Iterated Insights

Abstract

A growing body of research has identified a naturally occurring subgroup of male rhesus macaques that consistently spends less time interacting with other monkeys, initiates fewer affiliative contacts, and shows measurable differences in processing social information. These animals are not experimentally isolated or genetically engineered to mimic autism. They arise spontaneously within a highly social primate species, and their behavior varies along a continuum from highly social to markedly low-social. Over the past decade, researchers have found that this phenotype is stable across time, detectable early in development, associated with differences in face recognition and reciprocal social signaling, and linked to cerebrospinal-fluid vasopressin. Low-social males also show preserved object-recognition memory, suggesting that the phenotype is selective for social rather than general cognitive functioning. Remarkably, the same vasopressin system has been implicated in human autism-related social difficulties, and pedigree studies in rhesus macaques show a strong paternal familial contribution to social functioning.

These findings are highly relevant to the solitary forager hypothesis of autism. That hypothesis proposed that some autism-associated traits may represent naturally occurring variation in mammalian sociality and that comparable phenotypes might eventually be recognized within other social species. It further proposed that reduced social motivation need not imply generalized incompetence and may, under some ecological conditions, support a more independent style of learning and resource acquisition. The low-social rhesus macaque phenotype now provides a concrete primate example of the kind of variation the hypothesis anticipated. The monkeys show reduced social initiation, altered social recognition and reciprocity, early developmental differences in social attention, and biologically measurable variation in a neurochemical system also associated with autism, while retaining competence in at least some nonsocial cognitive tasks.

The evolutionary significance of this phenotype remains to be determined, but the comparison shifts the question in an important direction. Rather than viewing reduced social engagement only as pathology, these findings make it possible to ask how variation in social motivation, social recognition, and nonsocial cognition is organized within primate populations, how it is inherited, and under what ecological conditions it may impose costs or confer advantages. The rhesus findings therefore strengthen the case for studying autism-associated traits as part of the broader evolutionary neuroecology of sociality and provide a direct comparative framework for testing whether low-social-dependence cognition can represent a viable alternative strategy within a highly social species.

Keywords: autism; rhesus macaque; social motivation; vasopressin; behavioral ecology; individual differences; solitary forager hypothesis; evolutionary mismatch

1. Introduction

The solitary forager hypothesis proposed that some autism-associated traits may have been useful under conditions in which individuals acquired resources independently, interacted within small social networks, or depended less on continuous interpersonal coordination. Persistent interests, repeated practice, attention to physical regularities, and reduced reliance on social reinforcement were considered possible components of such a cognitive strategy. The hypothesis concerned combinations of traits and their ecological consequences, particularly within subclinical and relatively independent presentations, rather than requiring the adaptive value of every feature associated with an autism diagnosis (Reser, 2011). 

A subsequent article proposed a comparative method for investigating this possibility. Naturally solitary mammals and individuals with low social engagement within social species could reveal mechanisms relevant to autism-associated variation. The proposed targets included neuropeptide signaling, affiliative reward, social recognition, responses to social threat, and the motivational processes that influence which information enters sustained attention and working memory. The central prediction was that naturally occurring variation could illuminate these systems without first producing a lesion or introducing a pathogenic mutation into an otherwise highly social animal (Reser, 2014, published online in 2013). 

The rhesus-macaque research program provides a particularly informative test of that comparative proposal. It examines variation within a social primate species and connects directly observed behavior with developmental assessments, quantitative ratings, pedigrees, and neurochemical measurements. The comparison is therefore narrower anatomically and phylogenetically than an analogy assembled from distantly related species.

This article advances an explicit hypothesis: some autism-associated human variation involves differences in the regulation of conserved primate systems for initiating affiliation, recognizing social partners, and responding to social information. The ecological consequences of those differences depend on the demands and opportunities of the environment. The solitary forager hypothesis supplies one possible evolutionary explanation for such variation. Its distinctive prediction is that reduced social dependence can coexist with practical competence and may sometimes confer advantages when independent action has comparatively high returns.

The evidence reviewed below is a narrative synthesis of primary studies, interpreted through that framework. “Autism-like” refers to specified behavioral and biological correspondences, not a diagnosis of autism in monkeys.

Anticipating Autism-Like Variation in Other Social Species

One of the more striking aspects of the recent rhesus macaque findings is that the possibility of finding such a phenotype was explicitly raised in the original solitary forager hypothesis. In discussing the large differences in sociality both between and within animal species, I questioned whether autism-like variation might occur naturally within other highly social mammals, where it could easily go unrecognized.

“Perhaps populations of other social species, such as primates, have an equally low but consistent prevalence of autistic individuals as well.”

The point was not that chimpanzees or other primates should possess the complete human clinical syndrome. The prediction was that if autism partly reflects variation in older biological systems governing social motivation, affiliation, recognition, and attention to conspecifics, then corresponding low-social phenotypes should appear within other social species. Such animals might constitute a persistent minority of the population without ever being recognized as possessing a coherent behavioral phenotype. This is closely related to what has now been documented in male rhesus macaques: stable individual differences in social initiation and engagement occurring within an otherwise highly social primate population.

I also anticipated a methodological reason why such phenotypes might remain invisible. Animal researchers generally did not have an equivalent of a psychiatric diagnostic framework for identifying unusual combinations of social behavior, nor was there necessarily a reason to regard naturally occurring low sociality as a syndrome worthy of separate investigation.

“No formal diagnostic criteria are available for psychiatric or even social disabilities in other animals … but it would be interesting, although difficult, to see if there are analogues, or possibly homologues of autism in other species.”

This is remarkably close to the path that the rhesus research eventually took. The low-social monkeys were not produced experimentally in order to mimic autism. Researchers first identified naturally occurring variation in rhesus social behavior and then began to characterize the animals more systematically. They subsequently adapted the human Social Responsiveness Scale for macaques, identified differences in early face processing and social initiation, discovered the association with cerebrospinal-fluid vasopressin, demonstrated substantial familial structure, and experimentally altered particular social responses using vasopressin. A naturally occurring social variant became recognizable as an autism-relevant phenotype only after researchers developed the tools to measure it.

The later comparative article made the same point more generally by distinguishing variation between species from variation among individuals belonging to the same species.

“Perhaps both intra- and interspecific diversity can be utilized to investigate the autism spectrum; however, the data concerning interspecific diversity is currently much stronger.”

At the time, comparative evidence was indeed much stronger between species. Prairie and montane voles, for example, provided conspicuous contrasts in affiliation, pair bonding, separation responses, and oxytocin and vasopressin systems. What was largely missing was the other side of the prediction: naturally occurring variation within a highly social species. The rhesus work now begins to fill precisely that gap. Sociality in rhesus macaques is continuously distributed, and individuals at the low-social end show a stable constellation of reduced affiliative initiation and altered social-information processing without evidence of a generalized cognitive deficit.

The 2013 article went still further and suggested that the appropriate biological unit might be the entire quantitative distribution of sociality rather than a sharp distinction between normality and disorder.

“Could something similar be true throughout the normal distribution of sociality, including social deficits, ASDs, and other disorders of bonding, attachment, and empathy?”

That question is particularly relevant now. The macaque Social Responsiveness Scale does not reveal two separate populations of normal and autistic-like monkeys. Scores vary continuously. The same is true of directly observed social behavior, and cerebrospinal-fluid vasopressin tracks quantitative social variation rather than merely membership in an experimentally designated low-social group. The emerging picture is therefore one of continuous biological variation in the systems governing primate social engagement.

Seen retrospectively, the rhesus findings are unusually pertinent because they were not evidence used to construct the original hypothesis. Most of this research appeared years later. The earlier papers predicted that autism-like phenotypes might eventually be recognized within other social species, that existing methods might have failed to identify them, and that within-species social variation could ultimately be connected to molecular and neurobiological mechanisms. The low-social rhesus macaque literature now supplies a concrete example of exactly this kind of phenomenon. It does not by itself establish the solitary forager hypothesis, but it provides a substantial and unexpectedly specific comparative confirmation of one of the hypothesis’s original predictions.

Table 1. Behavioral similarities between autism and the low-social macaque phenotype

Behavioral feature	Autism-related pattern	What the macaque studies found
Reduced spontaneous social initiation	Some autistic people initiate fewer interactions, even when they can respond to others.	Low-social males initiated fewer approaches, nonaggressive contacts, and requests for grooming.
More time apart from others	Some autistic people prefer more solitude or engage less frequently with others.	Low-social monkeys spent more time beyond close proximity to other animals. This establishes reduced engagement; preference is an interpretation of the pattern.
Reduced affiliative contact	The frequency and preferred forms of physical or affiliative interaction can differ.	Higher Poor Social Motivation scores predicted less nonaggressive bodily contact and grooming. Grooming is the monkey-specific behavior; the comparison concerns affiliation, not an identical human action.
Difficulty recognizing individual faces	Face-identity recognition is impaired in a subgroup of autistic people, including people without intellectual disability.	Low-social monkeys did not demonstrate the expected face-recognition memory under placebo in the adult experiment; earlier developmental testing also identified face-recognition differences.
Better-preserved nonsocial than social recognition	Social-perceptual difficulties can coexist with preserved nonsocial abilities.	Object-recognition memory was intact in the same low-social animals whose face-recognition performance was impaired.
Atypical reciprocal social signaling	Spontaneous responses to others’ nonverbal communication can be less frequent or differently coordinated.	Under placebo, low-social monkeys did not show the expected selective affiliative response to friendly displays. Vasopressin increased appropriate reciprocal responding.
Context-dependent gaze differences	Gaze can be used differently when interpreting social information.	Infants later classified as low-social showed less gaze aversion to aggressive displays than future high-social animals.
Social awkwardness and reduced social confidence	Some autistic people have difficulty with conventional social timing, presentation, or confidence.	The macaque scale assessed these dimensions. Factor analysis separated social motivation from social attractiveness and inappropriate behavior.
Differences in peer play and solitary play	Participation in conventional peer play may differ; some children engage more independently.	Poor Social Attractiveness scores predicted less observed play. In a separate juvenile cohort, atypical social ratings were associated with more solitary play.
Social tension, anxiety, or avoidance	These can accompany autism, although they are distinct from low social motivation.	Tension and avoidance were assessed in macaque scales. The separate juvenile study linked atypical social ratings with anxiety-related behavior.
Differences in outward affiliative or comforting behavior	How concern, affection, or comfort is expressed can differ; expression alone does not establish the strength of the underlying feeling.	Comforting and related affiliative behaviors were included in the rating research.
Repetitive or stereotyped movements	Repeated movements and self-directed actions are part of the autism-related behavioral spectrum.	A repetitive-behavior item was retained in the validated macaque scale. Examples included rocking/swaying, spinning, digit sucking, and self-directed actions. Repetitive behavior contributed to a separate factor that did not predict observed social engagement.
Restricted or unusually narrow interests	Intense, focused interests are characteristic of many autistic people.	Restricted-interest content appeared in the original macaque questionnaire, but the relevant item was not retained in the revised adult scale. The defined low-social males have not been shown to possess autism-like circumscribed interests.
Insistence on sameness or strong routines	Some autistic people show marked preferences for predictability and familiar sequences.	The reviewed low-social studies did not establish a corresponding group difference in routines or resistance to change.

Table 2. Broader developmental, biological, familial, and ecological comparisons

Broader feature	Findings or relevant pattern in humans	Findings in the macaque research
Naturally occurring variation within a social species	Autism-associated characteristics arise within human populations, with related traits also occurring outside diagnosis.	Low-social animals were identified through observation of existing variation in social groups, rather than created through lesions, engineered mutations, or imposed isolation.
Continuous rather than strictly categorical variation	Autism-related genetic influences extend into population variation in social behavior and adaptive functioning.	Social behavior and macaque social-responsiveness scores vary continuously. “Low-social” usually identifies a selected portion of that distribution.
Measurement using related behavioral constructs	The human Social Responsiveness Scale measures autism-related social characteristics.	Researchers adapted it for macaques and validated the revised instrument against observed behavior in a program involving 349 animals.
Early developmental emergence	Autism-related differences emerge during development rather than being solely adult social preferences.	Face-recognition and social-gaze differences were detected in assessments at approximately three to four months, before later low-social classification.
Persistence over time	Autism is a neurodevelopmental condition with enduring characteristics, although functioning changes with age and experience.	Initial low-social classification predicted classification two years later.
Familial and inherited contributions	Both common inherited variation and other genetic influences contribute to autism liability.	In a study of 407 males, sons sharing fathers resembled one another in social functioning more strongly than sons sharing mothers.
Paternal age in humans versus paternal inheritance in monkeys	Large population studies associate older paternal age with increased probability of an autism diagnosis in offspring.	The monkey study found a paternal-family inheritance pattern: greater resemblance among paternal half-siblings. It did not demonstrate an effect of the father’s age.
Sex-dependent expression	Autism diagnoses are more frequent in males, and presentation can differ by sex.	The male pattern did not reproduce straightforwardly in 88 females: dominance rank was especially informative, and vasopressin did not predict female social functioning.
Lower cerebrospinal-fluid vasopressin	Studies found lower CSF vasopressin in sampled autistic children, with further human work supporting its relevance to social difficulties.	Lower CSF vasopressin distinguished low-social from high-social males and replicated across monkey cohorts.
Biology tracks the degree of social difficulty	Lower CSF vasopressin has been associated with greater autism-related social difficulties.	In a broader male sample, lower CSF vasopressin predicted higher macaque social-responsiveness scores and remained relatively stable within individuals.
Social responses can be changed through the same neurochemical system	A 30-child randomized pilot trial reported improved social outcomes after intranasal vasopressin.	Vasopressin improved face recognition and affiliative responding in a small low-social male experiment without increasing measured aggression.
Social motivation and social competence are separable	Wanting interaction, understanding it, and successfully participating in it are distinguishable dimensions.	The rating research separated social motivation from other social difficulties; earlier work distinguished putatively introverted from putatively lonely low-sociable males.
Reduced initiation need not mean complete exclusion	Social difficulties do not logically imply that all relationships or responses to others are absent.	Low-social males received comparable prosocial behavior during the longitudinal observations despite initiating less themselves. Separate medical records nevertheless showed more traumatic injuries.
Observed performance can depend on social conditions	Autistic interview performance improved when questions were made more explicit and structured.	In a different rhesus population, subordinate monkeys expressed previously acquired food-related knowledge when dominant animals were absent.
Possible ecological significance of reduced social dependence	The solitary forager hypothesis proposes that some autistic profiles could function competently under particular learning and subsistence conditions.	The low-social phenotype provides measurable variation with which to test independent learning, activity allocation, resource acquisition, and reproductive outcomes. Those ecological advantages have not yet been demonstrated in these males.

2. Defining the Low-Social Phenotype

2.1. Naturally occurring variation within a social population

The principal California National Primate Research Center studies concern animals born and reared in large outdoor social groups. Their low sociality was identified through observation rather than experimentally induced by isolation or genetic manipulation. These are provisioned research populations, however, so their social behavior develops within an environment that does not require independent food procurement for survival. 

Low-Social Rhesus Macaques and Autism Evolutionary Mismatch: Evidence for Subsistence-Reinforcement Coupling, Ecological Competence, and Context-Dependent Fitness

The primary observational distinction is straightforward. Investigators record whether an animal is apart from others, in proximity, in nonaggressive bodily contact, grooming, or playing. “Nonsocial” generally denotes being beyond arm’s reach of another animal and not engaged in social play. Individuals are then positioned along a distribution of observed sociality. This measure identifies how much social contact occurs; by itself, it does not reveal whether an animal away from others is exploring, feeding, resting, or manipulating objects. 

Low-Social Rhesus Macaques and Autism Evolutionary Mismatch: Evidence for Subsistence-Reinforcement Coupling, Ecological Competence, and Context-Dependent Fitness

Quantitative ratings complement these observations. Feczko and colleagues (2016) adapted the Social Responsiveness Scale for rhesus macaques, obtaining ratings for 105 animals. Scores showed a unimodal, positively skewed distribution, demonstrating measurable variation rather than two naturally separated classes. Talbot and colleagues (2020) subsequently evaluated and refined the instrument in 349 monkeys, producing a 17-item macaque Social Responsiveness Scale-Revised, or mSRS-R. Scores predicted observed social behavior and distinguished selected low- and high-social animals.

The relevant population pattern is consequently a continuum. Labels such as “low-social” identify a region of that distribution, often using study-specific thresholds. They should not be interpreted as a fixed percentage of wild macaques constituting a discrete biological morph.

2.2. Low interaction has more than one psychological interpretation

Capitanio and colleagues (2014) distinguished two behavioral patterns among low-sociable males. Some made tentative approaches but engaged in relatively little sustained interaction, suggesting a discrepancy between social interest and social attainment. Others interacted infrequently across potential partners and showed fewer indications of seeking additional contact. The authors interpreted these patterns as putatively lonely and putatively introverted, respectively. The interpretations were examined using experimental social probes, although subjective experience cannot be measured in monkeys as it can through human self-report.

This distinction is central to the evolutionary question. Low observed interaction can reflect low demand for affiliation, difficulty securing desired relationships, fear of particular partners, or some combination. Those possibilities imply different developmental processes and different ecological consequences. An account of low social dependence must distinguish them rather than treating every solitary episode as the same phenotype.

3. Behavioral Correspondences With Autism

The most informative parallels involve specific actions and task performance. The human comparison concerns dimensions found in some autistic people, not characteristics shared by every individual on the spectrum.

The initiation, reciprocity, and developmental findings come from distinct observational and experimental procedures. They should not be treated as eight independent discoveries in entirely separate populations, but together they characterize a specific social-processing profile.

3.1. Initiation is distinguishable from reception

Talbot and colleagues (2022) screened 95 males, comparing 20 low-social with 21 high-social animals. The low-social group initiated less prosocial behavior but did not receive significantly less prosocial behavior or differ in observed threat exchanges. A subset of 22 animals was reassessed two years later, demonstrating persistence of individual differences. The result identifies spontaneous social initiation as a meaningful component of the phenotype, without equating low initiation with complete exclusion from the group.

This finding allows a more precise human comparison. Social participation depends both on an individual’s actions and on the opportunities supplied by others. Reduced initiation need not imply absent capacity to respond, absent attachment, or universal rejection. These are separate outcomes requiring separate measurements.

3.2. Early differences concern the interpretation of social information

Sclafani and colleagues (2016) examined infant assessments from 50 males later classified as low- or high-social at ages one to four. Future low-social animals did not show the face-novelty preference demonstrated by future high-social animals. They also averted their gaze less often during aggressive displays. Initial looking during face familiarization did not explain the recognition difference.

The gaze result is particularly instructive. It does not support a uniform cross-species claim of increased eye avoidance. Instead, it implicates the adjustment of attention to the meaning of a social display. The comparable functional question in autism is how faces and gaze guide recognition, expectation, and response.

3.3. Social and repetitive-behavior dimensions are partly separable

The mSRS-R factor analysis in 233 males identified dimensions labeled Poor Social Motivation, Poor Social Attractiveness, and Inappropriate Behavior. Repetitive, disruptive, and contextually odd behavior contributed to the last factor, which did not predict the measured frequencies of social engagement. Thus, repetitive behavior was assessed, but it was not demonstrated to be an obligatory accompaniment of low social motivation.

The distinction prevents questionnaire contents from being mistaken for a demonstrated syndrome. Restricted interests, technical systemizing, and superior resource acquisition have not been established in the defined low-social male cohorts. The directly supported comparison centers on affiliation, social recognition, and reciprocal signaling.

Human findings also support examining cognitive components separately. Minio-Paluello and colleagues (2020) identified substantial variation in face-identity recognition within autism, including a subgroup with marked difficulties that were not reducible to general intellectual ability. Face recognition is therefore a useful comparative target because its variation can be measured independently of broad diagnostic or intelligence categories.

4. Vasopressin and the Biological Specificity of the Comparison

4.1. From observed sociality to a reproducible neurochemical association

Parker and colleagues (2018) identified lower cerebrospinal-fluid arginine vasopressin in low-social male rhesus macaques, replicated the result in another monkey cohort, and demonstrated within-individual stability in an additional cohort. The study then reported lower concentrations in a small sample of autistic boys compared with male medical controls. This established a direct cross-primate correspondence involving the same biological measure.

Oztan and colleagues (2021) extended the analysis to quantitative variation across 76 males. Forty-three underwent repeated sampling across approximately ten months, and cerebrospinal-fluid vasopressin was available for 75 animals in the larger analysis. Lower concentrations predicted greater mSRS-R social difficulties. The association therefore extended beyond a contrast between selected behavioral extremes.

This is significant for a dimensional account of autism. A biological measure can relate to differences distributed throughout a population, rather than identifying only the presence or absence of a clinical category. It also makes the comparison more specific than a shared description such as “withdrawn” or “introverted.”

4.2. Experimental modulation separates tested social and nonsocial functions

Talbot and colleagues (2024) administered nebulized vasopressin or placebo to eight low-social males in a within-subject design. Vasopressin improved face-recognition performance and appropriate affiliative responding to social displays, while already-intact object recognition remained unchanged. It did not increase measured aggression. A separate four-monkey component examined pharmacokinetics. The behavioral experiment therefore provides evidence of selective, reversible modulation rather than a general enhancement of every tested cognitive function.

The broader implication is that reduced spontaneous social functioning can involve the operating state of a modulatory system. A social-performance difference need not represent an inability of the organism to learn or remember all classes of information. This is a mechanistic basis for investigating uneven cognitive profiles, not a demonstration of globally intact or superior nonsocial cognition.

4.3. Human evidence continues the translational connection

Oztan and colleagues (2026) examined neuropathological specimens and associated data from 18 individuals. They reproduced a relationship between cerebrospinal-fluid vasopressin and autism-related social difficulties across a methodologically different sample. In concurrently collected postmortem specimens, cerebrospinal-fluid concentrations predicted hypothalamic vasopressin gene expression, whereas blood concentrations did not. This strengthens the interpretation of cerebrospinal-fluid vasopressin as a brain-relevant measure rather than an interchangeable peripheral hormone assay.

The existing studies do not yet identify one regional receptor map or complete circuit configuration shared by the two populations. Their contribution is a more focused biological bridge: naturally varying monkey social behavior and human autism-related social difficulties are associated with the same neurochemical system, and manipulating that system changes selected monkey social responses.

5. Familial Structure, Sex, and Developmental Context

Garner and colleagues (2023) analyzed social functioning in 407 males using behavioral observations and mSRS-R ratings. Sons sharing a father showed substantially stronger resemblance than sons sharing a mother. The pattern appeared with both measures and persisted across alternative analyses. Parental behavior was not itself measured, so the finding concerns pedigree-based resemblance rather than direct observation of low-social fathers producing low-social sons. It motivates molecular investigation of inherited and parent-of-origin contributions without establishing a particular mechanism.

Familial structure is relevant because the evolutionary question concerns transmissible variation, not simply transient differences in circumstance. At the same time, inherited liability can affect sensitivity to experience rather than specify behavior rigidly. The next task is to identify which developmental processes mediate the family resemblance and whether they are shared with corresponding human traits.

Sex provides another organizing variable. Oztan and colleagues (2024) assessed 88 females, with neuropeptide measurements in a subset of 16. The male relationships did not reproduce straightforwardly: mSRS-R scores were not significantly related to the same social-behavior measures, vasopressin did not predict social functioning, and dominance rank was especially informative. The authors interpreted the result partly through female matrilineal social organization.

This is evidence that the meaning of a social score depends on the life-history setting in which behavior occurs. It is not evidence that macaques reproduce the human sex ratio in autism; the original research program deliberately emphasized males. The appropriate comparison concerns sex-dependent pathways from biology and social circumstances to observed functioning.

Several publications also draw on overlapping cohorts. They collectively provide deep characterization of an interrelated research population rather than a series of wholly independent population replications. That structure is valuable for connecting behavior to biology, while making independent-site replication an important next step. 

Low-Social Rhesus Macaques and Autism Evolutionary Mismatch: Evidence for Subsistence-Reinforcement Coupling, Ecological Competence, and Context-Dependent Fitness

6. Implications for Autism-Associated Human Variation

6.1. Conserved social mechanisms can vary within a species

The rhesus findings support a comparative prediction made in the original solitary-mammal framework: autism-relevant social processes can be studied through naturally occurring variation in another primate. The combined behavioral and vasopressin evidence makes it reasonable to investigate some human differences as variation in the regulation of conserved social systems, rather than attributing them exclusively to language, schooling, or uniquely human conventions.

This inference concerns underlying systems, not an unchanged diagnosis inherited from a common ancestor. Recognition of partners, initiation of contact, and interpretation of social signals are component functions. They can be combined differently across species and individuals, and their importance can change with ecological circumstances.

A useful distinction follows. An individual can differ in the motivation to seek interaction, the ability to interpret social information, and the consequences of interaction for that individual. These dimensions can influence one another without being identical. Comparative research becomes more informative when it measures each directly.

6.2. Continuous liability does not imply a single cognitive axis

Robinson and colleagues (2016), using resources totaling more than 38,000 individuals, found genetic links between autism risk and variation in social behavior and adaptive functioning in the general population. This supports studying relevant dimensions beyond diagnostic boundaries. The macaque work supplies a parallel research structure: continuous behavioral variation related to measurable biological differences.

Continuity should not be confused with the proposition that autism is simply extreme introversion. The macaque findings already extend beyond interaction frequency to recognition and reciprocal signaling. Human autism additionally includes combinations of developmental, sensory, communicative, and repetitive-behavior characteristics that the monkey model does not reproduce as a complete entity.

Warrier and colleagues’ study of systemizing illustrates the importance of those combinations. In 51,564 participants, systemizing was heritable and genetically correlated with autism. Systemizing polygenic scores predicted restricted and repetitive behavior but not social difficulties in autistic participants. The findings support partly distinguishable dimensions rather than a necessary one-for-one tradeoff between social and nonsocial capacities.

The evolutionary hypothesis can accommodate this architecture. Selection could act on particular traits or combinations, while diagnosis reflects the interaction of several developmental influences. A relatively low need for affiliation would not automatically produce technical expertise; sustained interests would not necessarily entail poor social recognition.

6.3. Social salience could influence the development of expertise

Reser’s comparative article proposed that motivational systems influence which representations gain access to sustained attention and working memory. Applied here, the hypothesis is that differences in the salience of social information alter the experiences that individuals repeatedly process and learn from. A person who spends less time monitoring rapid interpersonal exchanges could, under appropriate conditions, devote more sustained attention to objects, procedures, or environmental regularities. 

Solitary_Mammals_Provide_an_Animal_Model.pdf

This is a developmental allocation hypothesis, not a claim that social disengagement necessarily saves brain energy or creates superior ability. The predicted outcome depends on what replaces the social activity. Time spent independently could involve productive exploration, repetitive practice, rest, anxiety, or little engagement of any kind. Those alternatives need distinct measures.

The macaque phenotype makes this hypothesis experimentally accessible. Investigators can ask whether differences in spontaneous affiliation predict what animals inspect, remember, manipulate, or learn when given comparable opportunities. Preserved recognition of objects establishes one starting point; it does not settle the larger question of expertise.

7. The Evolutionary Significance of Low Sociality

7.1. Social variation should be evaluated through ecological consequences

The solitary forager hypothesis predicts that the value of social investment changes with the returns to independent activity and interpersonal coordination. In a setting where access to resources depends heavily on coalition support or tolerance, reduced affiliation could be costly. Where resources can be obtained independently and interaction carries substantial competition or opportunity costs, the balance could differ.

This formulation does not require all low-social individuals to perform better than high-social individuals. It predicts a social phenotype by environment interaction: differences in success should depend on resource distribution, social density, familiarity, and the activities required. A universal disadvantage would support a different explanation from a context-dependent reversal.

Several evolutionary processes could maintain the relevant variation. Different environments might favor different social investments. A strategy’s value might depend on its frequency, such that a minority of less-initiating individuals benefits from relationships maintained largely by others. Alternatively, a broad range of social dispositions might yield similar lifetime outcomes. These are alternative models to distinguish with ecological and reproductive measurements.

7.2. Injury identifies a cost, not a complete fitness account

Myers and colleagues (2021) examined medical records from 152 males, including 73 low-social and 79 high-social animals. Low-social monkeys experienced more traumatic injury events, and greater nonsocial behavior and mSRS-R difficulties predicted higher injury rates. This prevents equal threat frequencies during sampled observations from being interpreted as equal accumulated social risk. The study does not establish the mechanism of every injury.

An ecological interpretation should ask how recognition, conflict avoidance, partner support, and opportunities to withdraw contribute to that difference. The finding could reflect a cost of reduced social engagement in the colony, a cost of particular social-cognitive difficulties, or both. It does not show that the entire phenotype is globally dysfunctional, nor does it demonstrate an offsetting benefit.

7.3. The survival value of sociality can change with the environment

Testard and colleagues (2024) analyzed ten years of data from free-ranging, provisioned rhesus macaques on Cayo Santiago before and after Hurricane Maria. Deforestation increased heat exposure and reduced access to shade. Following the disturbance, monkeys showed increased tolerance and reduced aggression, facilitating shared use of shaded locations. Social tolerance predicted survival after the hurricane but not before it.

This is direct evidence that the fitness consequences of a social characteristic can change when an ecological resource changes. In this case, the advantage shifted toward greater tolerance. The relevance to the solitary forager hypothesis is the demonstrated environmental contingency, not evidence that the hurricane favored low sociality.

The Cayo measures are also not interchangeable with CNPRC low-social classification. Social-network position, proximity tolerance, affiliative motivation, and social recognition overlap imperfectly. Their relationships should be established before using one as a substitute for another.

7.4. Reproductive viability and reproductive advantage are separate questions

Brent and colleagues (2014) studied 108 free-ranging female rhesus macaques and identified repeatable, heritable personality components, including one labeled Loner. None significantly predicted infant survival or interbirth interval. The study therefore did not identify a straightforward reproductive disadvantage associated with that less-social dimension, but it did not establish exact fitness equivalence or an advantage under a particular environment.

For the defined low-social male phenotype, the research synthesized here does not connect classification or mSRS-R scores to lifetime paternity. That missing measurement is decisive because maintaining some relationships, surviving, reproducing adequately, and outperforming another phenotype are different outcomes. 

Low-Social Rhesus Macaques and Autism Evolutionary Mismatch: Evidence for Subsistence-Reinforcement Coupling, Ecological Competence, and Context-Dependent Fitness

The evolutionary question is consequently specific: which combinations of motivation, social cognition, practical competence, and environmental conditions predict successful reproduction? The phenotype’s recurrence and familial structure justify this investigation, while the answer must come from its consequences.

8. Ecological Competence and Evolutionary Mismatch

A central implication of the solitary forager hypothesis is that practical competence should be measured independently of conventional social performance. The macaque work permits that distinction to be examined within one species, including conditions under which social circumstances affect the expression of learning.

Drea and Wallen (1999) provide an informative adjacent experiment. Rhesus monkeys learned color associations that predicted access to food. Subordinate animals performed poorly when tested alongside dominant families but immediately expressed the learned associations when tested separately. Their prior learning was therefore underestimated by performance in the mixed social setting. These were low-status rather than phenotypically low-social animals, but the experiment demonstrates that social arrangements can conceal food-related knowledge.

The proposed human implication is a distinction between possessing a useful skill and displaying it under a socially demanding assessment. It is an inference from the experimental principle, not evidence that the low-social monkeys are already demonstrably skilled independent foragers.

The relevant learning hypothesis is subsistence-reinforcement coupling. Hunger or another ordinary biological need could make a productive action worth observing because its outcome is valuable. An individual may attend to another’s food-producing behavior for instrumental reasons even when interaction itself supplies little reward. Practice can then be reinforced by the material consequence rather than depending primarily on praise or affiliation. This was a substantive component of the original solitary forager argument. 

Conceptualizing_the_Autism_Spectrum_in_T.pdf

The existing low-social cohorts have not been tested in a design comparing direct food-producing actions with otherwise equivalent socially mediated reward contingencies. Nor do their standard social ethograms establish that reduced affiliation is accompanied by more resource search, object exploration, or persistent practice. These are the specific extensions required to connect the present phenotype to ecological competence. 

Low-Social Rhesus Macaques and Autism Evolutionary Mismatch: Evidence for Subsistence-Reinforcement Coupling, Ecological Competence, and Context-Dependent Fitness

The mismatch hypothesis would be supported if the same individuals performed differently when unnecessary social interpretation, competition, or indirect instruction was removed. An ecological advantage would require more: improved acquisition or use of relevant skills, reduced costs, or better fitness outcomes under specified conditions.

9. A Focused Research Program

9.1. Measure what replaces social engagement

The first requirement is a complete activity budget. Social observation should be combined with measures of feeding, resource search, exploration, manipulation, locomotion, resting, vigilance, and repetitive behavior. This would determine whether low affiliation predicts a productive alternative allocation, a different preferred environment, or lower engagement more generally.

Partner identity and familiarity should be included. An animal that interacts infrequently but maintains reliable relationships may differ substantially from one that has few successful interactions despite repeated attempts. Reproductive partners, familiar companions, and unfamiliar competitors should not be collapsed into one category of “social stimulus.”

9.2. Separate learning, motivation, and social performance

Low- and high-social animals could learn matched food-puzzle tasks under three arrangements: an action directly opens access to food; the same action activates an automatic dispenser; or the same action is followed by food delivered by a familiar experimenter. Reward amount, delay, visibility, prior experience, and motor demands should be matched as closely as possible.

The automatic-dispenser condition is important. It helps distinguish a benefit of a transparent physical contingency from a cost specifically associated with human mediation. Tests should measure acquisition, retention, transfer, persistence after errors, attention to the apparatus and demonstrator, and willingness to engage. Ordinary feeding and voluntary participation are sufficient; withholding necessary nutrition is not part of the hypothesis.

Sociality should be analyzed continuously, alongside rank, anxiety-related behavior, age, and relevant biological measures. The informative result would be an interaction in which particular profiles respond differently to task organization, not merely evidence that every monkey prefers the easier condition.

9.3. Connect behavior to survival and reproduction

Behavioral characterization should be linked to complete reproductive histories, including access to potential partners, mating, paternity, offspring survival, and age at reproduction. Environmental comparisons should examine density, resource distribution, social stability, and opportunities for spatial separation.

A useful test would compare the relative success of individuals under changes in those conditions. If low social initiation is costly in one setting but neutral or advantageous in another, the result would support context-dependent selection. If particular social-recognition difficulties predict costs across settings while low interaction frequency does not, those dimensions should receive different evolutionary interpretations.

9.4. Match the human and macaque constructs

Human studies should distinguish desired contact, spontaneous initiation, response to invitations, face recognition, interpretation of social signals, and nonsocial task performance. Self-report can establish experiences that cannot be inferred directly in animals, including whether limited interaction is preferred or distressing.

Cross-species comparisons should then examine matched functions rather than total questionnaire scores alone. A relation between the same biological measure and the same component function would be more informative than a broad resemblance between diagnostic labels. This approach also accommodates autistic individuals who desire substantial connection but experience recognition, communication, sensory, or access barriers.

10. Conclusion

Naturally occurring low sociality in rhesus macaques provides substantive comparative evidence relevant to the original solitary-mammal framework. The findings connect observed differences in affiliation with early social-information processing, enduring individual variation, familial structure, and a neurochemical system also implicated in human autism-related social difficulties. Experimental results further show that selected social functions can change while a tested nonsocial function remains preserved.

The central implication is that some autism-associated characteristics can be investigated as dimensions of primate social organization rather than only as deviations from an assumed uniform social phenotype. Motivation, recognition, reciprocity, and participation need not vary together, and their consequences depend on the circumstances in which they operate.

The solitary forager hypothesis proposes that particular combinations of these traits could support viable or advantageous forms of reduced social dependence. The macaque research now makes that proposition more tractable. The next empirical step is to connect the established social phenotype to independent learning, resource acquisition, and reproductive outcomes. That connection would determine when reduced social investment constitutes an effective ecological strategy, when it imposes costs, and how both possibilities inform human variation.

References

Brent, L. J. N., Semple, S., MacLarnon, A., Ruiz-Lambides, A., Gonzalez-Martinez, J., & Platt, M. L. (2014). Personality traits in rhesus macaques (Macaca mulatta) are heritable but do not predict reproductive output. International Journal of Primatology, 35, 188–209. DOI: 10.1007/s10764-013-9724-6.

Capitanio, J. P., Hawkley, L. C., Cole, S. W., & Cacioppo, J. T. (2014). A behavioral taxonomy of loneliness in humans and rhesus monkeys (Macaca mulatta). PLOS ONE, 9(10), e110307. DOI: 10.1371/journal.pone.0110307.

Drea, C. M., & Wallen, K. (1999). Low-status monkeys “play dumb” when learning in mixed social groups. Proceedings of the National Academy of Sciences, 96(22), 12965–12969. DOI: 10.1073/pnas.96.22.12965.

Feczko, E. J., Bliss-Moreau, E., Walum, H., Pruett, J. R., Jr., & Parr, L. A. (2016). The Macaque Social Responsiveness Scale (mSRS): A rapid screening tool for assessing variability in the social responsiveness of rhesus monkeys (Macaca mulatta). PLOS ONE, 11(1), e0145956. DOI: 10.1371/journal.pone.0145956.

Garner, J. P., et al. (2023). Rhesus macaque social functioning is paternally, but not maternally, inherited by sons: Potential implications for autism. Molecular Autism, 14, 25. DOI: 10.1186/s13229-023-00556-3.

Minio-Paluello, I., Porciello, G., Pascual-Leone, A., & Baron-Cohen, S. (2020). Face individual identity recognition: A potential endophenotype in autism. Molecular Autism, 11, 81. DOI: 10.1186/s13229-020-00371-0.

Myers, A. K., et al. (2021). Assessment of medical morbidities in a rhesus monkey model of naturally occurring low sociality. Autism Research, 14(7), 1332–1346. DOI: 10.1002/aur.2512.

Oztan, O., et al. (2021). Autism-associated biomarkers: Test–retest reliability and relationship to quantitative social trait variation in rhesus monkeys. Molecular Autism, 12, 50. DOI: 10.1186/s13229-021-00442-w.

Oztan, O., et al. (2024). Naturally occurring low sociality in female rhesus monkeys: A tractable model for autism or not? Molecular Autism, 15, 8. DOI: 10.1186/s13229-024-00588-3.

Oztan, O., Zhu, C., Nguyen, D. K. K., West, R. B., Garner, J. P., & Parker, K. J. (2026). Cerebrospinal fluid vasopressin concentration is a biomarker of autistic social impairment and hypothalamic vasopressin gene expression in humans. Autism Research, 19(3), e70181. DOI: 10.1002/aur.70181.

Parker, K. J., et al. (2018). Arginine vasopressin in cerebrospinal fluid is a marker of sociality in nonhuman primates. Science Translational Medicine, 10(439), eaam9100. DOI: 10.1126/scitranslmed.aam9100.

Reser, J. E. (2011). Conceptualizing the autism spectrum in terms of natural selection and behavioral ecology: The solitary forager hypothesis. Evolutionary Psychology, 9(2), 207–238. DOI: 10.1177/147470491100900209.

Reser, J. E. (2014). Solitary mammals provide an animal model for autism spectrum disorders. Journal of Comparative Psychology, 128(1), 99–113. First published online November 4, 2013. DOI: 10.1037/a0034519.

Robinson, E. B., et al. (2016). Genetic risk for autism spectrum disorders and neuropsychiatric variation in the general population. Nature Genetics, 48, 552–555. DOI: 10.1038/ng.3529.

Sclafani, V., et al. (2016). Early predictors of impaired social functioning in male rhesus macaques (Macaca mulatta). PLOS ONE, 11(10), e0165401. DOI: 10.1371/journal.pone.0165401.

Talbot, C. F., et al. (2020). A psychometrically robust screening tool to rapidly identify socially impaired monkeys in the general population. Autism Research, 13(9), 1465–1475. DOI: 10.1002/aur.2335.

Talbot, C. F., et al. (2021). The factor structure of the macaque social responsiveness scale-revised predicts social behavior and personality dimensions. American Journal of Primatology, 83(5), e23234. DOI: 10.1002/ajp.23234.

Talbot, C. F., Madrid, J. E., Del Rosso, L. A., Capitanio, J. P., Garner, J. P., & Parker, K. J. (2022). Rhesus monkey sociality is stable across time and linked to variation in the initiation but not receipt of prosocial behavior. American Journal of Primatology, 84(12), e23442. DOI: 10.1002/ajp.23442.

Talbot, C. F., et al. (2024). Nebulized vasopressin penetrates CSF and improves social cognition without inducing aggression in a rhesus monkey model of autism. Proceedings of the National Academy of Sciences, 121(49), e2418635121. DOI: 10.1073/pnas.2418635121.

Testard, C., et al. (2024). Ecological disturbance alters the adaptive benefits of social ties. Science, 384(6702), 1330–1335. DOI: 10.1126/science.adk0606.

Warrier, V., et al. (2019). Social and non-social autism symptoms and trait domains are genetically dissociable. Communications Biology, 2, 328. DOI: 10.1038/s42003-019-0558-4.