Referential Signals in Putty-nosed Monkeys

There is an extensive and highly contested literature surrounding referential communication among non-human animals. Referential signaling systems contain specific, discrete signals that are not dependent on context. The classic example of referential signaling is vervet monkeys alarm calls. Vervet monkeys have three different calls that are used to warn against three predator types. These calls produce specific behavioral responses that are appropriate for evading each type of predator. When responding to the calls, the monkeys act as if there is a predator, regardless of the context of whether they actually see a predator, leading some researchers to conclude that these signals are referential. Examples of referential signals are pretty rare and most alarm calls are thought to be affective signals. Affective signals convey information about the affective state of the caller; they signal the urgency of the threat, but not a specific cognitive representation of the predator. There is still a lot of debate around the extent to which primate alarm calls are referential, and whether all referential calls also have affective elements.

The purpose of this study was to examine whether putty-nosed monkeys give alarm calls based on predator types (referential signal) or the degree of threat (affective signal). Putty-nosed monkeys are a guenon species found in Western Africa. They spend most of their time in trees. They live in groups of 13-25 individuals consisting of a female harem with one male. Only males produce alarm calls which consist of patterns of “hacks” and “pyows” sounds. The two main predators of the monkeys are crowned eagle and leopards (large mammals). The experimenters played recordings of leopard and eagle sounds near groups of monkeys and recorded the subsequent alarm calls the males produced. They experimentally manipulated the type of predator played and the sound’s distance away from the group (to manipulate degree of threat). The study used spectrographic analysis to describe many dimensions of the entire alarm call including the duration of the call, the number of pyows and hacks, the call rate, the dominant frequency, the peak frequency, and the rate of change in dominate frequency throughout the call.

The study found that in general there was a stronger response (louder and longer) to eagle sounds than leopard sounds. Hack sounds were produced almost exclusively in the eagle alarm calls while pyows were the dominant sound used for large mammals (both in response to humans and leopard sounds). However, the experimenters could not conclude that the signals were referential because the pyows and hack sounds were not completely discrete and still only probalistically predicted the type of predator. Eagles are a greater threat to the monkeys than land predators, so it is possible that hacks are signaling a higher affective arousal rather than the cognitive concept of “eagle.” Interestingly, however, the experimenters also found that the signals were not consistent with a purely affective signal. They believe that the signal may be a combination of both referential and affective signaling, however a lot more experimentation is necessary before we understand how these two signals fit together.

This study further explores the concepts of referential signaling that we discussed in class. It gives an example of another species that may use some degree of referential signals. It introduces the idea that animals could be using both affective and referential signals at the same time. The study also gave some evolutionary theories for why referential signaling developed in some primate species and not others, which we did not discuss in class. The paper suggests that referential signals develop when there are distinct strategies for dealing with different predators. For many animals that live on the ground, there is only one plane of escape, and the response to predators is the same regardless of whether the predator is a bird or a leopard. If the response is always the same than there is no advantage to having referential signals because conspecifics do not need to know exactly what threat is approaching, just that there is a threat. However, many species of monkey live on both the ground and in trees and have different escape behaviors depending on what threat is approaching. This condition could be more important for some species than other. In putty-nosed monkeys there is a much greater threat from eagles than leopards, so they may not have developed as strict a signaling system for “large mammal” as they have for “eagle.”

Arnold, K. & Zuberbuhler, K. (2006). The alarm-calling system of adult male putty-nosed monkeys, Cercopithecus nictitans martini. Animal Behaviour, 72, 643-653.