Page 9 - Fall 2011

Table I. Some of the kinds of judgments that listeners can make from voices al., 2010) further point to a long evolutionary history of voice recognition abilities (Belin and Grosbras, 2010). Producing and recognizing familiar voice patterns thus antedates, by millions of years, the more lauded evolutionary development of speech and language in human communication and cog- nition. For discerning cohort, friend from foe, and recogniz- ing intimate family members—and being able to achieve this at a distance and in the dark—the preeminence of the famil- iar voice pattern in evolutionary biology can hardly be exag- gerated (Sidtis and Kreiman, 2011). Recognition of the familiar voices of animals that are not first-degree relatives is less common, but helps maintain prox- imity and promotes group cohesion in social animals by pro- viding a means of separating insiders from outsiders, even at a distance (Fig. 1). For example, female vervet monkeys can rec- ognize the voices of their own offspring but also of unrelated juveniles, and can associate those voices with the correct mother (Cheney and Seyfarth, 1980); and female baboons rec- ognize both the screams and threat grunts of unrelated indi- viduals (Cheney and Seyfarth, 1999). Playback experiments (in which recorded vocalizations are broadcast in the field to freely-behaving animals and responses are recorded) have shown that the extent of elephants’ defensive responses (bunching together, retreating) to the voices of elephants from the other family groups encountered within their range can be predicted by the frequency with which those animals are encountered. Response patterns imply an ability to recognize about 100 individuals (McComb et al., 2002). In the vast land- scape of biological vocal recognition, not to be neglected is the ability of nonvocal reptiles to recognize alarm calls of other species (Vitousek et al., 2007). Although these abilities are impressive, they pale in comparison to prodigious human abil- ities to recognize the voices of people we are not related to. Besides our friends, family, neighbors, and other associates (the “familiar-intimate” set), thanks to the media we are easily able to recognize and identify scores of people we have never spoken to or even met (the “familiar-famous” voices: actors, politicians, announcers, broadcasters), as well as fictional beings of endless variety (Bugs Bunny, Hal the computer, and Robby the Robot, for example). In fact, studies of familiar face recognition (Bahrick et al., 1975) and informal voice recogni- tion challenges suggest that there may not be an upper limit to the number of voices humans can recognize (Ladefoged and Ladefoged, 1980). In contrast, it is not clear how much attention listeners of any species actually pay to unfamiliar voices under normal circumstances. Most animals, including humans, treat unfa- miliar voices as part of the background of noise that sur- rounds them every day. As an example, imagine yourself on a busy street, surrounded by strangers talking to each other or on their cell phones. The voices we hear under these circum- stances, although ubiquitous, barely penetrate consciousness. In fact, in a study in which the original caller was surrepti- tiously replaced with a different talker during a telephone survey call, only 6% of subjects noticed the change (Fenn et al., 2011). In contrast, the voice of an approaching friend jumps out from a background of unknown voices, much as the sound of our own name emerges from the unattended voices of their offspring immediately after birth, suggesting calls have a genetic component (Scherrer and Wilkinson, 1993). These biological scenarios cast an eerie doubt on the traditional assumption that all voices, at the first instant, are unfamiliar. Voice recognition facilitates reunions between foraging parents and offspring that are mobile or located in a crowded crèche, helps animals ensure that care is provided to the cor- rect infant, and promotes bonding between mothers and infants. The wide distribution of voice recognition abilities across species, combined with the clear survival value of such abilities and their strikingly full-blown ontogenetic appear- ance, suggests that familiar voice recognition is evolutionari- ly very old. In fact, it may have appeared by the time that frogs emerged (Burke and Murphy 2007; Bee and Gerhardt, 2002; see Kreiman and Sidtis, 2011, for more review). Studies showing that primate brains may have voice-sensitive areas analogous to those seen in human infants as young as 7 months (Petkov et al., 2008; Petkov et al., 2009; Grossmann et 8 Acoustics Today, October 2011