The evolution of musicality: synchronization behaviours and rhythm perception in chimpanzees
Musicality - the ability to enjoy and produce music - is a human universal, grounded on a suite of biological and cognitive mechanisms. The individual, social and therapeutic benefits of music are multiple and uncontroversial, yet there is little agreement about how and why human musicality evolved. While modern musical uses may not entirely reflect the evolutionarily old contexts that promoted protomusicality, many modern benefits of music are connected to the pleasure that arises from rhythmic synchronization of bodies to a musical beat or to other bodies. Excitingly, there are now indications that rhythmic synchronisation (RS) may not be uniquely human, as previously claimed. Current evolutionary theories of musicality increasingly agree that RS constitutes the most ancient component of musicality, and urgently demand cross-species, comparative empirical research. Indeed, extant evidence from nonhuman species is highly limited. To provide the empirical data needed to confirm or refute evolutionary theories of musicality, our project will systematically examine the presence, range and complexity of RS in our closest genetic relative, the chimpanzee. Informed by our own recent research and pilot data, the project will address questions regarding the range of latent perceptual RS and the social antecedents and consequences of RS. The project relies on observational and experimental studies, and uses novel methodological paradigms that rely on non-invasive remote eye-tracking.
Final report
The project purpose was to address the evolutionary origins of human musicality by examining the rhythmic abilities of chimpanzees in experimental and observational studies, and in theoretical works of synthesis. Several project objectives have been achieved (as detailed below), while some have been delayed by unforeseen technical issues (see section 3) and the COVID-19 pandemic, which prevented field work during 2020 and 2021, given restrictions to travel and to work closely with endangered nonhuman apes, which can be easily infected by human respiratory viruses.
MAIN RESULTS
1.1. FUNCTIONAL AND MECHANISTIC THEORIES IN THE EVOLUTION OF HUMAN RHYTHM COGNITION – Theoretical synthesis paper
The study has been published and it has received positive feedback, being described as a "wonderfully inclusive survey, which will be a very useful reference for the field" and as an "impressively wide-ranging and thorough review of theory and research regarding the evolution of rhythmic processing in the context of human musicality". Besides providing a comprehensive review of the field, the study also proposes an encompassing research agenda of previously neglected aspects.
1.2. NATURE BEATS: THE WHAT, WHY, AND HOW OF ANIMAL RHYTHMIC BEHAVIOUR – Workshop and book
Research carried out in the project revealed the need to bridge relevant areas of studies. Given an invitation from Springer Nature, we decided to address this in a collective volume on the topic, which was planned as the output of a workshop to be organised at Lund University Primate Research Station Furuvik (LUPRSF) in October 2020 but postponed until May 2022 due to pandemic-related restrictions (https://portal.research.lu.se/en/activities/nature-beats-the-what-why-and-how-of-animal-rhythmic-behaviours). The workshop gathered leading names in the field and revolved around six pre-established themes derived from the areas of expertise of the participating researchers. The event was highly appreciated by participants and a networking success. The collective volume – an originally not planned project output - is the first book to systematically address rhythmic phenomena in animals and humans from multiple perspectives and disciplines. The volume gathers contributions from leading researchers in the field and is the outcome of a workshop organised as part of the project.
1.3. RHYTHMIC BEHAVIOURS IN CHIMPANZEES – Observational study
Data collection for the study was completed in February 2020. Besides the two initially planned sites (LUPRSF, N=6; MONA Foundation, N=14), we also added Kolmården Zoo (N=20) given observed differences in activity levels between the chimpanzees at LUPRSF and MONA. Finalising data analysis required more time than originally planned as 1) data volume greatly exceeded expectations: we found 3093 rhythmic bouts, compared to expecting hundreds; 2) complexity of data analysis increased, as 7 relevant variables were added besides antecedents and consequences; 3) the whole dataset is double coded by independent coders, given topic and methodological novelty. A first pass of data analysis has been completed for the whole dataset, and a second pass is completed for 560 bouts. Full study completion is expected by summer 2023.
The preliminary results obtained so far provide many valuable insights for the evolutionary understanding of musicality, showing that rhythmic behaviours are frequent in chimpanzees and exhibit musicality-relevant features, such as periodicity, rhythmic engagement, and presence in both sexes. Rhythmic behaviours were found primarily in socio-communicative situations and exerted social influence on conspecifics. This is consistent with evolutionary theories that view musicality as an inherently social phenomenon. Our results, however, seem to capture an earlier stage in this evolutionary process, i.e., a stage when rhythmic signals were widespread in the social interactions of ape species that existed before the split between the chimpanzee and human lineages. Rhythmic behaviours were also found in the context of solitary play, where they exhibited ‘music-like’ features, such as periodicity, intentionality, and immersion into playful rhythm production. For ensuing publications and new research ideas, please see the publication list.
2. EXECUTION STATUS FOR DELAYED STUDIES
2.1. EYE-TRACKING STUDIES
Two major studies are included here, one aimed at determining if 1) chimpanzees are capable of ‘gaze entrainment’ (i.e., synchronising gaze shifts to rhythmic stimuli) and if 2) chimpanzee perceive meter structure (i.e., exhibit pupil dilation when exposed to meter deviants). The design, stimulus production and literature review for the eye-tracking studies are completed, but data collection was delayed by technical issues (detailed further below) and by the pandemic. We assess, however, that study completion is feasible despite a lack of direct funding, since all the preparations are in place. In addition, the Department of Philosophy co-funds a research assistant to offload our (unpaid) time to finalise 1.3., which also increases our possibility to finalize 2.1.
To test “gaze entrainment” we have produced 128 short films that represent rhythmic and non-rhythmic sequences in a social and non-social context. In the social context, an experimenter taps rhythmic patterns on fruit pairs, while sounds are emitted simultaneously with the taps. In the non-social condition, the taps are ‘performed’ by a star that appears on top on each fruit in the same rhythmic patterns as in the social condition. The stimuli vary further with respect to structural arrangements, timbres and tempi in order to test for a) the generalisation of gaze entrainment across tempi and timbres; b) the capacity of chimpanzees to internalise more complex rhythmic patterns, such as those composed of three or four subunits.
For testing “meter perception” we use an odd-ball paradigm where random deviations or omissions are applied to a standard quaternary rhythmic pattern. The study includes three conditions: auditory only (short soundtracks), visual only (short films), audio-visual (soundtracks combined with short films). A total of 24 odd-ball manipulations (eight / condition) have been created. The rhythmic patterns are instantiated by percussion sounds (bells) in the auditory condition, and by pulsating circles in the visual condition.
For new ideas generated by these planned studies, see Forthcoming publications in the publication list.
THE UNFORSEEN TECHNICAL ISSUES mentioned above were caused by a major hardware upgrade of the eye-tracking system at LUPRSF. Prior to the start of the project, the technical requirements were in place. LUPRSF was equipped with a purpose-built eye-tracking system (SmartEye), which was integrated with a software for stimulus presentation and data analysis (GazeTracker). In our work, we are dependent on SmartEye as a system provider, since SmartEye systems implement head profiles. These are built from tagged head poses of each study subject, and allow the system to capture gaze behaviours. Regular eye trackers that require calibration are not suitable for research with non-verbal subjects, since it is not possible to instruct them to attend to specific areas on the presentation screen for a given amount of time. The required head profiles for the LUPRSF chimpanzees had been built in the year prior to project start.
Soon after project start, however, SmartEye informed us that our system was too old for receiving updates and technical assistance and had to be replaced with a newer one. A substantial discount was offered for the old system, which was turned in in the fall of 2018. On-site tests were performed by SmartEye technicians (Autumn 2018), and a new system was delivered in April 2019. With the new system, however, it was impossible to build reliable head profiles for chimpanzees. We corresponded intensely with SmartEye, and tested several unsuccessful solutions. In February 2020 SmartEye arrived on-site to collect material and troubleshoot. Subsequent tests were hampered by pandemic-related restrictions but were restarted in the fall. After long insistences that inbuilt restrictions may hinder tracking facial features that would be aberrant by human standards (such as chimpanzee faces), SmartEye inspected the software code (April 2021). Code errors were found, and a new version of the software was delivered.
By June 2021 we could finally build reliable head profiles but encountered a new major technical issue: the new SmartEye software was incompatible with GazeTracker (the stimulus presentation and data analysis software). We thus secured funding and acquired a new software suite recommended by SmartEye, - MAPPS-suite (licensed by EyesDX), which was delivered in the fall of 2021. At first, licensing errors prevented us to access the software, but these were resolved within a month. We then found that only one component of four was delivered and receiving the missing components took another month. The software, however, did not perform as expected and an inspection of the code revealed encompassing issues. Since then, we have corresponded with an EyesDX programmer, to virtually re-develop components of the MAPPS suite. On September 15th, 2022, we could finally confirm reliable eye tracking and reliable integration of the SmartEye software with the MAPPS software at all required levels. We thus expect to initiate data collection for the eye-tracking studies in the coming months.
2.3. RHYTHMIC ENGAGEMENT AND SYNCHRONISATION IN INTERACTIONS BETWEEN HUMANS AND CHIMPANZEES
Planned for 2020 and then 2021, the study could not be carried out due to pandemic-related restrictions. Due to the resource demands of the study, it will not be possible to conduct it in the future without dedicated funding for salary and travel costs.
3. DISSEMINATION
The project and its preliminary results have been presented in conferences (see publication list), and during the workshop – a major internationalisation event.
MAIN RESULTS
1.1. FUNCTIONAL AND MECHANISTIC THEORIES IN THE EVOLUTION OF HUMAN RHYTHM COGNITION – Theoretical synthesis paper
The study has been published and it has received positive feedback, being described as a "wonderfully inclusive survey, which will be a very useful reference for the field" and as an "impressively wide-ranging and thorough review of theory and research regarding the evolution of rhythmic processing in the context of human musicality". Besides providing a comprehensive review of the field, the study also proposes an encompassing research agenda of previously neglected aspects.
1.2. NATURE BEATS: THE WHAT, WHY, AND HOW OF ANIMAL RHYTHMIC BEHAVIOUR – Workshop and book
Research carried out in the project revealed the need to bridge relevant areas of studies. Given an invitation from Springer Nature, we decided to address this in a collective volume on the topic, which was planned as the output of a workshop to be organised at Lund University Primate Research Station Furuvik (LUPRSF) in October 2020 but postponed until May 2022 due to pandemic-related restrictions (https://portal.research.lu.se/en/activities/nature-beats-the-what-why-and-how-of-animal-rhythmic-behaviours). The workshop gathered leading names in the field and revolved around six pre-established themes derived from the areas of expertise of the participating researchers. The event was highly appreciated by participants and a networking success. The collective volume – an originally not planned project output - is the first book to systematically address rhythmic phenomena in animals and humans from multiple perspectives and disciplines. The volume gathers contributions from leading researchers in the field and is the outcome of a workshop organised as part of the project.
1.3. RHYTHMIC BEHAVIOURS IN CHIMPANZEES – Observational study
Data collection for the study was completed in February 2020. Besides the two initially planned sites (LUPRSF, N=6; MONA Foundation, N=14), we also added Kolmården Zoo (N=20) given observed differences in activity levels between the chimpanzees at LUPRSF and MONA. Finalising data analysis required more time than originally planned as 1) data volume greatly exceeded expectations: we found 3093 rhythmic bouts, compared to expecting hundreds; 2) complexity of data analysis increased, as 7 relevant variables were added besides antecedents and consequences; 3) the whole dataset is double coded by independent coders, given topic and methodological novelty. A first pass of data analysis has been completed for the whole dataset, and a second pass is completed for 560 bouts. Full study completion is expected by summer 2023.
The preliminary results obtained so far provide many valuable insights for the evolutionary understanding of musicality, showing that rhythmic behaviours are frequent in chimpanzees and exhibit musicality-relevant features, such as periodicity, rhythmic engagement, and presence in both sexes. Rhythmic behaviours were found primarily in socio-communicative situations and exerted social influence on conspecifics. This is consistent with evolutionary theories that view musicality as an inherently social phenomenon. Our results, however, seem to capture an earlier stage in this evolutionary process, i.e., a stage when rhythmic signals were widespread in the social interactions of ape species that existed before the split between the chimpanzee and human lineages. Rhythmic behaviours were also found in the context of solitary play, where they exhibited ‘music-like’ features, such as periodicity, intentionality, and immersion into playful rhythm production. For ensuing publications and new research ideas, please see the publication list.
2. EXECUTION STATUS FOR DELAYED STUDIES
2.1. EYE-TRACKING STUDIES
Two major studies are included here, one aimed at determining if 1) chimpanzees are capable of ‘gaze entrainment’ (i.e., synchronising gaze shifts to rhythmic stimuli) and if 2) chimpanzee perceive meter structure (i.e., exhibit pupil dilation when exposed to meter deviants). The design, stimulus production and literature review for the eye-tracking studies are completed, but data collection was delayed by technical issues (detailed further below) and by the pandemic. We assess, however, that study completion is feasible despite a lack of direct funding, since all the preparations are in place. In addition, the Department of Philosophy co-funds a research assistant to offload our (unpaid) time to finalise 1.3., which also increases our possibility to finalize 2.1.
To test “gaze entrainment” we have produced 128 short films that represent rhythmic and non-rhythmic sequences in a social and non-social context. In the social context, an experimenter taps rhythmic patterns on fruit pairs, while sounds are emitted simultaneously with the taps. In the non-social condition, the taps are ‘performed’ by a star that appears on top on each fruit in the same rhythmic patterns as in the social condition. The stimuli vary further with respect to structural arrangements, timbres and tempi in order to test for a) the generalisation of gaze entrainment across tempi and timbres; b) the capacity of chimpanzees to internalise more complex rhythmic patterns, such as those composed of three or four subunits.
For testing “meter perception” we use an odd-ball paradigm where random deviations or omissions are applied to a standard quaternary rhythmic pattern. The study includes three conditions: auditory only (short soundtracks), visual only (short films), audio-visual (soundtracks combined with short films). A total of 24 odd-ball manipulations (eight / condition) have been created. The rhythmic patterns are instantiated by percussion sounds (bells) in the auditory condition, and by pulsating circles in the visual condition.
For new ideas generated by these planned studies, see Forthcoming publications in the publication list.
THE UNFORSEEN TECHNICAL ISSUES mentioned above were caused by a major hardware upgrade of the eye-tracking system at LUPRSF. Prior to the start of the project, the technical requirements were in place. LUPRSF was equipped with a purpose-built eye-tracking system (SmartEye), which was integrated with a software for stimulus presentation and data analysis (GazeTracker). In our work, we are dependent on SmartEye as a system provider, since SmartEye systems implement head profiles. These are built from tagged head poses of each study subject, and allow the system to capture gaze behaviours. Regular eye trackers that require calibration are not suitable for research with non-verbal subjects, since it is not possible to instruct them to attend to specific areas on the presentation screen for a given amount of time. The required head profiles for the LUPRSF chimpanzees had been built in the year prior to project start.
Soon after project start, however, SmartEye informed us that our system was too old for receiving updates and technical assistance and had to be replaced with a newer one. A substantial discount was offered for the old system, which was turned in in the fall of 2018. On-site tests were performed by SmartEye technicians (Autumn 2018), and a new system was delivered in April 2019. With the new system, however, it was impossible to build reliable head profiles for chimpanzees. We corresponded intensely with SmartEye, and tested several unsuccessful solutions. In February 2020 SmartEye arrived on-site to collect material and troubleshoot. Subsequent tests were hampered by pandemic-related restrictions but were restarted in the fall. After long insistences that inbuilt restrictions may hinder tracking facial features that would be aberrant by human standards (such as chimpanzee faces), SmartEye inspected the software code (April 2021). Code errors were found, and a new version of the software was delivered.
By June 2021 we could finally build reliable head profiles but encountered a new major technical issue: the new SmartEye software was incompatible with GazeTracker (the stimulus presentation and data analysis software). We thus secured funding and acquired a new software suite recommended by SmartEye, - MAPPS-suite (licensed by EyesDX), which was delivered in the fall of 2021. At first, licensing errors prevented us to access the software, but these were resolved within a month. We then found that only one component of four was delivered and receiving the missing components took another month. The software, however, did not perform as expected and an inspection of the code revealed encompassing issues. Since then, we have corresponded with an EyesDX programmer, to virtually re-develop components of the MAPPS suite. On September 15th, 2022, we could finally confirm reliable eye tracking and reliable integration of the SmartEye software with the MAPPS software at all required levels. We thus expect to initiate data collection for the eye-tracking studies in the coming months.
2.3. RHYTHMIC ENGAGEMENT AND SYNCHRONISATION IN INTERACTIONS BETWEEN HUMANS AND CHIMPANZEES
Planned for 2020 and then 2021, the study could not be carried out due to pandemic-related restrictions. Due to the resource demands of the study, it will not be possible to conduct it in the future without dedicated funding for salary and travel costs.
3. DISSEMINATION
The project and its preliminary results have been presented in conferences (see publication list), and during the workshop – a major internationalisation event.