Seyfarth 1985; Mitchell and Thompson 1986; Marler et al. 1991; Hauser
1996, 1997). One reason for this difference in frequency is that it is more
difficult to catch someone who simply remains silent than someone who
actively falsifies information. From an empirical perspective, however,
researchers working on lies of omission are faced with a far more difficult
problem, for they are forced to interpret the absence of a response or
behavior. To show that an animal has committed a lie of omission, the same
conditions as discussed earlier must hold, plus an extra one. Thus, as for lies
of commission, there must be a reliable association between a call type and
some event internal or external to the organism; receivers should show a
reliable response to such calls, and there must be some behavioral flexibility on the part of callers such that call production is not rigidly triggered by
the relevant stimulus. The fourth condition for a lie of omission is that situations that reliably lead to calling in some circumstances fail to do so in a
different social context. Given this, studies of withholding information are
necessarily tied to studies of what Marler and colleagues (Marler et al.
1986a, 1991) have described as “audience effects,” analyses of the social
conditions mediating call production as opposed to suppression.
An important goal in studying lies of omission is to document the costs
associated with withholding information. If animals have mechanisms for
punishing those who are caught withholding information, this would indicate that lies of omission are perceived as such by conspecifics. Thus, if an
animal fails to announce the discovery of food and is then caught by another
group member, there should be a penalty imposed on the silent discoverer.
For example, a male caught withholding information about food might be
attacked by a dominant animal, denied shared food in the future, or rejected
by a female in future mating attempts. Although the first possibility is nearly
immediate, and thus easily observed, the latter requires long-term studies
of social reciprocation.
To flesh out these ideas, we return to some of the themes discussed above,
exploring when animals attempt to withhold information, how they benefit
from such lies, and the costs incurred if caught. We focus on cases where
animals withhold food and alarm calls because the absence of production
can be contrasted with the many contexts in which such calls are typically
given. To broaden the diversity of contexts, we also discuss calls given by
animals during mating and the contexts in which animals are selectively
silent about their sexual behavior.
When domestic chickens find food, they give a distinctive food call. When
domestic chickens detect an aerial predator, they give a different-sounding
alarm call than when they detect a ground predator. Such calls are functionally referential in that they appear to be highly correlated with particular contexts, and when receivers hear these calls, they respond in highly
specific ways, approaching food calls, looking up to aerial predator alarm
calls, and scanning the horizon in response to ground predator alarm calls.
Playbacks of these calls show that they reliably elicit different, adaptive
118
W.T. Fitch and M.D. Hauser
1996, 1997). One reason for this difference in frequency is that it is more
difficult to catch someone who simply remains silent than someone who
actively falsifies information. From an empirical perspective, however,
researchers working on lies of omission are faced with a far more difficult
problem, for they are forced to interpret the absence of a response or
behavior. To show that an animal has committed a lie of omission, the same
conditions as discussed earlier must hold, plus an extra one. Thus, as for lies
of commission, there must be a reliable association between a call type and
some event internal or external to the organism; receivers should show a
reliable response to such calls, and there must be some behavioral flexibility on the part of callers such that call production is not rigidly triggered by
the relevant stimulus. The fourth condition for a lie of omission is that situations that reliably lead to calling in some circumstances fail to do so in a
different social context. Given this, studies of withholding information are
necessarily tied to studies of what Marler and colleagues (Marler et al.
1986a, 1991) have described as “audience effects,” analyses of the social
conditions mediating call production as opposed to suppression.
An important goal in studying lies of omission is to document the costs
associated with withholding information. If animals have mechanisms for
punishing those who are caught withholding information, this would indicate that lies of omission are perceived as such by conspecifics. Thus, if an
animal fails to announce the discovery of food and is then caught by another
group member, there should be a penalty imposed on the silent discoverer.
For example, a male caught withholding information about food might be
attacked by a dominant animal, denied shared food in the future, or rejected
by a female in future mating attempts. Although the first possibility is nearly
immediate, and thus easily observed, the latter requires long-term studies
of social reciprocation.
To flesh out these ideas, we return to some of the themes discussed above,
exploring when animals attempt to withhold information, how they benefit
from such lies, and the costs incurred if caught. We focus on cases where
animals withhold food and alarm calls because the absence of production
can be contrasted with the many contexts in which such calls are typically
given. To broaden the diversity of contexts, we also discuss calls given by
animals during mating and the contexts in which animals are selectively
silent about their sexual behavior.
When domestic chickens find food, they give a distinctive food call. When
domestic chickens detect an aerial predator, they give a different-sounding
alarm call than when they detect a ground predator. Such calls are functionally referential in that they appear to be highly correlated with particular contexts, and when receivers hear these calls, they respond in highly
specific ways, approaching food calls, looking up to aerial predator alarm
calls, and scanning the horizon in response to ground predator alarm calls.
Playbacks of these calls show that they reliably elicit different, adaptive
118
W.T. Fitch and M.D. Hauser