Seeing beyond the image: Contextualising autism in art to shape aesthetic experience

Abstract

We explored whether providing information that artistic photography depicts individuals on the autism spectrum and their special interests influences viewers’ preferences. Our findings demonstrated a positive impact of providing such information on participants’ ratings of aesthetic emotions and judgments. The present study suggests that artistic activities showing autistic individuals can serve as positive self-advocacy tools when framed by contextual information.

There is a growing number of art exhibitions showcasing works by artists on the autism spectrum (AS) ¹ . One reason for this trend could be the increasing number of self-diagnosed individuals and the improved availability of diagnostic tools at an early age. Nevertheless, there remains an ongoing discussion about whether information regarding an artist's autism diagnosis should be made salient to the visitors of such exhibitions. The argument in favour of providing this information is that it could enhance the presence of artists on the AS in public discourse and promote inclusivity. Conversely, the counterargument is that highlighting information about the AS may distract viewers from engaging with the subject of the artwork, shifting their focus towards the artist's diagnosis rather than the artistic subject or merit.

Existing research on artwork preferences demonstrates that contextual information about an artist's gender or nationality significantly influences the perception and evaluation of artworks. Artworks labelled as created by male artists are more likely to be recognised as famous and valued as more expensive compared to those attributed to female artists (Hoffmann & Coate, 2022). Additionally, American participants tend to show a stronger preference for artworks labelled as created by American artists over those by Italians (Mastandrea et al., 2021).

However, no research to date has examined how contextualising image content as depicting individuals on the AS may influence preference judgements. One suggestion can be drawn from Trawiński et al. (2024), who explored viewers’ responses to paintings featuring Black and White sitters in a gallery setting, while measuring their attitudes towards and experience with both communities. The study revealed that contextualising image content through viewers’ social experiences with the Black community positively influenced aesthetic appeal, particularly the experience of pleasure, when the depicted individuals were Black. Further evidence comes from Szubielska et al. (2020), who found that providing information about an artist's visual impairments enhanced aesthetic appreciation and emotional responses, especially for artworks whose content is difficult to interpret. Together, these findings suggest that informational framing, whether about the artist or the subject, may shape the viewer's aesthetic experience.

In the present preliminary online study, we aimed to extend previous findings by investigating how manipulating contextual information about image content may influence preference judgements. To do so, we selected 12 artistic photographs by Paweł Bajew, depicting individuals on the AS (both males and females) alongside objects indicating their special interests (see Figure 1(A)). These artworks were originally created for a calendar produced by a foundation supporting children and adults on the AS. The primary purpose of the calendar was to serve as a form of self-advocacy for the individuals portrayed, who are beneficiaries of the foundation.

Figure 1.

(A) Example of an Image with Contextual Information (English Translation: “Adam has autism. He is passionate about literature”). (B) Self-Assessment Manikin scales used in this study. Parallel coordinate plots illustrating the results for aesthetic judgement (C) and aesthetic emotions (D) as a function of context.

Participants viewed the photographs one at a time, at their own pace, in a fixed order. Their task was to rate the emotional states evoked by each photograph in terms of emotional valence (negative vs. positive), arousal (low vs. high), dominance (low vs. high), origin (heart/automatic vs. mind/reflective), and significance (low vs. high) using the 9-point Self-Assessment Manikin scales (Szubielska & Imbir, 2021; see Figure 1(B)). Following this, participants made aesthetic judgements regarding their understanding and liking of each image using a 9-point Likert-type scale (1 = low, 9 = high; Szubielska & Imbir, 2021).

One hundred undergraduate students (56 females; M_age = 22.10, SD = 1.58, range = 20–28 years) with no formal art training participated in this study. Half of the participants were assigned to the experimental condition, where they viewed the photographs accompanied by contextual information about the portrayed individuals, including their name, AS, and a brief description of their special interests. The other half was allocated to the control group, to which no contextual information was provided.

To examine the effect of contextual information on participants’ aesthetic emotions and aesthetic judgements, two separate multivariate analyses of variance (MANOVAs) were conducted. The first MANOVA assessed the effect of context (context vs. no-context) on aesthetic emotions, including the variables: emotional valence, arousal, dominance, significance, and origin/reflectiveness. The second MANOVA examined the effect of context on aesthetic judgments, specifically liking and understanding ratings.

The analyses revealed a statistically significant effect of context on both the aesthetic judgement variables, Λ = .52, F(2, 97) = 45.29, p < .001, η_p² = .48 and the aesthetic emotions, Λ = .51, F(5, 94) = 17.93, p < .001, η_p² = .49 (see Figure 1(C) and (D)). For aesthetic judgements, participants rated the images with contextual information significantly higher in terms of liking and understanding (all: F's > 49.30, p's < .001, η_p²'s > .335). Similarly, for aesthetic emotions, participants with contextual information reported significantly more positive emotional valence, higher arousal, greater dominance, a more reflective rather than automatic response, and greater personal significance (all: F's > 28.26, p's < .001, η_p²'s > .224).

Three possible explanations may account for the positive effect of contextual information on aesthetic emotions and judgements. First, the social desirability account suggests that making information about the AS salient may prompt individuals to report a positive response more readily, driven by a desire to conform to perceived social expectations (see Trawiński et al., 2024, for a similar argument about race), particularly given the demographic background of our participants. Second, the framing account explains that providing contextual information about the AS alters aesthetic experiences and judgements by influencing how an identical stimulus is perceived in a different context (Szubielska et al., 2020). Third, the “double empathy problem,” the mutual misunderstanding between autistic and non-autistic individuals (Milton, 2012), may explain neurotypical participants’ attitudes towards neurodivergent people. In our study, those in the contextual condition may have better understood the depicted behaviours, such as reserved facial expressions or lack of smiling, leading to greater appreciation of the photographs (Crompton et al., 2025).

In terms of potential limitations, we acknowledge that a fully balanced study paradigm would require a control condition in which participants receive information unrelated to autism. This may limit the interpretability of our findings, as one cannot exclude the possibility that the mere presence of any contextual information could enhance meaning-making and lead to higher aesthetic evaluations. Moreover, while our study demonstrates a positive effect of contextual information on aesthetic emotions and judgements, it does not clarify how individuals arrive at these decisions or whether individual differences in dispositional factors may modulate this response. Finally, it suggests that using self-reported measures may be suboptimal for studying the issues of perception of neurodiversity, and there is a need to use less socially biased methods. In the follow-up studies, we aim to address these issues.