Dance displays in gibbons: biological and linguistic perspectives on structured, intentional, and rhythmic body movement

The study investigates dance-like displays in female crested gibbons (*Nomascus*), a behavior previously receiving limited scientific attention. While the concept of dance is typically associated with humans (defined as intentional, rhythmic, and non-mechanically effective body movement), this study explores its applicability to non-human primates, specifically focusing on crested gibbons. The authors aim to characterize these displays using both biological and linguistic tools, bridging the gap between biological and linguistic perspectives on animal communication. Previous research has documented these displays (sometimes referred to as "bobbing") anecdotally and in limited contexts, primarily focusing on their role as proceptive signals during oestrus. However, a detailed analysis of their structure, intentionality, and rhythmic organization has been lacking. This study addresses this gap by analyzing close-range video recordings of captive gibbons from multiple species, aiming to describe the structural organization of the movements, assessing the intentionality of the behavior, and quantifying the rhythmic patterns. Furthermore, a survey of experts in crested gibbon behavior supplements the analysis with observations from both captive and wild settings. The central hypothesis is that gibbon dance displays represent a structured, intentional, and rhythmic form of visual communication that exhibits regularities and rules akin to those found in human dance, despite acknowledging the non-homologous nature of the behaviors in humans and gibbons. The study also explores potential evolutionary pathways, considering the phylogenetic relationships between gibbon species and comparing the displays to similar behaviors in related primates.

Existing literature on crested gibbon communication primarily focuses on their vocalizations, which include elaborate song bouts. While anecdotal observations of dance-like displays in captive individuals existed, systematic studies were scarce. Lukas et al. (2002) provided some early observations of "bobbing" behavior, noting its increased frequency during oestrus. Fan et al. (2016) presented a more comprehensive analysis of dances in wild *Nomascus nasutus* females, characterizing them as rhythmic body movements and suggesting that they serve multiple functions beyond proceptive signaling, including strengthening inter-sexual bonds and intrasexual competition. However, these studies lacked rigorous analyses of the displays' structure, intentionality, and rhythmicity. The current understanding of gibbon gestural communication is also limited, with no prior comparative assessments. This study aims to build upon these previous works by providing a more detailed and nuanced analysis of gibbon dance using both biological and linguistic frameworks, incorporating data from both captive and wild gibbon populations.

The study employed a multi-faceted methodology combining video analysis and a survey of experts. Video data consisted of 37 initial videos showing dance events, with 11 later discarded due to external disturbances. The remaining 26 videos, featuring seven captive female crested gibbons from four *Nomascus* species, were used to analyze intentionality, structural grouping, and rhythm. The definition of a "dance event" was based on abrupt body stiffening accompanied by rhythmic twitching movements. Intentionality was assessed using criteria established in great ape research, focusing on audience-checking, persistence, and elaboration. Structural grouping analysis involved coding movement patterns (up-down, left-right, diagonal, etc.) and systematically identifying groups of similar movements within dance bouts. The study drew upon existing approaches to the structural organization of human dance, focusing on segmentation and grouping of similar behaviors, using changes in movement parameters to establish group boundaries. Three hypotheses were tested: a null hypothesis of random movement sequences, a one-level grouping hypothesis, and a two-level grouping hypothesis (where some groups are nested within larger groups). Rhythm analysis, using methods previously applied to gibbon songs, involved quantifying inter-onset intervals (IOIs) between twitch movements and calculating rhythmic ratios (r) to assess isochrony. Videos were analyzed in BORIS software at reduced speed to accurately code twitch onsets. A self-administered online survey was distributed to primatologists and zoo/rescue center staff to gather information on dance occurrence, contexts, and demographic characteristics of dancing individuals. Inter-observer reliability tests were performed using Cohen's kappa and intra-class correlations to validate the consistency of coding.

The study found evidence supporting the intentional nature of crested gibbon dances. Audience-checking behaviors were observed in 61% of dances, indicating sensitivity to the presence of an audience. Persistence and elaboration, though less frequently observed, were also documented. The analysis of grouping structure revealed a range of complexities. Homogeneous dances (single group of movements), simple dances (multiple groups without nested structure), and complex dances (with nested groups) were identified. The majority of dances (79%) could be explained by a one-level grouping structure, while complex dances suggested the presence of a two-level grouping. Changes in posture and direction of movement were identified as primary factors influencing group boundaries. Rhythm analysis indicated a predominantly isochronous pattern across all species, suggesting a consistent temporal structure regardless of individual differences in IOI length. The survey confirmed that dances are exclusively performed by sexually mature females, across all age classes, targeting both conspecifics and humans in a variety of contexts. In captive gibbons, dances occurred not only in copulatory contexts but also in social interactions and during periods of arousal or frustration. No effect of hormonal contraception was observed on dance behavior. Three respondents reported observing similar dances in siamangs, suggesting potential phylogenetic connections.

The findings demonstrate that crested gibbon dances represent a complex form of visual communication that is intentional, rhythmic, and structurally organized. The presence of isochrony, previously documented in vocal communication of gibbons, extends to visual displays, underscoring the importance of rhythmic patterns in primate communication. The variable grouping structures suggest a potential for complex signaling, although further research is needed to clarify the semantic content conveyed by different dance types. While the primary function of dances in wild gibbons is likely proceptive signaling, the observation of dances in diverse contexts in captivity raises questions about their flexibility and the potential for multiple functions. The observation of dances directed toward humans suggests a capacity for attention-getting in interspecific interactions, though alternative interpretations, such as tension-release behavior, are also considered. The potential phylogenetic links to similar displays in siamangs and hoolocks warrant further investigation. Comparison of gibbon and human dance, although sharing commonalities in rhythmic and structured movement, highlights distinct evolutionary trajectories.

This study provides the first comprehensive analysis of dance displays in crested gibbons, demonstrating their intentional, rhythmic, and structured nature. These findings expand our understanding of primate communication and offer a unique model for studying complex gestural signals. Future research should investigate the variability of dances across individuals and species, explore potential links to reproductive fitness, and examine the semantic content of different dance types. Investigating gestalt perception in gibbons and comparing their dance perception to that of humans would provide further insights.

The study primarily utilizes data from captive gibbons, potentially limiting the generalizability of findings to wild populations. While the survey expands the data scope, the inherent biases in self-reported data must be considered. The limited sample size in certain analyses restricts the power to draw strong conclusions about certain aspects of dance, such as two-level grouping structures. The opportunistic nature of video collection could introduce biases related to the contexts captured. Further research is necessary to address these limitations and broaden our understanding of this fascinating behavior.