Comparative psychology and cooperative cognition in animals

Get Complete Project Material File(s) Now! »

Chapter 3 Testing prosociality in kea


Kea (Nestor notabilis) were presented with a series of conditions designed to examine whether they exhibit prosocial tendencies and in particular, reciprocity. Two out of four kea had an initial preference for the prosocial token which rewarded both themselves and a partner. While prosocial token preference did increase for three out of four kea when they alternated taking turns with a partner, it did not decrease in a subsequent control condition in which the experimenter replicated the token choice made by the partner in the previous alternating trials. Finally, prosocial token choice only decreased for two kea and actually increased for one in an asocial control condition where no partner was present. I found no evidence of reciprocity in kea and at present, it remains unclear whether two of the four kea exhibited spontaneous prosocial choice in condition 1, or if this preference was attributable to other factors. Due to our small subject pool and subject’s prior inexperience with token exchange paradigms, further research needs to be undertaken before conclusions can be made about whether the selective pressures faced by kea facilitated the evolution of prosociality in this species.


Prosocial behaviour, whereby an individual confers benefits to another at no cost to themselves is ubiquitous in humans and is likely to have played a role in the evolution of human cooperative behaviour (Amici et al., 2014; Horner et al., 2011). Animals also demonstrate behaviours indicative of prosocial tendencies in naturalistic settings, such as mutual grooming, sharing food, assistance in conflict and cooperative hunting (de Waal et al., 1997; Koyama et al., 2006; Schino, 2007). Prosociality in animals has typically been examined using two different paradigms, the Giving Assistance Test (GAT) and the Prosocial Choice Test (PCT). In the GAT subjects must decide between giving instrumental help to another and doing nothing. GAT studies with chimpanzees have demonstrated that chimpanzees were willing to provide assistance to a conspecific by giving a conspecific a tool they required to access a reward (Yamamoto et al., 2009; Yamamoto et al., 2012), unlocking a door that prevented a partner from accessing a food reward (Warneken et al., 2007), releasing food so it could be reached by a conspecific (Melis et al., 2011) and pulling a handle that helped the conspecific access the reward (Greenberg et al., 2012).
In PCT tasks subjects must choose between two options, a selfish option (1/0) which only yields a reward for the subject and a prosocial option (1/1) that yields a reward for both subject and partner. For the most, PCT tasks have examined prosociality in animals using tokens or platforms (Amici et al., 2014). In platform tasks, subjects can choose between either a prosocial platform containing rewards for both subject and partner or a selfish platform which only contains rewards for the subject (Amici et al., 2014). Similarly, token tasks typically require subjects to choose between two tokens, one selfish and one prosocial. Selection of the selfish token only confers a reward to the subject, whereas selection of the prosocial token confers a reward to both subject and their partner. Prosociality studies with chimpanzees using platform tasks have found little evidence of other regarding preferences in chimpanzees (Silk, 2005; Jensen et al., 2006; Vonk et al., 2008). In similar platform tasks, Cronin et al., (2009) and Stevens (2010) reported that cottontop tamarins did not behave prosocially towards group members, including long-term mates. In contrast, platform studies with long tailed macaques (Massen et al., 2011) and capuchin monkeys (Lakshiminarayanan et al., 2008; Takimoto, Kuroshima & Fujita, 2010) reported that these species were sensitive to the outcomes received by others.
One of the first token exchange paradigms examining prosocial behaviour in animals was conducted by de Waal and colleagues (de Waal et al., 2008) in capuchins. They found that capuchins had a significant preference for the prosocial token and concluded that capuchins possessed other regarding preferences. However, this study was subsequently criticized for a number of reasons (House, et al., Prosociality 2014; Jensen et al., 2006; Amici et al., 2014). Firstly, subjects were only required to choose between one of two tokens and so it was argued that the presence of only two possible choices might create a bias for a particular token. Furthermore, this study did not have any control conditions such as a partner absent condition. This made it impossible to discern whether the capuchin’s preference for the prosocial token was motivated by a prosocial preference or simply a preference for the token that conferred more visible rewards (Amici et al., 2014). Finally, rewards were always visible to both participants which may have masked prosocial tendencies and fostered competitive and selfish attitudes by subjects towards partners (Horner et al., 2011; Amici et al., 2014).
Horner et al., (2011) examined prosociality in chimpanzees with a token exchange experiment designed to avoid some of methodological pitfalls present in the capuchin study by de Waal et al., (2008) and other prosociality experiments (Massen et al., 2010). Rather than presenting chimpanzees with only two tokens at a time to choose from, Horner et al., (2011) presented subjects with 30 tokens in a bucket, 15 of each colour and did not make rewards were visible to both subject and partner. This study also presented subjects with partner-absent control conditions. Chimpanzees chose the prosocial token significantly more than the selfish token suggesting that, similar to the results found in previous GAT studies, chimpanzees may have other regarding preferences.
Based on the study conducted by Horner and colleagues (Horner et al., 2011), Suchak and de Waal investigated prosociality in capuchin monkeys. Previous studies with capuchins have already suggested that they possess prosocial tendencies (de Waal et al., 2008; Lakshiminarayanan et al., 2008; Takimoto et al., 2010) and the authors sought to investigate if this increased when subjects took turns and whether this increase might be based on contingent, or non-contingent reciprocity (Suchak de Waal, 2012). Their study included two new conditions, an alternating turns condition and a yoked control condition, which was designed to assess whether an increase in prosocial preference in the alternating turns condition was due to prosocial behaviour or simply due to the rewards conferred by the prosocial token. These conditions allowed researchers to directly test whether capuchins, like humans, demonstrate contingent reciprocity, which requires a repertoire of cognitive skills such as skills of memory, temporal discounting and numerical discrimination (Stevens et al., 2005; Stevens & Hauser, 2004). Suchak and de Waal found that the capuchins prosocial tendencies increased significantly when they alternated taking turns with their partner. However, token choice was not contingent upon the choice that their partner had made in the previous trial, suggesting subject’s choices were not based on some sort of tit for tat mental scorekeeping scenario like contingent reciprocity. The authors suggested that situations that involve joint action enhance prosociality in capuchins and that subjects could produce outcomes similar to that produced by contingent reciprocity through non-contingent reciprocity, without possessing the complex cognitive mechanisms required for contingent reciprocity (Suchak & de Waal, 2012).

Chapter 1 Introduction
1.1 Comparative psychology and cooperative cognition in animals
1.2 The kea
1.3 Examining cooperative cognition in kea
Chapter 2 Kea perform similarly to elephants, chimpanzees and humans across a range of  cooperative tasks
2.1 Introduction
2.2 Method
2.4 Results
2.5 Discussion
Chapter 3 Testing prosociality in kea 
3.1 Introduction
3.2 Method
3.3 Results
3.4 Discussion
Chapter 4 Kea do not show evidence of inequity aversion in a classic token exchange paradigm 
4.1 Introduction
4.2 Method
4.3 Results
4.4 Discussion
Chapter 5 Concluding remarks 
5.1 Summary of the main findings
5.2 Cooperative cognition in kea and implications for comparative psychology
5.3 Future directions
5.4 Final remarks
Cooperative Cognition in kea

Related Posts