Persuasive conversation as a new form of communication in Homo sapiens

Abstract

The aim of this paper is twofold: to propose that conversation is the distinctive feature of Homo sapiens' communication; and to show that the emergence of modern language is tied to the transition from pantomime to verbal and grammatically complex forms of narrative. It is suggested that (animal and human) communication is a form of persuasion and that storytelling was the best tool developed by humans to convince others. In the early stage of communication, archaic hominins used forms of pantomimic storytelling to persuade others. Although pantomime is a powerful tool for persuasive communication, it is proposed that it is not an effective tool for persuasive conversation: conversation is characterized by a form of reciprocal persuasion among peers; instead, pantomime has a mainly asymmetrical character. The selective pressure towards persuasive reciprocity of the conversational level is the evolutionary reason that allowed the transition from pantomime to grammatically complex codes in H. sapiens, which favoured the evolution of speech.

This article is part of the theme issue ‘Reconstructing prehistoric languages’.

1. Introduction

Our aim in this article is to account for the specificity of Homo sapiens' language in a framework of continuity with both the communication of its hominin predecessors and that of non-human animals based on a quite specific meaning of ‘conversation’. Our proposal relies on three argumentative steps. The first concerns the idea that (human and animal) communication has an intrinsically persuasive nature; the second is that, in order to enhance their persuasive abilities, archaic members of the genus Homo developed narrative forms of communication using pantomime (narrative persuasion is the trait distinguishing human communication from animal communication); the third is that H. sapiens' communication is characterized by the use of narrative strategies in conversational contexts. In such a schema, the transition from narrative communication to narrative conversation coincides with the transition from pantomimic forms of persuasion (typical of archaic members of the genus Homo) to grammatically complex forms of persuasive conversation (typical of members of H. sapiens).

2. Which model of communication?

We adhere to the view that modern language is, from a functional point of view, in continuity with animal communication. From this perspective, understanding the functional role of animal communication is key towards understanding language evolution. Two major models have been proposed to account for such a role: the informative and the manipulative model. According to the former, heavily inspired by Shannon & Weaver's [1] perspective of communication, animal communication's main function is to exchange information through the production and perception of signals, which are encoded by a sender and decoded by a receiver [2–5]. A crucial underlying assumption of this model is that communication rests on cooperative information sharing between the participants and has mutual advantages for them. The discovery that vervet monkeys [6] use different alarm calls to designate different predators (leopards, eagles and snakes) and that these calls convey sufficient information for listeners to adopt distinct adaptive responses even in the absence of a direct stimulus has provided important evidence in favour of the information model (for a discussion [5,7]).

Dawkins & Krebs [8,9] advanced an alternative model based on the view that communication is fundamentally competitive and evolves to manipulate receivers, rather than to faithfully inform them: signals serve to make the receiver respond in ways that provide advantages for the sender, regardless of any benefits for the receiver. In their opinion, ‘communication … could be characterized as a means by which one animal makes use of another animal's muscle power’ [8, p. 283]. The manipulative model has had a strong impact on animal communication scholars (e.g. [10–13]), but has also been criticized for referring to a passive conception of the perceiver compared to the active role of the signaller. Indeed, as highlighted by Seyfarth et al. [5], proponents of the manipulative model assume that ‘listeners are automata that can be manipulated to respond in ways beneficial to the signaller as long as the right nervous system buttons are pushed’ [5, p. 4]. In the light of these criticisms, it may be useful, as suggested by Owren et al. [13], to replace ‘manipulation’ with ‘influence’.

In this article, we adhere to the model of influence arguing that, in addition to delineating a path of continuity between animal and human communication, it paves the way to the possibility of looking at the specificities of H. sapiens' communication without forgetting the evolutionary links that language maintains with the most archaic forms of communication.

3. Persuasion, narrative and pantomimic communication in archaic hominins

Adhering to the model of communication as influence, we contend that language evolved to modify/change the beliefs and attitudes of other people in order to make the recipient act in a certain way rather than simply informing her/him. Albeit sharing the same functional role with animal communication, humans developed a specific way to influence others, which accounts for the uniqueness of language: in the transition from communication to language, a generic ability for influencing others is transformed into a specific persuasive ability governed by the emergence of storytelling [14]. From this perspective, the origin of language is strongly tied to the origin of narrative [15–17] because stories are a very powerful way to persuade others (e.g. [18,19]). It is our claim that narrative characterized the communication of our hominin predecessors long before the advent of H. sapiens and the emergence of modern language (see also [20–23]). In line with this claim, it has been suggested that forms of narrative were likely present in hominins of the early Pleistocene, such as Homo ergaster [21,24,25]. Before going into the details of language phylogeny, some preliminary clarifications are needed.

At a general level, narrative can be defined as ‘a primary resource for configuring circumstances and events into more or less coherent scenarios involving the experience of persons’ [26, p. 74]. Its persuasive power mainly depends on the ‘emotional effects' that stories elicit in the audience members in reference to both the characters (e.g. through forms of empathic simulation) and the plot (e.g. through expectations about the end) [27]. Stressing the importance of these effects, Green & Brock [28] proposed the transportation-imagery model (TIM) according to which individuals understanding stories may become highly engaged (transported) in a narrative, experiencing a state of intense cognitive and emotional focus on the story (see also [29]). Indeed, several investigations have demonstrated that higher scores on the post-exposure transportation scale were associated with a stronger persuasive impact of stories (e.g. [28,30,31]). An important element of the process of transportation into a story world is imagery production, i.e. the generation of visual, mental representations of a narrative scene that has properties similar to the representations elicited by external stimuli [32]. This appears to be crucial to explain the persuasive power of narrative: mental imagery, facilitating mental simulation of events in a story, may lead individuals to adopt beliefs and intentions that are implicated by the sequence of events. In fact, while immersed in a story, people are less aware of the real-world facts that contradict assertions made in the narrative and, therefore, they ‘may be less likely to disbelieve or counterargue story claims, and thus their beliefs may be influenced’ [28, p. 703, our emphasis].

Considering the effects of a narrative on audience members in reference to forms of engagement, transportation and imagery is a way to account for the difference between the persuasive effects of argumentative discourses and the persuasive effects of stories. In argumentation speakers' persuasive intentions are explicit. This has the consequence of generating counterarguments, namely direct confutations toward an overtly persuasive message or in response to a counter-attitudinal statement (e.g. [33]). On the contrary, in the case of narrative, strong immersion into a story prevents (or reduces) counterarguments against story assertions because it creates a lifelike experience and provides solid connections with characters [19]. As stressed by Bilandzic & Busselle [19] ‘we do not need an argument when we have an image’ [19, p. 212]. From these considerations, it follows that narrative is a powerful tool to expand the intrinsically persuasive nature of communication.

It is our claim that forms of narrative characterized the communication of our hominin predecessors long before the advent of H. sapiens and the emergence of modern language (see also [20–22]). Two lines of evidence can be developed to support this claim: the first, which is founded on the idea that narrative is a modality of thought long before a modality of communication, concerns the cognitive systems which allowed archaic hominins to mentally represent reality in a narrative format; the second, which relies on the idea that narrative communication predates verbal communication, regards the expressive system archaic hominins could rely on to share their narrative representations with others.

According to the first line of evidence, we contend that to shed light on cognitive systems underlying narrative representations of reality it is necessary to focus on the core proprieties of stories. Although narratology has provided a very articulated picture in this regard [34], there is great consensus that at the basis of a narrative, there are three essential features: character, plot and time. From that, it follows that the cognitive systems underlying narrative must be capable of processing these three features.

Theoretical models privileging the role of the character (e.g. [35–38]) consider the mindreading system crucial for narrative processing. The idea underlying these models is that the human mind treats fictional characters as real persons: we adopt the system of mindreading used to understand people's goals and motivations in daily social interactions also to give meaning to the actions described in narratives. This system is critical for the transportation of audience members into the life of the story characters as it allows simulation of the thoughts, feelings, beliefs and desires of the protagonists [39]. Perspectives attributing a major role to the temporal sequence of events that comprise the plot (e.g. [34,40]) explain narrative processing with reference to systems of spatial and temporal projection [20,41–43]. These systems are responsible for the process of scene construction (e.g. [44]), which involves binding together multiple elements of imagined scenes, like those described in a story, including contextual details (e.g. sounds, feelings, thoughts, people and objects). The mindreading system and the systems of spatial and temporal projection work hand in hand in narrative processing and represent the core of a neurocognitive network that can be defined as the narrative brain [14].

The systems constituting the narrative brain at the basis of the narrative representation of reality are devices not specific to language. The mindreading system has evolved to make sense of other's behaviours; the systems of spatial and temporal projection have evolved to allow individuals to detach from the present context imagining different spatial and temporal scenarios. This allows corroboration of the view they are independent from language and preceded it phylogenetically, on the one hand, and that they were likely present in archaic hominins, on the other [20,21,45]. From these considerations, it is legitimate to hypothesize that narrative representations made possible by the narrative brain constrained the form of the expressions used by our predecessors in communicating their thoughts to others. In other words, it can be assumed that narrative communication followed narrative representation.

The second line of evidence regards the transition from narrative representation to narrative communication. To account for this transition, it is necessary to bring into play an appropriate expressive system able to convey narrative representations. Pantomime appears to be a suitable candidate. The idea that pantomime characterized the communication of archaic hominins is at the core of most models that endorse a gestural/visual origin of language [20,46–51]. Some of these models adopt a general view of pantomime (e.g. [50,52]), which is considered as a form of iconic gesturing (for a discussion [53]). Here, we refer to a more specific conception in line with the definitions advanced by Żywiczyński et al. [54] and McNeill [55]. Żywiczyński and colleagues [54] described pantomime as a nonverbal, mimetic and non-conventionalized means of communication in which events are represented iconically by coordinated movements of the whole body. The reference to event representation rather than individual actions or objects makes pantomime an ideal means of expression for narrative representation. In this sense, pantomime is different from isolated gestures, which are characterized by movements with clear boundaries and correspond to discrete concepts. Indeed, as highlighted by McNeill, pantomime ‘is dumb-show, a gesture or sequence of gestures conveying a narrative line, with a story to tell, produced without speech’ [55, p. 2]. It is relevant that execution and comprehension of pantomime gestures strongly rely on mental imagery abilities (e.g. [56]), which also contribute to the process of scene construction [57]. These considerations open the way to the possibility that pantomimic narrative, being able to evoke visual representations of a narrative scene that enhance the transportive experience into story world, is an effective means of persuasion.

That said, it is important to highlight that, although we refer to McNeill's definition of pantomime, our proposal differs from his evolutionary model. According to McNeill [58], pantomime does not constitute a precondition of human language because it is produced in the absence of speech, i.e. it does not have the two modes of semiosis (codification for speech and imagery for gesture) that characterize language, the combination of which allows one to express an idea simultaneously. For that, in McNeill's view, speech and gesture were evolutionary ‘equiprimordial’. However, challenging this view, experimental studies testing the relation between modality, interaction and the emergence of language found that, at least in the early stages, gesture played a dominant role compared to vocalization in pantomime [59–61]. Specifically, these studies showed that when gesture is produced in conjunction with vocalizations, this does not necessarily increase the power of a communicative act (i.e. accuracy in identifying communicated meanings) over gestural communication. This does not mean, however, denying the multimodal nature of communication, i.e. the fact that multimodality is the default mode of interaction of human communication [62–65]. What we are stressing here is that there is evidence revealing that gesture has a greater potential for ‘bootstrapping’ a communication system than non-linguistic vocalization (e.g. [66,67]). That said, pantomime would certainly have gradually included vocalizations [47], which would have the effect of enhancing the overall message (as we will argue below, vocal communication evolved for complex syntax that, in turn, developed to facilitate persuasive communication). Following these considerations, we take the view that pantomime was a precursor of human language because, as shown, it is characterized by the two properties that distinguish human language from animal communication: its persuasive nature and its narrative format.

4. The rise of modern language: from communication to conversation

We suggest that the ‘invention’ of modern language by H. sapiens was constrained by the need to find an effective way to construct persuasive arguments for resolving conversational disputes. At the basis of our proposal is the idea that conversation is characterized by a form of reciprocal persuasion between the interlocutors. This proposal, strongly connected to the conception that language has a narrative foundation, can be outlined as follows: (i) the narrative representation of reality characterizes the cognition (the system of beliefs) of individuals in a perspectival form, i.e. each subject always represents reality from a certain point of view; (ii) conversation is a form of communication characterized by the different (often, divergent) points of view of interlocutors; (iii) the different points of view of interlocutors constitute a selective pressure towards forms of communication founded on reciprocal persuasion; (iv) conversation is the evolutionary reason, as we will see in ‘A triggering factor for grammar and speech’ section, for the invention by H. sapiens of a complex grammar and articulated forms of speech in order to provide arguments aimed at both supporting and defending one's own point of view. That said, although modern language has distinctive traits, it is nevertheless in continuity with archaic forms of communication. Specifically, pantomime represented the primitive form of communication capable of making humans able to persuade by telling stories. After stressing the importance of the features that allow us to consider pantomime as a form of protolanguage that acted as a platform for the invention of modern language, it is now time to analyse its limitations, which prompted modern humans to find a new expressive system.

(a). Pantomime on a platform of trust

Pantomime is a powerful tool for persuasive communication, but not an effective tool for persuasive conversation. Accordingly, the emergence of modern language is connected to the overcoming of some traits of pantomimic communication which make it unsuited for reciprocal persuasion between the interlocutors. The selective pressure towards persuasive reciprocity of the conversational level is the evolutionary reason that allowed the transition from pantomimic protolanguage to verbal and grammatically complex forms of expression.

Our argument starts from the proposal that pantomime can be framed within the Platform of Trust (PoT; [68,69]), an assumption that Wacewicz & Żywiczyński [70] consider a ‘deep’ design feature of language. In their opinion, human communication is characterizable in terms of ‘information donation’: unlike other animals, humans can flip ‘the default setting from expecting manipulation to expecting honesty’ [70, p. 170]. Referring to information donation means referring to an ‘informative altruistic act’ (informing others of things even when informers themselves have no personal interest in the information) which, according to some models of language evolution [52], distinguishes human communication from an animal form of communication. This fact places communication within the broader cooperative character of human nature [71]. Moreover, interpreting communication as an altruistic act within the PoT tends to consider pantomime in a perspective strongly tied to the information model of communication. According to the authors,

the minimal requirements for successful pantomimic communication (…) need to be complemented by a social cognitive trait of motivation: of the producer (i) to inform the receiver and (ii) to do it truthfully, and of the receiver to (i) trust such messages as non-deceptive and (ii) to respond appropriately [70, p. 176].

In this view of communication, the receiver's appropriate response can be conceived as the point of convergence between the closure of a communication cycle and, through the exchange of roles between interlocutors, the opening of a new cycle. The thesis of communication as a sequence of communicative exchanges based on the receiver's appropriate response leads only to a proto-conversation (communicative exchanges governed by cooperation and PoT) and not to a conversation in the proper sense (communicative exchanges governed by cooperation/competition and reciprocal persuasion). Conversation, indeed, involves a factor which is underestimated in PoT-based models. The distinction between honest and deceptive information, in fact, is not the only option in place: conversational exchanges are also characterized by the notion of opinion. Humans converse because their narrative representations of reality entail different visions of the world (the diversity of opinions characterizes conversational communication). Conversation is the communicative exchange in which the speaker defends her/his own perspectival view from the perspectival views of other interlocutors [72,73] (this is not the only function of conversation, but defending one's own views from those of others only happens in conversation). Interlocutors' reciprocal commitment to persuade each other allows conversation to move forward—conversation involves a form of progression or a sequence of actions [74]. The interlocutor does not only respond appropriately to what the speaker says: she/he provides counter arguments in support of her/his point of view and in order to convince the other, who in turn is trying to convince her/him.

In §2, we stressed the persuasive role of narrative considering how transportation and engagement can prevent counterarguing. However, this is not the only way to convince others using narratives. When it comes to the reciprocal persuasiveness of conversation, the argumentative dimension plays a major role. Counterarguing between interlocutors is the lifeblood of conversation: without a reciprocal exchange of argument to support one's own opinions, conversation gets stuck. Research on conversation in individuals with autism spectrum disorders (ASD) corroborates such a view. Several studies have attested that subjects with ASD have difficulties in conversational exchanges (e.g. [75–77]) and that these difficulties negatively affect their persuasive abilities [78]. Specifically, it has been shown that individuals with ASD have problems in counterarguing: their communicative exchanges tend to get stuck with no possibility of moving forward [79, p. 20]. Turn-taking impairments in subjects with ASD can be interpreted in terms of inabilities to hold and defend their point of view in conversation: the exchange of roles of turn-taking [71,80] is the surface of a deeper phenomenon that can be interpreted within communicative exchanges governed by the logic of diverging points of view. In a conversational scenario of this kind, it has to be recognized that competition plays a major role in human communication. This does not mean denying that receivers are predisposed to trust the message; nor denying that in conversation interlocutors are engaged in collaborative exchanges. It means that competition plays as important a role in communication as cooperation does. These observations allow us to make some reflections on the role of pantomime in the transition from archaic to modern forms of communication.

A clue supporting the ineffectiveness of pantomime as a conversation tool is its mainly asymmetrical nature. Indeed, to the best of our knowledge, experiments that have tested the communicative efficacy of pantomime are based on communicative tasks in which a group of participants is asked to describe an object or an event by using gesture without speech, and another group of participants is asked to guess the content of the pantomimic act (e.g. [59,60,66,81]). Exchanges of this kind are cases of interactive communication, but not forms of conversation. It is not a coincidence that most of the examples of pantomimic communication concern the case of leaders who persuade individuals to act in a certain way. McBride [22], for example, sets a likely scenario of an archaic pantomimic storytelling involving an alpha male who has just returned to his group after a hunt:

Alpha mimed the hunt story … Perhaps the mime included some abbreviating gestures like starting with; Go Stream, or Antelope Baby. The mime was easy to understand … The play metasignal communicates that what followed was ‘not real,’ … but they had no difficulty recognizing the story. What had been transmitted was something gestures alone could never do, present a whole story, a metaphor for an event [22, p. 3].

This example, which represents the prototypical case of trust in the interlocutor, is a form of asymmetrical communication: the sender persuades; the receiver is persuaded. Such a form of communication is not effective for conversational exchanges based on persuasive reciprocity.

From the asymmetrical character of pantomime, it does not follow that pantomime is not an effective tool for face-to-face communication; it follows that pantomime is not an effective tool for conversation. To account for conversation, it is necessary to refer to a new expressive tool capable of articulating thoughts in the form of arguments. As shown above, pantomime can express narrative with the aim of persuasion because it exploits the processes of transportation and imagery production. However, if we explain the persuasive power of stories only referring to these processes, we risk not recognizing narrative elements in arguments and rhetorical elements in narrative [19]. Acknowledging that narrative has an argumentative dimension is a way to maintain that, along with transportation and imagery, elaboration and inference—cognitive activities of the audience associated with deeper reflection of the story—also contribute to make narrative a tool of persuasion. In this regard, Bilandzic & Busselle [19] suggested that when people are strongly immersed in a story, they also make more inferences about the implications of the narrated events. In line with this view, it has been found that the argument strength of narratives affects their persuasiveness: stories with strong arguments are more persuasive than stories with weak arguments [82]. That said, if part of the persuasive power of narrative relies on argumentative strategies, it is legitimate to imagine that narrative abilities further developed to make cognitive activities underlying argumentation (elaboration and inference) possible: the reciprocal persuasion required by the conversational level led to the development of representational structures supporting inferential processes. From this point of view, conversation was a triggering factor for the emergence of grammar.

(b). A triggering factor for grammar and speech

Within the scenario we are outlining, grammar is the product of selective pressures governed by pragmatics: the use of language in conversational contexts is the triggering factor for the transition from archaic forms of language to modern ones. Specifically, the evolution of complex grammar is an adaptive response to the use of stories in conversational exchanges. The point to start with is that defending one's beliefs from the beliefs of others means articulating thoughts and expressing them in specific representational structures. This implies that the enhancement of the argumentative structure of narrative depends on the enhancement of propositional structures on which inferences are based (for a similar point, see [83]). Indeed, as Sperber & Wilson [84] stated:

There is a very good reason for anyone concerned with the role of inference in communication to assume that what is communicated is propositional: it is relatively easy to say what propositions are, and how inference might operate over propositions. No one has any clear idea how inference might operate over non propositional objects: say, over images, impressions or emotions. Propositional contents and attitudes thus seem to provide the only relatively solid ground on which to base a partly of wholly inferential approach to communication [84, p. 57].

From these considerations it follows that in order to improve the argumentative structure a sophisticated expressive system is needed at the grammatical level: syntax, indeed, contributes to shape in a propositional format the informative content necessary to construct persuasive arguments in conversations. In other words, interactions taking place in conversational contexts push syntax to expand the argumentative structures necessary for persuasion in such contexts, thus facilitating inferential processes. Such a view is consistent with the idea of ‘syntax in conversation’ carried out by interactional linguistics [85]. The triggering role that conversation plays for the evolution of grammar emerges exactly within the primacy of interaction characterizing the ‘grammar at work’. In a perspective of this kind, conversation provides selective pressures that lead to complex syntax.

Accepting the idea that conversation is the triggering factor for grammar (a way to explain why grammar emerged), the question to be addressed now is how grammar shaped language. The emphasis placed on the primacy of pragmatics and the reference to interactional linguistics lead towards a model of grammaticalization interpretable within the framework of cultural evolution [86,87]. Maintaining that ‘a sine qua non for grammaticalization is the transmission of linguistic knowledge from one generation to the next’ [87, p. 382], Heine et al. stress the role of cultural transmission. Similarly, Christiansen & Kirby [88] suggest that the growing complexity of grammar is ‘due to the process of transmitting language across generations through the narrow filter of children's learning mechanisms' [88, p. 303]. Cognitive architectures and processes allowing children to learn language represent important constraints in the construction of a complex grammar [89]. In this view, the construction of grammar is a cultural rather than a biological factor:

Language is manifestly a socially learned, culturally transmitted system. Individuals acquire their knowledge of language by observing the linguistic behaviour of others, and go on to use this knowledge to produce further examples of linguistic behaviour, which others can learn in turn (…). The fact that language is socially learned and culturally transmitted opens up a second possible explanation for the design features of language: those features arose through cultural, rather than biological, evolution. Rather than traditional transmission being another design feature that a biological account must explain, traditional transmission is the feature from which the other structural properties of language spring [89, p. 3594].

An alternative model to grammaticalization regards language as a biological organ rather than a cultural construct. There are two ways of interpreting language as a biological organ. The first, which can be referred to the Chomskyan tradition, considers syntax an all-or-nothing phenomenon with no precursors in the animal kingdom: the origin of grammar is a qualitative leap due to a catastrophic jump [90–92]. Such a view has raised a lot of criticism and has been termed as miraculous by Corballis [20]. Some scholars contrasted the saltationism perspective and, following the path marked by Pinker & Bloom ([93], see also [94]), try to reconcile Chomsky with Darwin. They propose that Chomsky's model (especially Minimalism) is compatible with the gradualism of natural selection. Within this framework, Progovac [95,96] proposes a form of protosyntax working as a basic element for the construction of complex syntax, which gradually evolves in stages driven by natural/sexual selection. In her view, a small clause can be considered the foundation for the evolution of more complex stages, according to the principle that a higher stage ‘preserves and builds upon, the achievements of the previous stages' [96, p. 10].

That said, it should be emphasized that the two ways of understanding the evolution of grammar are not necessarily in contrast with each other: language is a bio-cultural hybrid [97] that cannot be interpreted in exclusive reference to cultural or biological evolution. The model of grammaticalization based on cultural transmission and constrained by the bottleneck of the child's learning systems is a way of looking at the evolution of language from the point of view of linguistic changes: languages transform themselves to be assimilated within already existing cognitive structures. The development of a grammatically complex expressive code, on the other hand, produces adaptive challenges for the brain: these challenges can be interpreted as accommodations (to use a Piagetian terminology) of processing systems to language [98]. A coevolution process of this type fits well with the feedback loop that Benitez-Burraco & Progovac [99,100] pose at the base of modern language: sophistication of language structure coevolves with the enhancement of conversational abilities to reduce reactive aggression, which represents the characteristic trait of H. sapiens' self-domestication. More generally, this feedback loop can also be extended to mindreading and narrative representation: it is plausible that more complex forms of mindreading and narrative coevolved with (proto-)language and were influenced by it both in evolution and development. For example, as for mindreading, it has been suggested that pre-existing mentalizing abilities are further boosted and amplified through language, conversation and interaction both in terms of coevolution [101–104] and cognitive development [105,106].

The transition to a complex grammar imposed by the use of language in conversation also has relevant consequences for the evolution of the expressive code. In this framework, conversation not only plays a triggering factor for the development of grammar, but also leads to the increasing dominance of speech over gesture. Although there is wide debate about the ‘modality transition problem’, with some authors [107] maintaining that there is no definite solution to this problem yet, it is possible to speculate that while gesture in pantomime played a dominant role in the early stages of language evolution, as it can help bootstrap a communication system, pantomime would gradually include vocalizations. This inclusion took advantage of the later adaptations that would ensure the development of vocal signalling into a much more complex medium than in its initial stage [49]. The two systems would continue to co-evolve, leading to the fully integrated multimodal platform characterizing human communication today. This multimodal platform incorporates two semiotic modes at once: codification in speech and imagery in gesture; speech is analytic, segmented and combinatoric; gesture is global and synthetic [58,108]. Starting from this dual semiosis, it is legitimate to speculate that speech would gradually take on the grammatical properties of human language, while gesture would become semantically and pragmatically co-expressive with speech. On this latter point, Kendon [62] identifies two main functions of co-speech gestures: substantial and pragmatic gestures. Substantial gestures contribute to the utterance content; pragmatic gestures contribute negotiating aspects of the situational embedding. Suggesting that vocal communication evolved for complex syntax does not mean denying that syntax can emerge from gesture, as the existence of sign languages attests (e.g. [109,110]). However, it has to be highlighted that this occurs when ‘communicators are forced to rely solely on gesture, that is, when gesture … must assume the full burden of communication’ [111, p. 34]. When a multimodal system is in place as default mode of interaction, speech typically takes on the burden of grammar.

5. Conclusion

The thesis of this article is that humans are distinguished from other animals by their ability to tell stories for persuasive purposes. In order to persuade by telling stories, our ancient predecessors used pantomime, an expressive medium characterized by an asymmetrical kind of persuasion. To engage in conversation and implement a form of reciprocal persuasion, H. sapiens had to abandon pantomime and invent a new, more effective, expressive system. The construction of a complex grammar linked to the use of verbalization is the response of modern humans to the selective thrusts in favour of a communication based on an egalitarian role of reciprocity: grammar enhances and strengthens the persuasive skills requested by the logic of persuasive reciprocity typical of conversational exchanges by strengthening the argumentative power of narrative. That said, the origin of modern language does not represent a breaking point with respect to the most archaic forms of communication: the conversational skills of H. sapiens exploit persuasive abilities already present in other species of hominins, reconstructing them at a new level. In a perspective of this type, the appearance of H. sapiens on the communicative scene is the appearance of a species capable at the same time of a highly innovative type of communication and a solid connection with the past.

Data accessibility

This article has no additional data.

Authors' contributions

The article is the outcome of a collaborative effort between the authors. For the final draft, F.F. wrote §§ 1, 4 and 5. I.A. wrote §§ 2 and 3.

Competing interests

We have no competing interests.

Funding

We declare we received no funding for this study.