How can we distinguish between gestures and other behaviors? Should we identify them based on their form or function? Are gestures limited to the visual modality? Does it make a difference whether a gesture is produced in isolation or whether it is combined with a facial expression or a sound? How do these different types of communicative behaviors contribute to the conveyed meaning? These questions reflect some of the current debates in the field of primate communication research (Liebal et al. Reference Liebal, Waller, Burrows and Slocombe2013; Slocombe et al. Reference Slocombe, Waller and Liebal2011), which investigates if and how the communicative patterns of our closest relatives, apes and monkeys, provide insight into the evolution of human language (Arbib et al. Reference Arbib, Liebal and Pika2008; Hewes Reference Hewes1973; MacNeilage Reference MacNeilage1998; Tomasello Reference Tomasello2008; Zuberbühler Reference Zuberbühler2005). Although nonhuman primates’ gestures are fundamentally different from human co-speech or co-sign gestures, I propose that taking a comparative perspective to human communication might be beneficial to answer the issues I want to raise with regard to the article by Goldin-Meadow & Brentari (G-M&B). This commentary specifically refers to the question regarding which criteria should be used to define a gesture and whether it is sufficient to focus on manual gestures only, while not considering the information conveyed by other body parts than the hands.

In comparative gesture research with nonhuman primates (hereafter: primates), much attention has been dedicated to differentiating communicative gestures from other, noncommunicative behaviors (Liebal & Call Reference Liebal and Call2012). However, definitions of gestures and the criteria used to identify them vary substantially across studies (Liebal et al. Reference Liebal, Waller, Burrows and Slocombe2013; Slocombe et al. Reference Slocombe, Waller and Liebal2011). For example, some researchers define primate gestures as expressive movements of the limbs or head, as well as body postures that appear to initiate a desired action (Pika et al. Reference Pika, Liebal and Tomasello2003). Others refer to them as discrete, mechanically ineffective physical movements of the body, which include the whole body or only the limbs and head, but they exclude facial expressions and static body postures (Hobaiter & Byrne Reference Hobaiter and Byrne2011). And yet others also consider facial movements as gestures (Maestripieri Reference Maestripieri and King1999). As a consequence of the current lack of agreement about which sensory modalities (visual, tactile, auditory) and which body parts (whole body, manual gestures, postures, face) should be considered when defining a gesture, systematic comparisons of findings across studies and species are difficult if not impossible.

Furthermore, the majority of primate gesture research investigates gestures in isolation and neglects the fact that they might be accompanied by other communicative behaviors, such as facial expressions and vocalizations (Slocombe et al. Reference Slocombe, Waller and Liebal2011). Consequently, there is currently very limited knowledge about whether a gesture produced alone conveys different information than a gesture combined with other facial expressions or vocalizations (but see Taglialatela et al. Reference Taglialatela, Russell, Pope, Morton, Bogart, Reamer and Hopkins2015). Although it is currently debated whether primate gestures have specific meanings (Hobaiter & Byrne Reference Hobaiter and Byrne2014), or whether their meaning is defined by the context they are used in (Call & Tomasello Reference Call and Tomasello2007), it is a crucial question if and how the combination with other behaviors modifies the communicative function of a gesture (Liebal et al. Reference Liebal, Waller, Burrows and Slocombe2013).

These two current drawbacks in primate gesture research – the lack of a shared, comprehensive gesture definition and the dominance of a unimodal approach to primate communication – also become evident in the article by G-M&B. For example, they define gestures as “manual movements that speakers produce when they talk” (para. 3), but seem to largely ignore other body parts that might contribute to the linguistic information conveyed by the hands. A more critical discussion as to why G-M&B limited their scope to manual, visual gestures would have been helpful to better understand why they defined a gesture the way they did – also in regard to the challenge of differentiating between signs and gestures. For example, did they not consider facial gestures because they would not refer to facial movements as gestures? This would reflect a similar debate in primate communication research, since facial expressions are often referred to as a distinct class of signals compared to visual, tactile, or auditory gestures (Liebal et al. Reference Liebal, Waller, Burrows and Slocombe2013), because unlike gestures, facial expressions seem to represent expressions of emotional states rather than voluntarily produced communicative behaviors (Tomasello Reference Tomasello2008).

There are, however, sign language studies that highlight the role of the body, the head, or the face, and their contribution to the linguistic message primarily provided by the hands (Liddell & Metzger Reference Liddell and Metzger1998; Nespor & Sandler Reference Nespor and Sandler1999). For example, Sandler (Reference Sandler2009) showed that mouth gestures can function as linguistic symbols, which co-occur with the linguistic description of an event. Similarly, movements of the upper face, like squints or brow raises, are part of the intonational system of Israeli Sign Language (Dachkovsky & Sandler Reference Dachkovsky and Sandler2009). Although Dachkovsy and Sandler (Reference Sandler2009) did not refer to these upper-face movements as gestures, they defined them as linguistic facial expressions, which are fundamentally different from emotional facial expressions. Finally, Sandler (Reference Sandler2012a) proposed that different body parts contribute to the articulation of grammatical information in Al-Sayyid Bedouin Sign Language. In this emerging sign language, younger signers isolate different parts of the body for linguistic functions, while older signers use their body gesturally (Sandler Reference Sandler2012a). Although these are just some examples from a limited set of studies, they impressively demonstrate that by focusing on one modality only, it is difficult to capture the complexity of a communicative system. With regard to the G-M&B target article, the focus on manual gestures only might not be sufficient for a comprehensive discussion of the similarities and differences between signed and spoken languages and their relationship with gesture, as long as it remains unclear what a gesture is.