“
Experiments like Lack's indicate considerable flexibility. The results should
not be overinterpreted as indicating tendencies to reflexive responding, as
Krebs and Dawkins (1984, p. 385) did: *that animals are susceptible to being
"tricked" by the crude dummies of ethologists ... makes it Ukely that natural
selection will favor similar exploitation by other animals'. In real events
stimuli are not isolated. Animals trying to 'trick' other individuals will
inevitably supply information from many sources, some of which may be
contradictory. Natural selection will have opposing effects, favoring exploita-
tion on the one hand and flexible coping procedures on the other.
Surely one of the hardest questions about flexibility is: to what extent can
signalers anticipate responses to their signaling and control that signaling to
influence the behavior of other individuals? Can they choose whether to
signal, and perhaps even what signal to use, based on expectations of the
responses they may eUcit? The parrot Alex can, with Enghsh words (e.g.
Pepperberg, 1990), but can birds have similar control over their species-
specific signaling? Evidence of such effects is inconclusive.
So-called 'audience effects' are sometimes cited as evidence of volitional
control of signaling; some may well be. In many if not most cases, however,
plausible alternative explanations have not been ruled out (Smith, 1990, pp.
211-214). For instance, does an individual ground squirrel (e.g. Spermophilus
beldingi) behave differently on detecting a predator if it is near or not near
its close relatives? It is more likely to utter trills when near close relatives
(Sherman, 1977), but is this because in such a situation it is more likely alertly
to monitor a predator, or because its audience elicits the calls? If the former,
then the audience effect does not influence signaling directly. The influence
is indirect, through an effect on the signaler's monitoring behavior. If a high
probability of staying attentive is part of the information that the vocalization
provides about the signaler's behavior, then the presence of relatives may
be simply a condition for the monitoring rather than a basis for a decision
to vocalize. The point is that we can not learn whether animals make decisions
about whether to signal until we have fully grasped the requisite conditions
(and thus the regular correlates) of signaling. If an individual retains freedom
with respect to those correlates, then its signaling can be modulated by
audience effects and the like. However, if the correlates are regular and thus
represented by the 'messages' of the signal, there is little opportunity to signal
electively. Signaling behavior is useful for cognitive research only when the
referents of signals have been carefully studied.
Surprisingly, a signaler can also be its own audience. An unanticipated
effect of an individual's vocal signahng on its own hormonal states was
discovered by Cheng (1992). The ovarian follicles of female Ring Doves,
Streptopelia risoria, who cannot coo because of experimental brain lesions,
severed syringial nerves or deflated air sacs do not mature. If the doves are
exposed to playback of their own previously recorded coos, the follicles do
mature. Cheng proposed that 'vocal self-stimulation' might also be important
in physiological responses to other signahng - a male passerine's singing, for
instance, or a human's crying, talking or singing in the dark. Any such
physiological changes would alter the bases for cognitive processing, and thus
for social responsiveness as well.
”
”
Russell P. Balda (Animal Cognition in Nature: The Convergence of Psychology and Biology in Laboratory and Field)