The assessment by vervet monkeys of their own and another species' alarm calls

In Amboseli National Park, Kenya, both superb starlings, Spreo superbus, and vervet monkeys, Cercopithecus aethiops, give acoustically different alarm calls to different classes of predators. The 'raptor alarms' of starlings and vervet monkeys are so called because they are given exclusively to avian species that attack from the air. In contrast, while vervets give 'leopard alarms' to a narrow array of terrestrial predators, starlings give terrestrial predator alarms to a wide variety of species, including some birds. Habituation experiments demonstrate that monkeys compare vocalizations according to their referents, not just their acoustic properties: vervets who learned to ignore playback of a starling's raptor alarm subsequently also ignored playback of a vervet's eagle alarm. Experiments also demonstrate that vervets are sensitive to the breadth of referential specificity exhibited by different calls. Subjects who learned to ignore playback of a starling terrestrial predator alarm subsequently also ignored playback of both vervet leopard and vervet raptor alarms. Free-ranging vervet monkeys, Cercopithecus aethiops, who have learned to ignore one type of call given by an unreliable signaller will subsequently also ignore an acoustically different call given by the same signaller, but only if the calls have similar referents (Cheney & Seyfarth 1988). After repeatedly being played an intergroup 'chutter' in the absence of other groups, vervets not only habituate to this chutter but also ignore an intergroup 'wrr' (a call with consistently different acoustic features) given by the same individual. Such transfer of habituation from one call to another does not occur, however, if the two calls have different referents. Vervets who have habituated to another monkey's leopard alarm call played repeatedly in the absence of leopards, for example, nevertheless still respond to the same signaller's eagle alarm call. Thus, vervet monkeys (and probably other non-human primates) appear to compare calls according to their meaning, not just their acoustic properties. In this paper, we describe playback experiments designed to investigate further the criteria used by vervet monkeys when comparing vocalizations that have consistently different acoustic properties. Using the same habituation/dishabituation technique employed earlier (Cheney & Seyfarth 1988), vervets were presented with their own species' predator alarm calls and the alarm calls given by a sympatric bird, the superb starling, Spreo superbus. Both vervet monkeys (Struhsaker 1967; Seyfarth et al. 1980) and starlings (Cheney & Seyfarth 1985) give alarm calls with consistently different acoustic properties to raptors and to terrestrial predators, and vervets respond to both their own alarm calls and the alarm calls of starlings as if the calls denote specific types of danger. Vervets respond to both their own and starlings' raptor alarm calls, for example, by looking up into the air or running into dense bush (see below). If monkeys compare not only their own but also other species' vocalizations according to the referents of the calls, habituation to one species' raptor (or terrestrial) predator alarm should generalize to the corresponding alarm call of the other species. If monkeys respond to vocalizations according to the objects and events they denote, we should also expect them to be sensitive to the breadth of referential specificity exhibited by different calls. In English, for example, we use words with very specific meanings (or narrowly defined referents) like 'praying mantis' or 'banana' . We also use words like 'insect' or 'food' that denote a much broader class of objects and have, as a result, meanings that are more general. In making judgments about synonymy we take these differences into account. Sometimes praying mantis and insect can be used interchangeably; more often they cannot, because the meaning of insect is too broad. 0003-3472/90/100754+11 $03.00/0 9 1990 The Association for the Study of Animal Behaviour