To examine autistic traits in chimpanzees, we must consider the idea of social structure in detail. Something that humans share with many primate species is group living, and the social interactions that come with it. The structures and organizations of primate social systems are complex and vary widely [25]: The fission–fusion groups of chimpanzees and capuchins; the semi-solitary lifestyle of orangutans; the gorilla harem and pair-bonded douroucoulis; gelada and baboon troops made up of multiple subunits. These are just a few examples of broad variation in social structure. All social groups form a web of interlinked components, including group size, male-to-female ratio, mate choice, mating system, sexual dimorphism, male–female dispersal, and social hierarchy.

Ecological and behavioral constraints determine individual interactions that form a group, but the group, in turn, constrains individual behavior [24]. For example the costs of social living increase with group size as larger groups experience higher levels of competition and aggression [26]. Social structure also determines behavioral differences between the sexes. South American squirrel monkeys (Saimiri spp.) for example, are distinct in their social structure: S. boliviensis live in larger, female dominated groups, where females form coalitions and show more aggression to each other than in S. sciureus [27]. However, in S. sciureus, males, instead of females, form coalitions, and show more aggression to each other than in S. boliviensis [28].

Social interactions are thus dependent on social structure. Recent research addressed the benefits to individuals of cooperative behavior. Sussman and Garber [29] showed that across 51 species, affiliative behavior contributed up to 91.7 % of social interactions. By comparison, antagonistic behavior contributed to less than 1 %. Where spatial foraging allows for lower feeding competition, the benefits of social interaction are higher to the individual, and become a focal consideration of group structure [29]. Understanding what drives group social structure and cohesion is key to establishing species and population differences in social behavior [24]. Only by defining species-typical behavior can we then address individual and species differences in atypical behavior.

One requirement of group living is information sharing. This is beneficial for communication about predators and food location, but can also lead to kleptoparasitic behavior, a form of information “eavesdropping.” Spider monkeys, for example, use calls when foraging to manipulate group size. Individuals were more likely to emit food calls when there was an abundant resource, and groups of subordinate males called less frequently than groups of dominant males [30]. This suggests that individuals used their knowledge not only of food availability but also of social rank to determine how to respond to resource availability. Since subordinate ranks experience higher interference competition [30, 31] if food is scarce, subordinates will try to minimize group competition by keeping this information to themselves.

Social rank is integral to group communication, and has formed the basis for studies of primate social knowledge. If getting along in a group means knowing your neighbors, then knowledge of “who’s who” is beneficial to individual success in a group. This is especially true in a hierarchy, where dominance rank determines resource gain [30, 31]. Dominance rank in a female baboon hierarchy is determined by data on number of supplants, aggressive, and submissive behavior [32]. Hierarchical structure therefore predicts not only resource use but also social behavior. For example, higher ranking female baboons groomed significantly more with maternal kin than lower ranking females [32]. Social knowledge is a useful tool by which individuals mediate their “privilege” of resource use through monitoring others.

So what does “knowledge” refer to? Knowledge of a conspecific’s whereabouts should rely on relevant sensory information. In many primate species, vocalizations play a role in social interactions. Researchers can take advantage of this to test individual knowledge. Playback experiments in vervets and baboons that live in large troops consisting of smaller family units have shown that females understand and respond to vocalizations of conspecifics as social cues. Seyfarth and Cheney, who studied these social interactions for over 20 years, describe the role of social knowledge to an individual: “to achieve a complete understanding of her society… she must be able to step outside her own sphere of interactions and recognize the relations that exist among others…” ([33]: 209). This “sphere” refers to the female’s immediate matrilineal family, thus suggesting that females must become familiar with all other individuals in their troop. For the female, this involves observing the interactions of other individuals and inferring the appropriate associations. Thus, when she hears an anomalous call sequence, such as a subordinate individual threatening a dominant individual, this call sequence does not fit with the information that female retains on group social rank. Subjects will respond to an anomalous call sequence by looking toward the source for a longer time than when an expected call sequence is played [34].

Other experiments suggest that baboons are able to interpret the direction of threat-grunts from a dominant female as context specific (a threatened versus friendly interaction). They adjust their behavior as to whether they interpret these threats to be directed toward them [35]. These experiments demonstrate the role of social knowledge for group living. They also suggest the role of social inference in these interactions. As in other species, however, there is little evidence that baboons can differentiate between what they know and what others know. This suggests they do not develop a Theory of Mind as humans do. It does, however, suggest that they have a rudimentary ability of perspective taking, that is, a basic level of intersubjectivity.

These findings suggest that social knowledge is an evolutionary precursor to Theory of Mind. This fits with the development of social interaction in early human development. Emotional intersubjectivity [36] is present in infants from 3 months old, who attune to others’ emotions using eye contact. Emotional intersubjectivity is the most basic human social interaction and is likely a precursor to cognitive intersubjectivity, or Theory of Mind. Cognitive perception begins its study in children from 9 months old [37]. At this stage, children develop triadic interactions with other people, and rely on shared visual attention for communicating with others [38]. By the age of four, children understand desires, and are able to attribute false-belief [39]. It is during this stage of development that autism in humans may become apparent, with ToM deficits apparent in the diagnosis of autism in children [10].

Evidence in apes also suggests they demonstrate advanced perspective taking. Gomez [36] offered evidence of emotional intersubjectivity in infant gorillas, such as the use of persistent eye contact and elaborate repertoires to gain the attention of a human playmate. Chimpanzees parallel early human development in gaze following and in distinguishing between directed and non-directed gaze [23]. Chimpanzees can also attribute differences in knowledge between individuals [13]. However, findings suggest chimpanzees do not develop shared attention [23] or attribute false belief [15], as humans do. This suggests that, although other primates differ from humans in their ability to perceive the thoughts and actions of others, they also share some of these abilities. Theory of Mind should thus not be defined as a purely human attribute, but as part of a spectrum of social knowledge, understanding and perspective taking. Using this definition, we can establish the importance of studying these abilities in other primates for understanding neurodevelopment. In using primate social relationships as a model for understanding neurodevelopment, we are trying to establish how autism would manifest itself in other species. We will consider examples of typical and atypical social behavior in primates, and how these traits would fit within the current definition of autism. This definition is categorized into the triad of impairments, and in addition we will examine traits relating to three other aspects of behavior: intersubjective and empathic behavior, sensory sensitivity, and propensity for specific skill or focus in one area.

We now return to our earlier question: If we looked for autistic traits in other primates, what would we look for? For ideas of how to approach such traits, we can draw on the observations of those who have studied great apes in the wild. Chimpanzees are an ideal study species: they are one of our closest relatives, they have been studied extensively in both captivity and the wild [40–42], and much is known about their social behavior and emotional expression [43, 44]. Goodall [43] describes chimpanzees’ broad repertoire of social behaviors, which help to maintain group cohesion, and to reassure and reconcile. These include physical contact, such as patting, kissing, and embracing; allogrooming; sharing food; showing concern towards others; and helping and protecting kin or companions during fights [43]. They also exhibit context-dependent vocalizations, with call production determined by social context such as the sex and rank of the individual [45].

We can group these behaviors into categories that fit within the definition of autism. We have already identified social behavior (social interaction with peers; use of facial expression; sharing) and communicative behaviors (vocalizations; physical contact). Examples of stereotypic behavior are described in situations where individuals experience severe stress. In the wild these examples are specific to infants who have lost their mother, and display behaviors such as rocking, staring, hanging upside down for long periods of time, as well as inappropriate social behavior toward peers and adults [43]. In captivity, long-term effects of social deprivation are exhibited in fewer social interactions and subordinate behavior of deprived individuals [46].

Based on these traits we can hypothesize what behaviors might highlight developmental deficits among chimpanzees. But species differences in social structure, as we have seen, are likely to influence how a trait of “autism” manifests itself. Let us consider these traits in a less social ape. Orangutans are mostly considered solitary except when mating and raising young, or when abundant resources may favor low levels of fission–fusion [47]. Observations of social groupings led Galdikas [48] to describe them as semi-solitary apes. Solitary play is also observed much more in infant orangutans (50 % of observation time) compared to infant chimpanzees (15 % of observation time) between aged 10 and 11 months [49]. Allogrooming (grooming with another) is observed in orangutans, but rarely outside the natal unit [50]. In captive groups, adult and juvenile orangutans are observed to spend approximately 50 % of their time in social interactions. For juveniles this includes 24 % of total time in contact play, and for adults, 10 % of total time allogrooming [51]. Social behaviors amongst adults include hand fondling, touch, arm extension, grabbing, and mouthing. Although social interactions are clearly a part of orangutan behavior, expected deficits in social behavior may be less profound in these more solitary species.

However, communication is a still a key behavior for individuals to interact with mates, young, and rivals. Orangutans have a broad repertoire of communication, including facial expressions, touch and visual signals for such interactions [52]. Thus, deficits of some aspects of social behavior would still be expected in orangutan neurodevelopmental disorders. The record of stereotypic behavior in orangutans is little documented in the literature, reflecting, perhaps, a difference in behavioral focus between species, rather than species-differences in behavior.

Gorillas represent another social species with an alternative social structure. Gorillas live in harems, with females often transferring between groups, resulting in weak social ranks [53, 54]. This may result in the loss of maternal kin bonds, and could explain why few social interactions are observed between unit females [55]. Allogrooming in western lowland gorillas (Gorilla gorilla gorilla) is observed notably less than in mountain gorillas (G. beringei beringei), and is mainly only between mothers and infants. This is not because of an observed difference in social structure. Rather it is considered an effect of resting time and adherence of plant species to hair in the mountain gorillas [55], leading to an increased frequency of allogrooming. These species differences in social behavior are important considerations for predicting behavioral deficits.

Although social interactions do not seem as important to gorilla group life as they do for chimpanzees, as with orangutans, communication is still a necessary part of these interactions. Gorillas are observed to use a variety of context-dependent close calls, and frequency of usage varies with dominance rank of both those giving and receiving the call [56]. Gestural communication has also been observed in captive gorillas [57], and is thought to be used to signal intent. Atypical behavior has been documented in captive gorillas; infants that are hand reared by people show more solitary play and aggression than those reared by their mothers; however, interacting with group members helped to mitigate these effects [58]. Stereotypic behaviors observed in these infants include rhythmic body movements (such as rocking) and finger sucking; the frequencies of these behaviors increased when an individual was under stress, such as when introduced to conspecifics.

We have now established species typical behaviors and how they differ as a function of social structure. From this, we have deduced atypical behavior under the triad of impairments. What about other aspects of autism? Deficits in empathic behavior are harder to address; empirical studies of empathy in primates are few [59–61]. However, observations of concern towards others [43] and perspective taking in chimpanzees [15], would predict lower perceptual abilities with neurodevelopmental deficits.

Savantism, the propensity for a focused skill or hobby, is also difficult to address. We must consider skill use within species’ societies, such as tool use to obtain food or completing a problem-solving task in the lab. As with humans, some apes are smarter than others, referred to as “geniuses” of the ape world [62]. In examining “savant” like behaviors in apes, we may consider not just intelligence, but look for focused, skilled, and repetitive behavior typical to savantism.

A final consideration is sensory sensitivity. Autistic children can be sensitive to a variety of sensory stimuli [63]. In working with autistic children, the author has observed every day sensations, such as touch, loud noises and even wearing clothes, can be distressing for individuals. They can become highly stimulated by detailed movement, such as the movement of a lift or a bag blowing in the breeze. Response to sensory stimuli may be classified as stereotypic, with stimulation involving repetitive behaviors such as hand flapping, rocking, and jumping up and down. In primates, repetitive behavior is seen in response to a lack of environmental stimuli (for a review see [64]). We may consider sensory sensitivity in primates a type of stereotypic behavior, predicting an increase in fearful behaviors as well as a withdrawn interest in the environment, as behavioral deficits.

From the evidence reviewed here, stereotypic behaviors appear to occur under situations of social stress or lack of appropriate social interaction [46, 58]. They have also been mainly documented in ape infants. If we are to establish behavioral parallels between the triad of impairments found in autistic humans, and behavioral deficits in nonhuman primates, care must be taken in how to interpret this behavior. This can be understood in more detail by examining the work of Harlow and his colleagues.

Harlow demonstrated in infant macaques that social deprivation from birth has severe behavioral consequences [65, 66]. Monkeys raised alone exhibited inappropriate aggressive behavior when housed with conspecifics [67], and females raised without mothers specifically showed a lack of maternal behavior. This included disinterest towards the infant, lack of maternal contact, and indifference to having infants removed from their cage, as well as aggressive behavior to the point of infanticide [68]. These studies were the first to address the role that social attachment plays in individual social behavior, and demonstrate more drastically the atypical behaviors observed in wild and captive apes.

Research into deprivation extends beyond behavior observation. A study examining social deprivation in rhesus macaques found structural differences in the brains of socially deprived individuals compared to socialized monkeys [69]. Differences were found in the basal ganglia; in the caudate nucleus, which is implemented in learning, especially feedback processes; in the putamen, which regulates movement; and in the nucleus accumbens, which is involved in reward, addiction, aggression, and fear. These findings suggested that abnormal sensory input in developing infants permanently affected the neurochemical structure of brain regions affected by social and environmental stimuli.

Interpreting atypical behaviors thus advocates caution [64]. Although stereotypic behaviors have been well observed, little is understood of the mechanisms that cause them [70]. A trait that may appear as a neurodevelopmental deficit, such as lack of social interaction or stereotypic behavior, may in fact be the result of environmental deprivation, such as infant isolation or traumatic captive experience. In people it is clear that stereotypic and atypical social behavior can develop from neglect, abuse, or trauma during childhood [71], or as a result of atypical neurodevelopment, as seen in children with autism [6]. In primates, these atypical behaviors clearly develop during socially deprived circumstances. However, it is possible that stereotypies also occur naturally, as a result of neurodevelopmental deficits. We outline the predicted associations between atypical behavior and neurodevelopment in Fig. 1.

Given the potential similarities of behavioral deficits derived from different origins, it is crucial to take individual history into account when examining the manifestation of any such “autistic” trait. Table 1 considers these behaviors in more detail, and documents examples of behavioral deficits in both wild and captive apes. Based upon these behaviors, we can hypothesize what behavioral deficits we might expect to find as a result of neurodevelopmental deficits (see Table 2). Predicted traits are taken from items on the Hominoid Personality Questionnaire (HPQ [72]) and are grouped triadically as in the diagnosis of human autism.