The Social Role of Inhibition in Primates and Other Animals, by Robin Dunbar & Susanne Shultz
Abstract
Posted March 19, 2022
The capacity to inhibit prepotent actions (inhibitory self-control) plays an important role in many aspects of the behaviour of birds and mammals. Although a number of studies have used it as an index of foraging skills, inhibition is, in fact, also crucial for maintaining the temporal and spatial coherence of bonded social groups. Using two sets of comparative data, we show that, in primates, the capacity to inhibit behaviour correlates better with the demands of social contexts than the demands of foraging contexts, whereas a more generalised capacity for causal reasoning correlates with both social and foraging contexts.
In addition, we confirm the Passingham-Wise Conjecture that the capacity to inhibit prepotent action is unique to anthropoid primates, and suggest that it may be crucial for the spatio-temporal integrity of their unique bonded social groups.
1. Introduction
The capacity to inhibit prepotent responses (self-control) is an important cognitive skill that plays a crucial role in many contexts. In humans, for example, it is a strong predictor of both anti-social behaviour as well as the stability of romantic relationships. It also appears to be neurophysiologically demanding in that, in mammals at least, species differences in the ability to exercise self-control covary with brain size.
Many studies have emphasised a role for inhibition in foraging: when animals forage optimally, they have to be able to bypass a less valuable immediate reward in order to gain a more valuable future one.
However, inhibition can also be important in a social context for species that live in bonded social groups. These groups face severe coordination problems during foraging. In order to maintain group cohesion, individuals need to be able to inhibit the inclination to continue feeding when others want to rest (or rest when others want to feed). Failure to do so results in groups becoming scattered, and eventually fragmenting.
For species that form aggregations (unstable flocks or herds), differences in the rate of gut fill result in animals’ time budgets getting out of synchrony, causing groups to fragment and disperse on a timescale of hours. Joiner-leaver models remind us that, whereas small groups typically attract members, large groups lose them, especially when food patches become exhausted or the stresses of group-living become too great. In such groups, individuals maintain weak (or simply casual) relationships with other members of group.
With no centripetal force to maintain group cohesion, animals simply drift apart such that group membership varies on a daily basis. Even in herd-forming primates like the gelada, herds are increasingly likely to fragment during the day’s foraging as their size increases. In contrast, species that form stable social groups (congregations) that travel and sleep together are demographically stable over many years (subject to births and deaths). These groups, individuals have strong bonds with each other that create networks of interconnectedness. Such groups, which are particularly characteristic of primates, face significant challenges because they have to maintain cohesion in the face of the pressures that cause aggregations to fragment.
The central problem that bonded groups face is the risk that individuals’ activity budgets get out of synchrony, since that leads to groups dispersing. Mechanisms are needed to suppress the tendency for an individual to continue with an activity (e.g. feeding) when the rest of the group wants to do something else. In socially flocking weaver birds (Quelea spp.), for example, the pressure to go to roost with the rest of the flock results in low-ranking individuals preferring to lose weight by staying with the flock rather than continue feeding unhindered at an abundant food source from which they had previously been excluded by other group-members. In bonded groups, animals also need to minimise actions that will destabilise groups and cause individuals to leave the group. This requires individuals to be willing, for example, to suppress the desire to steal each other’s food and to avoid escalating agonistic contests.
So far, however, no attempt has been made to determine whether the capacity to inhibit behaviour is more important in the social or the ecological domain, or equally important to both.
The answer bears on the debate over domain-general versus domain-special cognition, and hence whether cognition itself is under selection. To determine whether self-control is more important in the social or the ecological (foraging) domains, we analyse data on performance on two widely used inhibition tasks (Go/No-Go task and A-not-B task) sourced from two separate databases for a range of primate species.
MacLean et al. also used performance on an additional task (the cylinder task) as an index of self-control. Although widely used as a test of self-control, doubts have been expressed as to whether this task actually indexes inhibition. The cylinder task asks animals to choose between two ends of a tube in order to access a food reward, one of which is more accessible than the other (either because one end is blocked or because the reward is nearer one end). This suggests that it might, in fact, be better characterized as a causal reasoning task. If so, it may be a proxy for generalised reasoning rather than specialised cognition, and may hence be used in a wider range of contexts than is self-control.
As indices of foraging demand, we use the percentage of fruit in the diet and the size of the home range (or territory), both of which have frequently been used to test similar hypotheses. Fruits are much less predictable than foliage, and are usually viewed as being a cognitively more challenging diet. They also vary spatially in both quality and quantity, providing the ideal conditions for optimal foraging decisions.
Similarly, large home ranges are assumed to be cognitively demanding in terms of the mental mapping skills and the fact that foraging animals have to choose between near and distant locations on the basis of profitability. In primates, both percent fruit in the diet and range size are strongly influenced by habitat conditions and hence impact on nutrient acquisition. If inhibition relates to foraging efficiency, performance on such tasks should correlate positively with one or both of these ecological indices: the more patchy the resource distribution, the more animals will need to be able to inhibit the temptation to stay with a proximate resource rather than delay reward to go to a richer one that is further away.
In respect of the social domain, our focus is on the stresses that coordination problems impose on group stability and the role that inhibition is likely to play in ensuring that animals stay together during foraging. For social species, the larger its groups and the longer its day journeys the more important inhibition will be for maintaining group cohesion As social domain indices, we use mean social group size and mean day journey length. Coordination problems increase as a function of both group size and the distance animals have to travel after leaving a point of convergence (e.g. a sleeping site or waterhole) since both make it more likely that individuals’ activity cycles will get out of synchrony. Baboons provide a well studied example: group fragmentation is more likely as both day journey length and group size increase, with these effects acting independently. If inhibition is primarily a social skill that influences group cohesion, it should correlate positively with one or both of these indices.
It is important to be clear about the difference between range size and day journey length here since, viewed superficially, both look like foraging-relevant variables. However, functionally they are very different, and especially so for primates. Primates do not forage randomly in their environment, but rather move from one resource patch to another, often at some considerable distance. Range size determines the number of patches available to the group, but does not, of itself, determine the number of patches visited each day or the length of the day journey. Day journey length, by contrast, is a consequence, not a determinant, of the size of the group, the distribution of resource patches and the number of patches the group has to visit to satisfy its collective nutritional demand. In other words, range size defines the distribution of food sources that animals can choose between and hence the choices they make on where to forage, whereas day journey length is simply the means of achieving the ecological end of visiting the required number of patches (but not which patches to visit). The first is a resource choice issue, the second a routing issue and it is only the second that has significant implications for maintaining group cohesion.
… Passingham & Wise argue that, in primates, inhibition depends explicitly on the brain’s frontal pole, a brain unit that is found only in anthropoid primates.
[Footnote: Passingham, R.E. & Wise, S.P. The Neurobiology of the Prefrontal Cortex: Anatomy, Evolution and the Origin of Insight, Oxford University Press 2012]
The significant role of the prefrontal cortex (and, in particular, the ventrolateral prefrontal cortex) in supporting inhibition is confirmed by neuroimaging and lesion studies in humans. Since, with the exception of a few (mostly species-poor) orders (equids, delphinid, tylopods), large bonded social groups are found only in anthropoid primates, we also test the derivative hypothesis that the capacity to inhibit behaviour will be more highly developed in anthropoid primates than in other mammalian orders. In contrast, to the extent that the cylinder task indexes causal reasoning (and hence a form of generalised cognition), we might expect competence on this task to occur more widely outside the primates. …
4. Discussion
We have shown, using two very different inhibition tasks from different databases, that inhibition (self-control) is closely correlated with two key variables that affect group coordination (group size and day journey), but not with either of the two explicitly ecological indices (percentage of fruit in the diet and home range size).
This suggests that the capacity to inhibit prepotent responses has less to do with foraging per se than with the demands of maintaining group coordination whilst foraging.
This concurs with human evidence that the ability to inhibit gratification strongly predicts social skills, with poor capacity to do so being directly related to disruptive anti-social behaviour and poor ability to maintain stable relationships, both of which are indicative of an inability to negotiate compromises.
In primates, including humans, this ability is particularly associated with units in the prefrontal cortex (notably the frontal pole and the adjacent inferior frontal cortex), the part of the brain that expanded most during the course of primate and human evolution – and is most dependent for full adult functionality on experience and learning during socialisation.
The causal reasoning task (the cylinder task) gives more ambiguous results, loading with diet (and hence food-finding abilities) in some cases but on both factors in other cases. The lack of any clear difference in performance on this task between the mammal and bird taxa reinforces the suggestion that the cylinder task is a generalised cognitive skill whose primary function is related to foraging rather than specifically to complex social decision-making.
Nonetheless, it is worth observing that, in general, birds and prosimians performed less well on the cylinder task than rodents, carnivores and anthropoid primates. The latter probably engage in far more manipulation (or behavioural prediction in the case of carnivore hunters) and processing of their food than the former. Although several authors have claimed that birds perform just as well as apes and monkeys on this task, the species concerned (parrots, corvids, passerines) are all ones with large brains (for birds) and well known for their sophisticated cognitive abilities and the capacity to manipulate food items. They are not a statistically random sample of the birds. Nonetheless, this result adds weight to the suggestion that inhibition and causal reasoning have evolved independently of each other, and represent a clear case of mosaic evolution in cognitive skills and their underlying neural bases.
The mediation analysis makes it clear that by far the best model of the causal relationships between the three variables in the social cluster is that capacity to inhibit prepotent actions determines group size, and group size then determines day journey length. Biologically, this makes more sense than any alternative model. This suggests that self-control plays a crucial role in managing group size and its constituent social relationships rather than influencing day journey length directly.
These results feed into the longstanding distinction drawn between species that have stable social groups (congregations) and those that live in unstable herds (aggregations, or fission-fusion social systems). The former are characterized by intense affiliative relationships between individuals, mediated in primates by social grooming and the constant monitoring of social partners.
[MGH: Dunbar indicates elsewhere that relevant human substitutes for grooming can include campfires, laughter, singing, talking, and religious ritual.]
These kinds of bonded relationships are characteristic of anthropoid primates, but otherwise are found only and only a handful of other mostly species-poor mammalian orders which live in modest-sized groups (notably elephants, equids, tylopods, and delphinids) and in the monogamous pairbonds of other mammalian and avian orders. The capacity to inhibit and modulate behaviour is crucial for the continued viability of bonded social groups just as it is for pairbonds in that it ensures that the individuals remain together and hence synchronise their movements.
[Our data/argument] confirms that this capacity is unique to the anthropoid primates, as suggested by Passingham and Wise, and is probably associated with the fact that, at least for the present sample of mammals, bonded social groups are uniquely characteristic of this taxon. Of the non-anthropoid species studied … only elephants have bonded sociality above the level of monogamous pairbonds; however, elephants have a fission-fusion social system that does not depend on maintaining cohesion in large social groups, which may explain why, uniquely, they scored poorly on the inhibition task.
The level of coordination required to maintain the coherence of bonded groups is also likely to be associated with other specialised forms of cognition, such as the ability to understand other individuals’ intentions, the ability to realise the consequences of one’s actions, the ability to plan ahead and the ability to persuade others to adjust their behaviour, all of which also seem to be dependent on the frontal pole.
In some Old World monkeys, for example, individuals make explicit bids, or suggestions, about direction of group travel (often signalled by specific behaviours), with other group members then ‘voting’ on their preferences in order to arrive at a consensus. The capacity to infer the intentions of the signaller and to interpret the meaning of a signal is dependent on mentalising, a cognitive skill that is also confined to the anthropoid primates.
In humans, mentalizing skills of this kind are correlated both with the size of an individual’s social network and with the volume of the brain’s mentalising and default mode neural networks, a brain connectome involving both the frontal, parietal and temporal lobes and the limbic system and their substantial white matter connections that humans share with at least the cercopithecine monkeys. A likely explanation is that both mentalising and inhibition are required to maintain bonded social groups.
[You have now reached the end of this Social Science Files exhibit.]
The Source of today’s exhibit has been:
R.I.M. Dunbar and Susanne Shultz, The Social Role of Inhibition in Primates and Other Animals
[Posted on bioRxiv March 19, 2022]
doi: https://doi.org/10.1101/2020.10.26.354852
Social Science Files displays multidisciplinary writings on a great variety of topics relating to evolutions of social order from the earliest humans to the present day and future machine age.
‘The Heller Files’, quality tools for Social Science.
Today’s ‘horizon’ nomad days photo was taken by MGH from the tent-door (2017):