Moffett argues “Societies should be understood fundamentally as identity groups….” If so, then social organisms require evolved mechanisms for detecting identity and advertising it by credibly signaling information concerning group membership.
Humans and other organisms form societies to obtain and maintain benefits, such as food and protection. These benefits create a selection pressure that threatens society members: They lead to social parasites, who mimic cues of society membership to exploit the resources of a society. For example, ant colonies are often exploited by species of spiders, beetles, and other ants that mimic the chemical and behavioral signals and cues of colony membership (Mclver & Stonedahl, Reference Mclver and Stonedahl1993). Similarly, cuckoos exploit other bird species by tricking them into raising their young: cuckoo eggs mimic the appearance of the target host species' eggs, and cuckoo chicks mimic host chick behaviors (Davies, Reference Davies2011; Yang et al., Reference Yang, Liang, Antonov, Cai, Stokke, Fossøy and Røskaft2012).
Humans have culturally evolved similar mimicking strategies, such as duck decoys, whose analogues date to octopus lures in the Mariana Islands, from at least 1500 B.C.E. (Carson & Hung, Reference Carson and Hung2021; Moser, Buckner, Sarian, & Winking, Reference Moser, Buckner, Sarian and Winking2023). Mimicry also enables humans to pass as members of social groups to which they do not belong. Although there is cross-cultural evidence for visual and other markers of ethnic identity (McElreath, Boyd, & Richerson, Reference McElreath, Boyd and Richerson2003), these are vulnerable to parasitism. For example, hairstyle influences perceived race, such that changing one's hairstyle can change how they are grouped by others (Sims, Pirtle, & Johnson-Arnold, Reference Sims, Pirtle and Johnson-Arnold2020).
Given the threat of social parasitism, reliably signaling identity to conspecifics in a society is essential, especially in large societies where individuals may not be known to all other members (Przepiorka and Diekmann, Reference Przepiorka and Diekmann2021).Footnote 1 Here, we argue that vocalizations are a widespread form of honest identity signaling – and are ideally suited to delineating societal units of the type Moffett argues for. Vocal signals, including cues present in language, music, and other, non-linguistic utterances, play an essential role in displaying shared identity in humans, because (1) human auditory perception is sensitive to subtle acoustic features that can uniquely characterize a vocal signal; (2) vocal signals are efficient, requiring minimal energy to produce; and (3) vocal signals can be complex, as in coordinated vocalizations in music, providing a rich medium in which to embed identity signals.
First, human auditory perception is specialized for the processing of speech and music (Singh & Mehr, Reference Singh and Mehr2023; Zatorre, Belin, & Penhune, Reference Zatorre, Belin and Penhune2002), with a high degree of sensitivity to subtle differences in the acoustic content of speech and music tokens. This makes it possible for vocalizations to function as identity signals. For example, infants and adults exhibit social preferences for unfamiliar adults whose accents are familiar (Giles & Billings, Reference Giles, Billings, Davies and Elder2004; Kinzler, Dupoux, & Spelke, Reference Kinzler, Dupoux and Spelke2007) or who sing songs that previously have been produced by a social partner (Mehr & Spelke, Reference Mehr and Spelke2018; Mehr, Song, & Spelke, Reference Mehr, Song and Spelke2016). Social inferences can also be drawn from other, non-linguistic and non-musical utterances: Listeners can reliably detect friendship/affiliation from the sounds of co-laughter, and can reliably detect if laughs are “real” or faked, whether or not the listeners had the same native language as the vocalizers (Bryant et al., Reference Bryant, Fessler, Fusaroli, Clint, Aarøe, Apicella and Zhou2016; Bryant et al., Reference Bryant, Fessler, Fusaroli, Clint, Amir, Chávez and Zhou2018).
Such effects evoke similar patterns in non-human species, where vocal identity signals are common, especially in the domain of territorial signaling (Mehr, Krasnow, Bryant, & Hagen, Reference Mehr, Krasnow, Bryant and Hagen2021). Such vocalizations are omnipresent across species, from whales, who use group-specific vocalizations (target article); to chimpanzees, who produce individual-specific and arguably group-specific calls (Crockford, Herbinger, Vigilant, & Boesch, Reference Crockford, Herbinger, Vigilant and Boesch2004; Desai, Fedurek, Slocombe, & Wilson, Reference Desai, Fedurek, Slocombe and Wilson2022); to red-winged blackbirds, where experimentally removing singing ability reduces their ability to control territory (Catchpole & Slater, Reference Catchpole and Slater1995).
Second, the energy costs of vocal production in humans are small (Titze, Reference Titze2021), certainly smaller than some other signals, such as those involving movement or physical identifiers such as elaborate clothing. The human vocal tract, including the low position of the larynx compared to other species, efficiently produces a wide range of sounds (Fitch, Reference Fitch2018), and humans demonstrate their proclivity for vocalization from birth, spontaneously vocalizing regularly, often in the first seconds or minutes of life (Soltis, Reference Soltis2004). In addition to distress signals, neonates produce protophones (non-cry, speech-like vocalizations) that may signal developmental progress to caregivers (Oller, Ramsay, Bene, Long, & Griebel, Reference Oller, Ramsay, Bene, Long and Griebel2021).
Last, human vocalizations are highly complex, providing a broad palette with which identity signals could be painted. For example, approximately 6000 languages are presently used, far surpassing the range of vocalization in any other species – yet linguistic diversity today likely accounts for only a small fraction of linguistic diversity across human history (Fitch, Reference Fitch2011). Concurrently, music varies along tonal, metrical, harmonic, instrumental, and contextual dimensions, among many others (Lomax, Reference Lomax1968; Mehr et al., Reference Mehr, Singh, Knox, Ketter, Pickens-Jones, Atwood and Glowacki2019), with links between its acoustical forms and behavioral functions (Hilton, Crowley-de Thierry, Yan, Martin, & Mehr, Reference Hilton, Crowley-de Thierry, Yan, Martin and Mehr2023; Mehr, Singh, York, Glowacki, & Krasnow, Reference Mehr, Singh, York, Glowacki and Krasnow2018; Yurdum et al., Reference Yurdum, Singh, Glowacki, Vardy, Atkinson, Hilton and Mehr2023), indicating that it too can function as a signal.
Both types of vocalization also vary on features that make them difficult to mimic. A given language may have only a few phonemes (e.g., 11, in Rotokas, from Papua New Guinea) or many (e.g., 144, in !Xun, from Southern Africa; Evans & Levinson, Reference Evans and Levinson2009). Although infants have the ability to learn any language, human audition is shaped by experience and adults have difficulty distinguishing sounds not native to their language, let alone producing them (e.g., Werker & Tees, Reference Werker and Tees1984). Moreover, the style of vocal music contains complex acoustic information that can be informative in signaling contexts, as in lullabies, which may signal a parent's attentional state to infants (Mehr & Krasnow, Reference Mehr and Krasnow2017).
Evidence for the attempted parasitization signals in humans also supports the role of vocalization in identity signaling: Criminals sometimes attempt to alter voice pitch or accent to disguise age, sex, nationality, or identity (Didla, Reference Didla2020). That these alterations are difficult and rare to compellingly produce suggests ongoing selection for greater abilities to produce and discriminate complex vocalizations. Indeed, aspects of linguistic and musical diversity may have evolved as an anti-parasite strategy in an arms race with hominin social parasites, given the putative role of these vocalizations in identity signaling.
Moffett argues “Societies should be understood fundamentally as identity groups….” If so, then social organisms require evolved mechanisms for detecting identity and advertising it by credibly signaling information concerning group membership.
Humans and other organisms form societies to obtain and maintain benefits, such as food and protection. These benefits create a selection pressure that threatens society members: They lead to social parasites, who mimic cues of society membership to exploit the resources of a society. For example, ant colonies are often exploited by species of spiders, beetles, and other ants that mimic the chemical and behavioral signals and cues of colony membership (Mclver & Stonedahl, Reference Mclver and Stonedahl1993). Similarly, cuckoos exploit other bird species by tricking them into raising their young: cuckoo eggs mimic the appearance of the target host species' eggs, and cuckoo chicks mimic host chick behaviors (Davies, Reference Davies2011; Yang et al., Reference Yang, Liang, Antonov, Cai, Stokke, Fossøy and Røskaft2012).
Humans have culturally evolved similar mimicking strategies, such as duck decoys, whose analogues date to octopus lures in the Mariana Islands, from at least 1500 B.C.E. (Carson & Hung, Reference Carson and Hung2021; Moser, Buckner, Sarian, & Winking, Reference Moser, Buckner, Sarian and Winking2023). Mimicry also enables humans to pass as members of social groups to which they do not belong. Although there is cross-cultural evidence for visual and other markers of ethnic identity (McElreath, Boyd, & Richerson, Reference McElreath, Boyd and Richerson2003), these are vulnerable to parasitism. For example, hairstyle influences perceived race, such that changing one's hairstyle can change how they are grouped by others (Sims, Pirtle, & Johnson-Arnold, Reference Sims, Pirtle and Johnson-Arnold2020).
Given the threat of social parasitism, reliably signaling identity to conspecifics in a society is essential, especially in large societies where individuals may not be known to all other members (Przepiorka and Diekmann, Reference Przepiorka and Diekmann2021).Footnote 1 Here, we argue that vocalizations are a widespread form of honest identity signaling – and are ideally suited to delineating societal units of the type Moffett argues for. Vocal signals, including cues present in language, music, and other, non-linguistic utterances, play an essential role in displaying shared identity in humans, because (1) human auditory perception is sensitive to subtle acoustic features that can uniquely characterize a vocal signal; (2) vocal signals are efficient, requiring minimal energy to produce; and (3) vocal signals can be complex, as in coordinated vocalizations in music, providing a rich medium in which to embed identity signals.
First, human auditory perception is specialized for the processing of speech and music (Singh & Mehr, Reference Singh and Mehr2023; Zatorre, Belin, & Penhune, Reference Zatorre, Belin and Penhune2002), with a high degree of sensitivity to subtle differences in the acoustic content of speech and music tokens. This makes it possible for vocalizations to function as identity signals. For example, infants and adults exhibit social preferences for unfamiliar adults whose accents are familiar (Giles & Billings, Reference Giles, Billings, Davies and Elder2004; Kinzler, Dupoux, & Spelke, Reference Kinzler, Dupoux and Spelke2007) or who sing songs that previously have been produced by a social partner (Mehr & Spelke, Reference Mehr and Spelke2018; Mehr, Song, & Spelke, Reference Mehr, Song and Spelke2016). Social inferences can also be drawn from other, non-linguistic and non-musical utterances: Listeners can reliably detect friendship/affiliation from the sounds of co-laughter, and can reliably detect if laughs are “real” or faked, whether or not the listeners had the same native language as the vocalizers (Bryant et al., Reference Bryant, Fessler, Fusaroli, Clint, Aarøe, Apicella and Zhou2016; Bryant et al., Reference Bryant, Fessler, Fusaroli, Clint, Amir, Chávez and Zhou2018).
Such effects evoke similar patterns in non-human species, where vocal identity signals are common, especially in the domain of territorial signaling (Mehr, Krasnow, Bryant, & Hagen, Reference Mehr, Krasnow, Bryant and Hagen2021). Such vocalizations are omnipresent across species, from whales, who use group-specific vocalizations (target article); to chimpanzees, who produce individual-specific and arguably group-specific calls (Crockford, Herbinger, Vigilant, & Boesch, Reference Crockford, Herbinger, Vigilant and Boesch2004; Desai, Fedurek, Slocombe, & Wilson, Reference Desai, Fedurek, Slocombe and Wilson2022); to red-winged blackbirds, where experimentally removing singing ability reduces their ability to control territory (Catchpole & Slater, Reference Catchpole and Slater1995).
Second, the energy costs of vocal production in humans are small (Titze, Reference Titze2021), certainly smaller than some other signals, such as those involving movement or physical identifiers such as elaborate clothing. The human vocal tract, including the low position of the larynx compared to other species, efficiently produces a wide range of sounds (Fitch, Reference Fitch2018), and humans demonstrate their proclivity for vocalization from birth, spontaneously vocalizing regularly, often in the first seconds or minutes of life (Soltis, Reference Soltis2004). In addition to distress signals, neonates produce protophones (non-cry, speech-like vocalizations) that may signal developmental progress to caregivers (Oller, Ramsay, Bene, Long, & Griebel, Reference Oller, Ramsay, Bene, Long and Griebel2021).
Last, human vocalizations are highly complex, providing a broad palette with which identity signals could be painted. For example, approximately 6000 languages are presently used, far surpassing the range of vocalization in any other species – yet linguistic diversity today likely accounts for only a small fraction of linguistic diversity across human history (Fitch, Reference Fitch2011). Concurrently, music varies along tonal, metrical, harmonic, instrumental, and contextual dimensions, among many others (Lomax, Reference Lomax1968; Mehr et al., Reference Mehr, Singh, Knox, Ketter, Pickens-Jones, Atwood and Glowacki2019), with links between its acoustical forms and behavioral functions (Hilton, Crowley-de Thierry, Yan, Martin, & Mehr, Reference Hilton, Crowley-de Thierry, Yan, Martin and Mehr2023; Mehr, Singh, York, Glowacki, & Krasnow, Reference Mehr, Singh, York, Glowacki and Krasnow2018; Yurdum et al., Reference Yurdum, Singh, Glowacki, Vardy, Atkinson, Hilton and Mehr2023), indicating that it too can function as a signal.
Both types of vocalization also vary on features that make them difficult to mimic. A given language may have only a few phonemes (e.g., 11, in Rotokas, from Papua New Guinea) or many (e.g., 144, in !Xun, from Southern Africa; Evans & Levinson, Reference Evans and Levinson2009). Although infants have the ability to learn any language, human audition is shaped by experience and adults have difficulty distinguishing sounds not native to their language, let alone producing them (e.g., Werker & Tees, Reference Werker and Tees1984). Moreover, the style of vocal music contains complex acoustic information that can be informative in signaling contexts, as in lullabies, which may signal a parent's attentional state to infants (Mehr & Krasnow, Reference Mehr and Krasnow2017).
Evidence for the attempted parasitization signals in humans also supports the role of vocalization in identity signaling: Criminals sometimes attempt to alter voice pitch or accent to disguise age, sex, nationality, or identity (Didla, Reference Didla2020). That these alterations are difficult and rare to compellingly produce suggests ongoing selection for greater abilities to produce and discriminate complex vocalizations. Indeed, aspects of linguistic and musical diversity may have evolved as an anti-parasite strategy in an arms race with hominin social parasites, given the putative role of these vocalizations in identity signaling.
Financial support
J. C. B. and S. A. M. are supported by grants from the Royal Society of New Zealand Te Apārangi (Rutherford Discovery Fellowship RDF-UOA2103 and Marsden Fund Standard Grant MFP-UOA2133).
Competing interests
None.