Social vs. Non-Social Attention

Social versus non-social attention refers to the relative prioritization the attentional system gives to socially relevant stimuli — faces, eyes, voices, gestures, biological motion, and emotional expressions — compared to non-social environmental features such as objects, patterns, and physical properties. In typical development, the attentional system is powerfully biased toward social information from birth: neonates preferentially orient to face-like patterns, infants track biological motion over mechanical motion, and adults involuntarily orient to faces in complex visual scenes. In autism spectrum disorder, this social attention bias is significantly reduced, producing a distinctive pattern of visual and auditory attention allocation that has profound implications for social-cognitive development.

Social Attention in Typical Development

The typical social attention system develops through a series of milestones that reflect the progressive specialization of neural circuits for processing social information:

• Neonatal face preference — Within hours of birth, human neonates preferentially orient to face-like configurations (two blobs above a single blob, resembling a face) over inverted or scrambled patterns. This innate bias, mediated by subcortical visual pathways (superior colliculus, pulvinar), ensures that the developing visual system receives disproportionate face input, driving the cortical specialization of the fusiform face area and other face-processing regions.
• Eye contact preference — By 2–3 months, infants preferentially fixate on the eye region of faces, especially when eyes are directed toward them (direct gaze). Direct gaze is processed as a social signal indicating communicative intent, triggering enhanced neural processing and autonomic arousal. By 4–5 months, infants follow others' gaze direction, using it as a cue to redirect their own attention.
• Biological motion preference — Infants as young as 2 days old preferentially attend to point-light displays of biological motion (walking figures) over scrambled motion with identical kinematic properties. This preference reflects an evolved sensitivity to animate motion that supports early social attention and the development of action understanding.
• Voice preference — Neonates show preferential neural responses to human speech over non-speech sounds, and to their mother's voice over unfamiliar voices. This auditory social bias complements visual social attention and supports early language acquisition.

Atypical Social Attention in Autism

The social attention system develops atypically in autism, with differences detectable in the first year of life:

• Reduced eye fixation — Eye-tracking studies consistently show that autistic individuals spend less time looking at the eye region of faces during social interactions and when viewing social scenes. This reduction is present in both children and adults and is correlated with the severity of social communication difficulties. Jones and Klin (2013) demonstrated that declining eye fixation between 2 and 6 months of age predicts later ASD diagnosis, suggesting that the reduction in eye fixation is progressive rather than present from birth.
• Altered face scanning — When viewing faces, autistic individuals show atypical scan paths: less time on the eyes, more time on the mouth or lower face regions, and sometimes fixation on non-feature areas (forehead, chin, ears). This altered scanning pattern reduces the extraction of socially critical information conveyed by the eyes — including gaze direction, emotional state, and communicative intent.
• Reduced preference for biological motion — Autistic individuals show reduced preferential attention to biological motion over scrambled motion displays. Toddlers later diagnosed with ASD fail to show the typical preference for upright biological motion, suggesting that the attentional system's innate bias toward animate motion is reduced.
• Reduced social orienting — In naturalistic settings, autistic children less frequently orient to their name being called, to the sound of others' voices, and to social events occurring in their visual periphery. This reduced social orienting limits the frequency of social attention episodes and, consequently, the amount of social information available for learning.
• Object over face preference — When presented with competing social (faces) and non-social (objects) stimuli, autistic individuals more frequently orient first to objects and spend more total time fixating on objects. This preference reversal — objects over faces — is one of the most distinctive features of the autistic attention profile and has been demonstrated using eye tracking from infancy through adulthood.
• Preserved attention to mouths and speech — Some autistic individuals show increased attention to the mouth region of faces, particularly during speech. This may represent a compensatory strategy that supports speech comprehension through visual speech information (lip reading) or may reflect an interest in the mechanical aspects of speech production rather than the social-communicative aspects conveyed by the eyes.

Neural Mechanisms

• Amygdala differences — The amygdala detects the biological significance of stimuli and modulates attention allocation accordingly, prioritizing stimuli that are socially or emotionally salient. Atypical amygdala function in autism — including altered activation to faces and eyes, and reduced modulation of attention by social significance — may underlie the reduced social attention bias. The amygdala theory of autism (Baron-Cohen et al., 2000) proposes that early amygdala dysfunction disrupts the social attention system, with cascading effects on social-cognitive development.
• Superior temporal sulcus (STS) — The STS is specialized for processing biological motion, gaze direction, and other socially relevant dynamic stimuli. Reduced STS activation in response to social stimuli is a consistent finding in autism neuroimaging studies, correlating with behavioral measures of social attention and social cognition.
• Fusiform face area (FFA) — The FFA, specialized for face processing, shows reduced activation during face viewing in many autistic individuals. However, this hypoactivation may be partly a consequence of reduced visual attention to faces rather than a primary face-processing deficit — when autistic individuals are instructed to attend to faces, FFA activation increases, though it may still differ in pattern from neurotypical activation.
• Subcortical social detection pathway — The fast, subcortical pathway (retina → superior colliculus → pulvinar → amygdala) that supports rapid, reflexive orienting to face-like stimuli may function atypically in autism. This subcortical route bypasses cortical analysis and provides a "fast track" for detecting social stimuli, and reduced functioning of this pathway would reduce the automatic, involuntary social attention bias from infancy.
• Reward system differences — Social stimuli are inherently rewarding for typically developing individuals, activating the mesolimbic dopamine system (ventral tegmental area, nucleus accumbens). The social motivation theory of autism proposes that social stimuli are less inherently rewarding for autistic individuals, reducing the attentional system's motivation to prioritize social over non-social information. Neuroimaging studies show reduced ventral striatum activation to social rewards in autism, supporting this account.

Developmental Consequences

• Social learning cascade — Reduced social attention from infancy creates a cumulative deficit in social learning opportunities. Each moment of reduced attention to faces, eyes, and social events represents a missed opportunity to learn about facial expressions, emotional states, communicative intentions, and social contingencies. Over months and years of development, these accumulated differences in social input lead to progressively divergent social-cognitive trajectories.
• Language development — Attention to faces and mouths during speech provides critical multimodal input for language acquisition. Reduced social attention limits exposure to the visual speech information (lip movements, facial expressions) that normally supplements and disambiguates auditory speech, potentially contributing to the language delays common in autism.
• Theory of mind — Understanding others' mental states requires attending to the cues that reveal those states — facial expressions, gaze direction, vocal prosody, and body language. Reduced attention to these social cues limits the data available for building a theory of others' minds, contributing to the theory of mind difficulties observed in autism.
• Emotion recognition — The ability to recognize and interpret emotional expressions develops through extensive exposure to emotional faces and voices. Reduced visual attention to the eye region — where the most diagnostic emotional information is located (particularly for differentiating fear, anger, and sadness) — limits the development of emotion recognition skills.

Measurement and Assessment

• Eye tracking — The primary method for studying social attention, providing precise measurement of where individuals look, for how long, and in what sequence. Metrics include: time to first fixation on faces versus objects, total dwell time on eyes versus mouth versus non-face regions, and fixation frequency during social scenes. Modern eye-tracking studies use naturalistic video stimuli to approximate real-world social attention patterns.
• Preferential looking paradigms — Presenting competing social and non-social stimuli side by side and measuring the proportion of looking time directed to each. This method is used with infants and nonverbal individuals and can detect social attention biases from the first months of life.
• Social orienting tasks — Measuring latency and frequency of orienting to social stimuli (name calling, face appearing in periphery, social sounds) versus non-social stimuli, in both controlled laboratory settings and naturalistic observations.
• Point-light biological motion displays — Assessing preferential attention to biological versus scrambled point-light animations to index the sensitivity of the motion processing system to social-biological stimuli.

Interventions Targeting Social Attention

• Gaze-contingent paradigms — Computer-based interventions that reward looking at faces and eyes by presenting engaging visual or auditory consequences contingent on gaze to social stimuli. These approaches use the principles of operant conditioning to increase the reinforcing value of social attention.
• Social skills programs — Programs like PEERS that explicitly teach social attention skills: where to look during conversations, how to read facial expressions, and how to monitor social cues in group settings.
• Video modeling — Using video demonstrations to show where social attention should be directed during interactions, making the implicit rules of social attention allocation explicit and learnable.
• Environmental structuring — Reducing non-social distractors in learning and social environments to increase the relative salience of social stimuli, supporting the natural development of social attention.
• Naturalistic developmental behavioral interventions — Approaches like Early Start Denver Model (ESDM) and PRT that create highly reinforcing social learning opportunities within naturalistic play, increasing the child's motivation to attend to social partners.