 |
 |
 |
|---|
 |
 |
 |
 |
 |
Antiphonal CallingAntiphonal calling occurs when a tamarin produces a long call immediately after hearing a long call produced by a conspecific. Given the stereotyped nature of this behavior, we are able to simulate an antiphonal exchange by broadcasting a pre-recorded or synthetic long call from a hidden speaker. To test, for example, which acoustic features of the long call are criticial to identifying it as a long call, we present subjects with manipulated exemplars of long calls. The logic here is that if subjects antiphonally call at the same rate to a normal call and a manipulated call, then the feature in the call that was changed was not functionally relevant for this behavior. If, however, the rate of antiphonal calling is significantly less for the manipulated call, then we conclude that the manipulated feature was important for antiphonal signal recognition.
To carry out these types of experiments, tamarins are placed in a testing apparatus in a sound attenuated booth. Behind, above and to the left of the tamarin is a concealed speaker. A set number of examplars of given stimulus types are broadcast from this speaker during each test trial. The stimulus types usually consist of one unmanipulated control version of the call and three manipulated versions. Stimul are broadcast approximately every 30 seconds. An antiphonal call occurs when subjects emit a long call within 5s of the stimulus broadcast. As an antiphonal response is a robust behavior, all trials are coded online. The video clip below provides an example of an antiphonal response. Links to PDFs of two papers using this experimental method are also given below.The units of perception in the antiphonal calling behavior of cotton-top tamarins (Saguinus oedipus): playback experiments with long calls
Amodal completion of acoustic signals by a nonhuman primate
Acoustic perception and the orienting responseFor the past four years, our lab has been using an orienting response to explore different aspects of perception and cognition in cotton-top tamarins. Our general approach is to place a single tamarin inside a test cage, located inside an acoustic chamber. Behind, above and to the left of the tamarin is a concealed speaker. Once the subject is looking down and away from the speaker, we initiate the playback of the test stimuli.
We videotape all trials and then code the response off-line by digitizing the trial and scoring it blind to condition; we achieve blind coding by marking the onset and offset of the playback and then using these markers to scroll through the trial, frame by frame, but with no sound feedback and no indication of condition. For each trial, we score one of four possible responses. A "Yes" response is scored if the subject turns and orients toward the speaker. A "No" response is scored if the subject fails to orient toward the speaker; head movement in a direction other than toward the speaker is not scored as a response. An "Ambiguous" response is scored if subject's movement is unclear (e.g., head movement up but not back toward the speaker). A "Bad" response is scored if the subject was moving, calling, or oriented toward the speaker at the time of the playback. In the video files below, we present examples of each of these four response measures. These videos are of an experiment that used a set of test stimuli in which speech syllables were used to represent numbers. Like many of our other experiments using this technique, we expect subjects to respond to test stimuli that are meaningfully different (perceptually or conceptually) from the stimuli presented during familiarization or habituation. The pdf link below provides one example of an experiment using the orienting response.
Cognitive Constraints on Cooperation
Altruism is unlikely to be common because the fitness costs of being nice provide a temptation to cheat--individuals that aren't nice are always better off that those that are nice. Yet, theory predicts that if individuals repeatedly take turns donating and receiving the niceness, this 'reciprocal altruism' (or reciprocity) can maintain niceness. However, because reciprocity involves current costs for future benefits, non-human animals may face cognitive constraints on their abilities to implement reciprocal strategies in cooperative contexts. Required cognitive abilities may include discriminating number, reducing discounting of future rewards, inhibitory control, cheater detection & punishment abilities, and recall of reputation. To test the role of cognitive constraints, I am evaluating cognitive capacities of cotton-top tamarins (Saguinus oedipus) and common marmosets (Callithrix jacchus). I will then test these species in cooperative situations that fall within and outside of their abilities to determine the effect of cognitive constraints on cooperation. In particular, I am assessing the tamarins' and marmosets' numerical and discounting abilities. These experiments are just beginning, but there appear to be interesting species differences.
Social and Vocal Responses to Predation
Music Perception
Comparative studies of music perception can contribute in at least two
ways to the debate on the evolution of music. First, we can carefully
control what if any exposure animals have to music, and thus rule out or
titrate the role of such exposure in various aspects of music perception.
Second, any human-like aspect of music perception found in an animal
cannot have evolved for the purpose of music, and thus if human abilities
are presumed to be homologous to the abilities found in animals, the human
abilities must be side effects of mechanisms evolved for other purposes.
We believe that comparative studies of music perception are poised to play
an important role in understanding the origins of music. Our research on
music perception involves several methodologies. Using an apparatus called
the skymaze, we can look for preferences for one kind of sound over
another in the cottontop tamarin. We also employ training methods, in
which an animal is played two sounds (e.g. melodies) and gets rewarded for
approaching one of the two. Once they have been trained to do this, we can
change the sounds (e.g. transpose the melodies) and see how they
generalize, to test whether musical relationships that are important for
humans are also heard by animals. Finally, we also use methods that have
been coopted from developmental work in human infants. In these
experiments, an animal is habituated to a particular class of sounds, and
then played a sequence of test sounds, some of which are from the
habituated class and some of which differ in some way. We measure whether
or not the novel test sounds produce an orienting response in the animal
as an indication of whether they hear them as different from the
habituated stimuli. All these methods can help us understand whether
animals share any of the musical perceptual abilities found in humans.