The search was aimed at the loudest, most social species
Comparative studies have tended to favor social, gregarious and talkative primates as models for language evolution, on the assumption that a rich social life drives vocal complexity [1]. Solitary and dispersed species were expected to be silent or communicatively poor, but that expectation was often presumed rather than tested, so their capacities rarely entered evolutionary theory [1]. Orangutans are the only great apes that live without permanent social groups, and they are also the only nonhuman primate known to share two features of language with humans: displaced vocal reference, where mothers delay alarm calls in order to teach offspring about past predator encounters, and vocal hierarchies, where vocal motifs are nested inside self-similar motifs [1]. Finding these traits in the most private and reserved great ape suggests that elaborate vocal patterns may depend on listeners who can stay quiet and attentive, while species with frequent, rapid social chatter face interruptions that constrain vocal elaboration [1].
Ape vocal learning is real but implicit, not deliberate imitation
In Guinea baboons, males living at the same social level produced more similar grunts than males at different levels, and acoustic similarity did not track genetic relatedness, which points to social interaction rather than inheritance as the driver [3]. The effect sizes were small, so this is a modest convergence rather than a dramatic copying of sounds [3]. The authors treat it as a product of sensory-motor integration and describe it as an implicit form of vocal learning shared with humans, in contrast to the goal-directed, intentional learning that characterizes human speech acquisition [3].
A separate review of primate vocal learning argues that standard animal learning theory, built on isolated lab animals and artificial cause-and-effect setups, fits real-life primate learning poorly [4]. In the field, learning often happens when other group members act as the causes of events, even without direct consequences for the learner, and the default mode may be acquiring scripts, meaning mental representations of how events typically unfold, rather than being conditioned through repeated trials [4]. Becoming communicatively competent, on this view, is more about learning how events unfold than about stimulus-response conditioning [4].
Vocal learning was never the required first step toward speech
A common assumption holds that vocal learning, the ability to imitate and modify vocalizations, was an obligatory precondition for speech evolution [2]. That assumption deserves scrutiny: by the time ape-like human ancestors emerged in Miocene forests, the ancestors of vocal-learning species already walked the earth and flew the skies, and millions of years of head start did not produce linguistic elephants, bats or birds [2]. This suggests that the humble vocal beginnings of hominids were the decisive factor for verbal evolution, and that speech should not be treated as the pinnacle endpoint of vocal learning evolution or as a parallel to far-related vocal learners [2]. Current evidence from great ape calls is described as offering new detail about the extinct vocal forms and functions that let human ancestors jump-start speech, shifting attention toward the hominid biotope, body, brain and behavior [2].
What this reframes about the origins of language
Taken together, these findings suggest that the elusiveness of vocal learning in our closest relatives is partly a measurement and sampling problem and partly a matter of what kind of learning to expect. Apes show implicit convergence and socially scripted learning rather than human-style intentional vocal imitation [3][4], and the most reserved great ape displays language-like vocal features that social-chatter models would not predict [1]. The evolutionary implication is that the (a)social prerequisites for primate vocal evolution, and for language among ancestral hominids, may need reframing around silence, attentive listening and impulse control as much as around social complexity [1]. Speech, on this account, is better explained by the specific conditions of hominid life than by a general capacity for vocal copying [2].
Sources used in this answer
Silence as an Overlooked Catalyzer for Primate Vocal Evolution.
Orangutans, the only great apes without permanent social groups, share displaced vocal reference and nested vocal hierarchies with humans, suggesting silence and attentive listening are underappreciated preconditions for vocal elaboration.
Vocal Learning Versus Speech Evolution: Untangling a False Equivalence
Vocal learning was not necessarily an obligatory precondition for speech, since vocal-learning lineages existed long before ape-like human ancestors without producing language, pointing instead to hominid-specific conditions.
Vocal convergence in a multi-level primate society: insights into the evolution of vocal learning
Male Guinea baboons at the same social level had more similar grunts, an effect not explained by genetic relatedness and interpreted as small-scale implicit vocal learning through social interaction.
Socially scripted vocal learning in primates
Primate vocal learning in the field fits poorly with reinforcement-based animal learning theory and may instead work by acquiring scripts of how events typically unfold, with conspecifics as event-causing agents.
Dynamic vocal learning in adult marmoset monkeys
Adult marmosets show dynamic vocal convergence, supporting an interactive process of vocal learning in primates relevant to language evolution.