both situations, they
looked up and fled. This shows that the false alarm call sounds like the true
alarm call and thus should be equally evocative. Although there may be
other acoustic cues that have been overlooked in the analysis, these
possible differences do not appear to be perceptually salient to listeners.
108
W.T. Fitch and M.D. Hauser
take our understanding further.
Lies of commission appear relatively often during interspecific interactions. Thus, in the nonvocal domain, we know of plovers that dupe their
predators by performing the injury-feigning display, predatory Photuris fireflies that mimic the mating-flash patterns of their congeners, thereby providing the mimic with a meal, snakes that play dead in order to avoid being
eaten, fish, birds, and mammals that enlarge some portion of their body in
order to look bigger, and insects and frogs that evolve coloration patterns
that resemble a sympatric but poisonous species (Lloyd 1984; Mitchell and
Thompson 1986; Burghardt 1991; Ristau 1991; Hauser 1996, 2000). In the
vocal domain, there are fewer examples, but Charles Munn’s (1986a, 1986b)
study of a mixed-species flock in Peru is perhaps one of the more compelling examples.
In a Peruvian rainforest, Munn noted that among the members of a
mixed-species flock of birds, some species appeared to be responsible for
finding food, whereas other species appeared to be responsible for alerting
the flock to danger.The fluidity with which these species interacted was spectacular, but perhaps more intriguing was the fact that the alarm-calling
species—the bluish-slate antshrike and the white-winged shrike tanager—
sometimes produced alarm calls when there were no predators in view.
These were not mistakes. Rather, the alarm calls were given almost exclusively when the antshrike or tanager was in direct competition with the foodfinding species over insect prey.As they approached the insect, the antshrike
and tanager produced an alarm call, causing the food-finding species to look
up and thereby forfeit its access to the insect. Surprisingly perhaps, this was
not a rare event. Out of 104 alarm calls recorded from the tanager, Munn
found that 55% were false alarms. Although this rate is high, we cannot conclude that it accurately represents the rate of deceptive alarm calls. It is
entirely possible that in some cases the bird detected an animal it considered
to be (or confused with) a predator when there was no predator at all. The
alarm call is a false alarm in the signal-detection sense—a perceptual error
that is likely to occur when sensitivities are set high, as are the costs of a miss
(i.e., of failing to detect the predator when it is present).
Is there any evidence that antshrikes and tanagers are capable of creating lies of commission? Let us return to our three conditions. First, the
antshrike and tanager alarm calls are often given during encounters with
predatory birds. Thus, there is a correlation between the signal and a specific context. Further, playback experiments of alarm calls given to actual
predators as opposed to virtual predators revealed no differences with
respect to the food-finding species’ responses; in both situations, they
looked up and fled. This shows that the false alarm call sounds like the true
alarm call and thus should be equally evocative. Although there may be
other acoustic cues that have been overlooked in the analysis, these
possible differences do not appear to be perceptually salient to listeners.
108
W.T. Fitch and M.D. Hauser