Visual art inspired by climate change—An analysis of audience reactions to 37 artworks presented during 21st UN climate summit in Paris

first indication of how features of environmental art affect the audience. More positive emo- tions were found if the artwork included something personal, was not too big, was presented with an explanation of how to "read" it, but did not include destruction, especially of known landmarks. Negative emotions, on the other hand, are likely triggered by polarizing and monu- mental artworks. Whereas some of these findings might have been caused by the specific selec- tion of artworks, they nevertheless point at some interesting relations between artwork features and how the artworks are perceived, which artists should be aware of. Limitations of the study Despite its uniqueness, the presented study has some important weaknesses that need to be acknowledged. First, the study’s correlational design does not allow drawing conclusions on causes and effects among the studied variables. In connection with its strongly exploratory nature on the artwork level, it has generated several hypotheses for future work, but it is not able to conclusively decide that a certain aspect is the cause of another effect. More controlled studies need to replicate and confirm the presented mechanisms and effects. Furthermore, the measurement quality of the variable "identification with the artist" is sub-optimal, which pre- vented its use in this study. Second, the literature on art perception is not definite on the order of emotions and cogni- tions–Locher [46] and Capo et al. [60] place, for example, the emotional reaction first, as we

did it in our model, while Leder [16] suggests an iterative process, in which emotions inform all the other steps in art perception in a feedback loop. To check for iterative feedback of emo- tions, we would have needed to follow our participants longitudinally through the experience probing them repeatedly about their emotional status and cognitive reflections. This was impossible to implement in the circumstances under which this data was collected. Further research should take into consideration whether emotions might inform environmental art perception through all steps. Third, many of the studied artworks were presented in designated art spaces such as galler- ies or open-air areas in connection to museums. This also skewed the sample toward people who describe themselves as art interested. The level of art expertise was found by Leder et al. [23] to influence the level of emotional reaction to "negative" art; even if we could not find an impact of art expertise on the emotional reactions and reflections to the artwork in our study, there is still the possibility that participants with a higher level of art expertise perceived the artworks differently than novices. Fourth, because of the purpose, venue, and timing of the exhibits, it is likely that the audi- ence was more concerned about climate change than the average Parisian resident or visitor, which may have produced different results from what would have been found in a more repre- sentative population sample. Because the number of audience members studied per artwork

was not large (around 25 per artwork), more detailed analyses were unfortunately not possible that might differentiate between artwork effects on a dedicated art audience and an audience consisting of people during their everyday lives. The willingness to reflect and think might be smaller for a general population sample. Perhaps the emotional triggers would be lost, or per- haps they would be stronger because the experience would be unexpected. Finally, even if the focus of our study was not to explain support for climate change policies in detail and we, therefore, did not include a large set of potential predictors of support, it can be argued that policy support is usually strongly impacted by variables such as political prefer- ences and worldviews, which we thus were not able to control for. If these variables were related to how people react to artworks (which they well might be), we might falsely assume that the variables in our model explain the differences in policy support, whereas the PLOS ONE Visual art inspired by climate change — An analysis of audience reactions PLOS ONE | https://doi.org/10.1371/journal.pone.0247331 February 19, 2021 13 / 18