Facial beauty and moral beauty stand as two fundamental pillars of social aesthetics, each playing a distinct yet interconnected role in human social interaction. Facial beauty, rooted in visual sensory and symbolic elements, is often shaped by physical features such as averageness and symmetry, while moral beauty, an abstract and physically independent construct, reflects humanity, virtue, and adherence to social norms. These two forms of beauty interact dynamically: the "Beauty-is-Good" stereotype highlights how facial attractiveness influences moral judgments, and the "Good-is-Beauty" stereotype demonstrates that moral goodness can alter perceptions of facial appeal. Despite extensive research on the neural mechanisms of individual aesthetic modalities, the neural underpinnings of their integration—how the brain processes and combines facial and moral beauty simultaneously—have long remained an unresolved question in neuroaesthetics.

To address this gap, the present study employed functional magnetic resonance imaging (fMRI) to investigate the neural correlates of integrated aesthetic evaluation. The research design aimed to identify three core neural networks: one for processing beauty, one for ugliness, and another for resolving aesthetic conflict. Building on prior findings, the medial orbitofrontal cortex (mOFC) was hypothesized to be central to beauty processing, given its established role in both facial attractiveness and moral goodness evaluations. The insular cortex, linked to negative stimulus processing, was expected to be involved in ugliness perception. Additionally, the medial prefrontal cortex (mPFC), a region sensitive to cognitive and emotional conflict, was predicted to mediate aesthetic conflict when facial and moral beauty information were incongruent.

Twenty-four right-handed female undergraduate students were initially recruited, with three excluded due to technical issues, resulting in a final sample of 21 participants (mean age = 21.62 years, SD = 1.63). Female participants were selected to avoid gender-related confounds and because they demonstrate more consistent aesthetic ratings of heterosexual faces. Stimuli included facial portraits (400 male Asian photographs) and moral descriptions (400 short sentences), pretested and categorized into beautiful, neutral, and ugly tiers via a 7-point scale by 68 pilot participants. These stimuli were randomly paired to form five experimental conditions: Facial Beauty-Moral Beauty (FBMB), Facial Ugliness-Moral Ugliness (FUMU), Facial Ugliness-Moral Beauty (FUMB), Facial Beauty-Moral Ugliness (FBMU), and a Neutral control condition (Facial Neutral-Moral Neutral).

During fMRI scanning, participants completed three runs of 80 trials each, evaluating the general aesthetics of paired stimuli on a 4-point scale within 2000 ms. The scanning protocol used a T2-weighted gradient echo sequence, with preprocessing (realignment, normalization, smoothing) and statistical analyses conducted via SPM8. Post-scan ratings confirmed the validity of stimulus selection: facial and moral stimuli showed significant differences across beautiful, neutral, and ugly categories (all ps < 0.001), with beautiful stimuli receiving the highest ratings and ugly stimuli the lowest.

Behavioral results revealed robust effects of stimulus congruence on aesthetic evaluation. FBMB received the highest aesthetic ratings (3.40 ± 0.30), while FUMU received the lowest (1.35 ± 0.24). Intermediate ratings were observed for FUMB, FBMU, and Neutral conditions, with FBMU rated significantly less attractive than FUMB and Neutral (ps < 0.001). Response times were shortest for congruent conditions (FBMB: 522 ± 129 ms; FUMU: 508 ± 116 ms) and significantly longer for conflicting (FUMB: 581 ± 122 ms; FBMU: 561 ± 120 ms) and Neutral (572 ± 137 ms) conditions, indicating that congruent aesthetic information facilitates faster processing.

Neurologically, the integration of facial and moral beauty recruited a common network involving the mOFC and middle occipital gyrus (MOG). The mOFC, a key reward-related region, showed graded activity across conditions—highest in FBMB and lowest in FUMU—reflecting its role as an integrated center for social beauty evaluation. The MOG, a core visual processing area, exhibited selective activation in the FBMB condition, suggesting enhanced visual processing for congruently beautiful stimuli. In contrast, no shared neural regions were identified for facial and moral ugliness alone. Instead, the bilateral insular cortex and supplementary motor area (SMA) were specifically activated in the FUMU condition (congruent ugliness), with insular activity suppressed in conflicting conditions, indicating sensitivity to pure ugliness.

Aesthetic conflict (FUMB + FBMU vs. FBMB + FUMU) uniquely activated the mPFC, bilateral inferior parietal lobe (IPL), right inferior temporal gyrus (ITG), and left cerebellum. Critically, only the mPFC showed a significant positive correlation between activation and conflict intensity (r = 0.448, p = 0.042), confirming its role as the primary mediator of emotional and cognitive conflict in integrated aesthetics. This aligns with prior research highlighting the mPFC’s involvement in resolving competing information.

The study’s findings advance our understanding of social aesthetics by demonstrating that integrated beauty evaluation relies on specialized neural networks. The mOFC serves as a hub for integrating positive aesthetic information, the insula for congruent negative aesthetics, and the mPFC for conflict resolution. Notably, the dissociation between beauty and ugliness networks challenges the view of aesthetics as a single continuum, instead supporting two independent systems. These results also emphasize that integrated aesthetics are more cognitively complex than single-modality evaluations, reflecting the multifaceted nature of social beauty perception.

Limitations of the study include the use of explicit evaluation tasks (which may introduce attentional biases) and the exclusive focus on female participants and male stimuli, precluding analysis of gender effects. Future research could adopt implicit tasks and diverse samples to further validate these findings. Nevertheless, this study provides novel neural evidence for the integration of facial and moral beauty, offering a foundation for understanding how humans navigate complex social aesthetic information in daily interactions.