In their target article, Ackermann et al. contribute to a long-standing debate concerning the extent to which the uniquely human propensity for language is the product of species-unique cognitive mechanisms (e.g., Hauser et al. Reference Hauser, Chomsky and Fitch2002; Penn et al. Reference Penn, Holyoak and Povinelli2008; Pinker & Bloom Reference Pinker and Bloom1990). Their comprehensive analysis of neurological and behavioral evidence strengthens the proposal for evolutionary continuity in the mechanisms underlying acoustic communication in human and nonhuman primates. Our goal in this commentary is to amplify their proposal by highlighting recent behavioral evidence from human infants between 3 and 6 months of age. This evidence, which documents how infants respond to vocalizations of humans and nonhuman primates, bears on Ackermann et al.'s formidable challenge to consider the evidence of evolved acoustic communication architecture within the broader faculties of human language.

Recent studies have documented that even in infants too young to speak, listening to human speech supports core cognitive processes, including the formation of object categories (Ferry et al. Reference Ferry, Hespos and Waxman2010; Fulkerson & Waxman Reference Fulkerson and Waxman2007; Waxman & Gelman Reference Waxman and Gelman2009). Perhaps more surprisingly, this precocious link between human language and cognition is initially broad enough to include the vocalizations of nonhuman primates. For 3- and 4-month-olds, nonhuman primate vocalizations (from a blue-eyed Madagascar lemur) also promote object categorization, mirroring exactly the effects of human speech. However, by 6 months, lemur vocalizations no longer have this language-like effect: Instead, the link to categorization is tuned specifically to human language (Ferry et al. Reference Ferry, Hespos and Waxman2013). These findings reveal that a link between language and object categories, evident as early as 3 months in human infants, derives from a broader template that initially encompasses vocalizations of human and nonhuman primates, and is rapidly tuned specifically to human vocalizations (see also Vouloumanos et al. Reference Vouloumanos, Hauser, Werker and Martin2010).

This striking ontogenetic evidence has strong implications for theories of language acquisition. It also offers insights into Ackermann et al.'s proposal for integrating primate-general and human-specific mechanisms of acoustic communication. We focus here on three. First, the evidence from human infants is consistent with the Ackermann et al.'s proposal that, broadly speaking, the faculties that give rise to human language may be related to those predating Homo sapiens (see also Fitch Reference Fitch2011; Stoeger et al. Reference Stoeger, Mietchen, Oh, de Silva and Herbst2012). What remains to be seen is how precisely the relations between homologous neural structures can be specified. For example, one promising investigation might be to ascertain whether infants' responses to human and nonhuman primate vocalizations engage the neural mechanisms described in the target article.

Second, the evidence from human infants converges with Ackermann et al.'s claim that human language acquisition may be built upon mechanisms that are specialized for acoustic communication. One must, however, consider the necessity of these acoustically-based mechanisms in human language acquisition. Although most humans acquire language in the aural-oral modality, our linguistic capacities are distinctly amodal. The signature of human language is not its perceptual form, but rather its ability to enable its users to express an infinite number of ideas using a discrete number of meaningful elements (Chomsky Reference Chomsky1965). Thus, a complete account of the evolution of human language will be one that considers not only the acoustic-spoken modality but also the visual-manual modality in which deaf infants naturally acquire language. One question is whether, given the evidence for evolved neural hardware underpinning acoustic communication, infants acquiring spoken language might have some advantage. Evidence from infants acquiring sign language casts doubt on this possibility (e.g., Goldin-Meadow & Mylander Reference Goldin-Meadow and Mylander1983; Newport & Meier Reference Newport, Meier and Slobin1985; Petitto & Marentette Reference Petitto and Marentette1991). More recent evidence from our lab underscores infants' flexibility in identifying language-like signals beyond human speech. If a novel signal (consisting of pure sine-wave tone sequences) is embedded within a social communicative exchange, infants endow the signal with communicative status and its effects mirror those of human speech in a subsequent categorization task (Ferguson & Waxman Reference Ferguson, Waxman, Knauff, Pauen, Sebanz and Wachsmuth2013).

Finally, evidence from infants can mutually constrain and inform developing theories of language evolution, acquisition, and usage. For example, we have recently discovered that unlike nonhuman primate vocalizations, zebra finch birdsong does not promote object categorization in human infants at any age (Perszyk & Waxman Reference Perszyk and Waxman2013). This outcome is consistent with claims that, although birdsong shares some structural features with human language, it lacks the links to meaning that characterize human language and, to a much lesser extent, nonhuman primate vocalizations (e.g., Berwick et al. Reference Berwick, Friederici, Chomsky and Bolhuis2013).

Ackerman et al.'s target article invites researchers across disciplines to engage in the larger enterprise of uncovering the origins of human language. Within this enterprise, the biggest leaps will be made by those who integrate seemingly disparate neurological, behavioral, and developmental evidence to unearth the evolutionary continuities and discontinuities in both modality-specific (e.g., vocalizations) and modality-independent capacities that provide humans alone with the capacity to acquire language.