Heather Bond Poje
University of California, San Diego
Anthropology Department
Behavioral coordination and intersubjectivity during grooming interactions in bonobos and baboons.
I. Goals:
Grooming has long been a frequent topic of study among primatologists because of its great social importance to primates and its seemingly ubiquitous nature. However, nearly all studies of grooming focus on the outcome of grooming (who grooms who, what their relative ranks are, whether grooming is reciprocated, etc), rather than the process of grooming itself (e.g. How is grooming initiated? How do the groomer and recipient negotiate shifts in the body part targeted for grooming? If grooming is disrupted, how is it reestablished? How do role switches occur? Why are some grooming sessions more stable than others? Is the behavior that we call “grooming” identical from one species to the next?). My real interest lies in finding a way to better grasp the inter-personal communication and shared understandings that two individuals may experience when sharing an interaction; in humans this is called “intersubjectivity”, though it is unclear whether or not nonhuman primates are capable of this type of shared experience. I feel that understanding the details of a complex social interaction are crucial to our understanding of not only the behavior in question (in this case, grooming) but also in the way that primates relate to one another, through communication and mutual coordination or coregulation.
My goals for this project are:
1. To understand the process of behavioral coordination between participants during grooming in bonobos and baboons. The extent of coordination is, potentially:
a) A reflection of shared understanding of the interaction (i.e. intersubjectivity)
b) An expression of the quality of the relationship.
2. To compare coordination (and hence, intersubjectivity) during grooming in adult
baboons vs bonobos.
3. To explore the development of infant participation and intersubjectivity during
grooming.
My working hypotheses for this project fall into 5 categries:
a) Behavioral Coordination/ Coregulation
Coordination can be seen in:
1. Smooth shifts between grooming targets
2. Infrequent pauses during grooming session
3. Little negotiation during initiations, role switches, and restarts after pauses.
Coordination occurs through:
1. Attention monitoring or sharing and mutual eye contact.
2. Physical contact
3. Communicative gestures and/or sounds
b) Intersubjectivity
Grooming sessions that are sustained with few breaks, even in the midst of other group events (e.g. entrances and exits of other individuals) and that exhibit a high degree of coordination during initiation and/or throughout the bout indicate that groomer and recipient share an understanding that the grooming bout is “about” grooming. This may be “intersubjectivity”.
c) Relationship quality
The quality of a relationship is expressed in the moment-to-moment interaction between individuals. Grooming both establishes relationships and serves as an indicator of the quality of a relationship. Well-coordinated grooming sessions are more likely to occur between pairs with higher quality relationships and who are more frequent grooming partners.
e) Species comparison
Bonobos should exhibit more use of mutual eye gaze during grooming. Other potential differences: body part targeted for grooming (greater focus on grooming the face in bonobos) and in body posture during grooming (lying down rather than sitting up in bonobos). These differences are likely to result in differences in the species’ capacity for intersubjectivity and in the degree to which coordination, especially through attention monitoring, is possible between individuals.
f) Development of behavioral coordination
Coordination in grooming between mother and infant should increase with infant’s age, as should the infant’s ability to participate in grooming in the role of the “recipient”. Infant’s focus of attention during the mother’s own adult grooming interactions should also change over time, with older infants more attentive to the attention and activities of the nearby adults.
II. Methods & Analyses
A. Subjects
Research is based on videotapes of olive baboons (wild, from Laikipia, Kenya) and bonobos (captive, from the San Diego Wild Animal Park). Subjects were videotaped between 2001-2003 (bonobos) and during the summer of 2003 (baboons). Research mainly focuses on grooming between adults (usually adult females but a few adult males are included to increase sample size), but the last phase of the project will be examining the development of infants’ participation in grooming as recipients of grooming from their mothers.
B. Types of Data
This project consists of three phases:
1. Coding and analysis of full-length grooming sessions between adults, at the system, macro, and micro levels of detail. I will be analyzing 10 sessions of grooming each for bonobos and baboons. I am currently working on this phase of the project (I have completed 6 of 10 for each species) and it is this phase I will be discussing in this workshop.
2. Micro-level coding and analysis of grooming initiations and instances of mutual eye contact for both species. This next phase (begins in March) will involve coding eye contact, physical contact, and communicative signaling between future grooming partners just prior to the start of grooming.
3. System, macro, and micro-level coding of mothers grooming infants for both species; the mother-infant behavioral coordination during grooming will be compared with behavioral coordination of two adults in order to identify how the infant’s capacity to participate in the role of “recipient” develops over time.
C. Data collection / Video coding
During phase 1 of the project, for each video, the following information is collected, as well as details about the duration of each grooming bout, disruptions or breaks in grooming,, and overall general details about each video.
1. System
State of the system (classified as grooming, grooming & activity, interaction, interaction & activity, noninteraction, and personnel change; for each of these (except personnel change), system state is either classified as coordinated or in discord regarding the attention of the members of the system).
Relative position—of groomer and recipient to one another (open-open, open-peripheral, open-closed, peripheral-peripheral, closed-closed, etc)
Proximity—of groomer, recipient, and infant (at 30 sec intervals): in physical contact (type of contact specified), within arms reach, within lunge, or beyond.
2. Macro
Activity—for groomer, recipient, infant, and any others in proximity (e.g. rest, groom, scratch, feed, forage, travel, climb, play, object play, clap, solicitation, touch, greet, follow, aggress, avoid, etc).
Posture—for groomer and recipient (e.g. sit, lean, quadrupedal, bipedal, supine, prone, on side).
3. Micro
Focus of attention (head direction)—for groomer, recipient, and infant (partner’s face, partner nonface, grooming target, nongroom activity, infant, other in system, other outside of system, self, external to system, object, camera, none/ground). Incidents of mutual eye contact are specified.
Physical Contact & Gestural communication—for groomer and recipient; short description of physical contact and of gesture/ signal used.
Grooming target (face, head, neck, chest, side, back, rump, upper arm, lower arm, hand, upper leg, lower leg, foot, tail).
Rate of grooming & “tool” (hand or mouth) used (at 30 sec interval)
Frame-by-frame analysis of mutual eye gaze incidents (for 1/2 sec prior and after contact) between groomer and recipient [this coding will occur in the next phase of the project].
D. Data analysis
Data analysis will involve descriptive statistics comparing overall durations of grooming bouts, grooming sessions, frequencies of disruptions, and numbers of group members present. Sequences of behaviors (activities, system states, micro-level communication) will be analyzed through sequential analysis. Additionally, based on initial comparisons, grooming sessions will be classified overall as “coordinated” or “disjointed” and will be analyzed based on chi-square analysis, and correlation matrices and factor analysis will be used to test for interaction with other factors such as rank, kinship, presence of infants or other group members, and group context, to determine which factors influence the overall pattern (this is relevant for comparison to other, more outcome-oriented studies of grooming). Preliminary analysis suggests that communicative signaling and mutual eye contact occurs in both species when negotiating grooming, but that eye contact and grooming of the face occur more frequently in bonobos and that this therefore may have implications for the evolution of attention in different species of primates.
III. Challenges
A. Methodological
• At this point in the project one of my main challenges has been finding adequate samples for coding. Some of the problems are that grooming sessions range in length quite a bit, they may not have complete initiations or follow-through to terminations, the angle of the camera may be less than ideal, and there may be breaks in the video where grooming is temporarily not visible. Unfortunately the biggest problem (unique to this project) is that I am using video that was filmed with a different project in mind and therefore grooming was not prioritized during filming.
• The length of time spent coding is (and, I believe, always will be) one of the primary challenges of doing this type of research. For this project in particular, I’m finding it challenging to balance the time spent collecting the time-consuming micro-level details with the need to code long sessions, as some grooming sessions are up to 30 min long.
• The sample size for my project (and, I suspect, most projects involving microethology) will inevitably be quite small making it difficult to analyze statistically.
•Additionally, comparisons of varying length grooming sessions, composed of different numbers of individuals, will be a hurdle to overcome during analysis of the data.
B. Theoretical
• I struggle with how to know that I’ve identified the differences that make a difference; differences that can be seen intuitively with the naked eye, and therefore should be available on video, are nevertheless difficult to pinpoint when choosing which elements of the video to code. My sample of coded videos is too small (even now) for good preliminary analysis, so I am still mostly working off hypotheses and don’t know yet if it will be successful.
• I have not determined a way to code video in detail beyond the dyad. Although my system-states can encompass multiple individuals, the coding of, say triadic grooming interactions is, unfortunately, in my project coded as simply two pairs of dyads. I hope that in the future I can adequately address this issue, and this is certainly something I would like to hear discussion about in this workshop.
• One of the potentially biggest challenges in coding behavior of multiple individuals with a relatively small sample size is the issue of individual differences and difference s of context (both long term and immediate)—if a difference is identified in the degree of behavioral coordination between different grooming sessions, it will be difficult to know if that is the result of individual or contextual differences, or if it is a potential species-wide difference (unless, as I hope, there is an identifiable widespread pattern).
C. Technological
• Data input is a challenge; I would loved to use integrated video—data collection software but could not find or could not afford one, and therefore I am making do with iMovie and Filemaker. The arrangement, though, is far from ideal.
•I have not yet tackled potential data analysis problems, but there will be several challenges inherent to it, including sample size and individual and contextual differences as mentioned above; presumably the actual data can be exported for analysis fairly easily from Filemaker to most statistical software programs but will certainly require at least some transformation.
D. Sociological
•There is an immense difficulty in displaying data and results, and this lends itself to a difficulty in explaining these ideas and issues to the general public. Ideally I would love to print out an enormous timeline involving simultaneous multiple behavioral modalities, but without that, it becomes quite challenging to convey the nature of the data to others.
• The topic and the methods I am using are not traditional in primatology, and therefore are, unfortunately, easily dismissed as irrelevant. Whether this will lead to problems in the future, both with publishing and when job hunting remains to be seen (but is a great concern of mine!).
I look forward to learning much from all of you as we discuss our research together!