Melissa A. Schoenlein¹ (), Melina O. Mueller², Karen B. Schloss³; ¹High Point University, ²Vanderbilt University, ³University of Wisconsin-Madison

Previous research found that people form color-concept associations for novel concrete concepts from direct exposure to color-concept co-occurrences (first-order relations) (Schoenlein & Schloss, 2022). However, people also have associations for abstract concepts lacking directly observable colors. How are color-concept associations formed for such concepts? We addressed this question by investigating whether participants could form color-concept associations from indirect co-occurrences (higher-order relations). In Experiment 1, participants were exposed to concept-concept co-occurrences: Nerky/Lomby tails belonging to Filk/Slub alien species, respectively. Then, participants were exposed to color-concept co-occurrences via images of warm-colored Filks/cool-colored Slubs. Last, participants rated how much they associated Nerky/Lomby tails with warm/cool colors (second-order relations). Participants associated warm colors more with Nerky tails and cool colors with Lomby tails (p<.001), despite them never directly co-occurring. In Experiment 2 we evaluated if concept relation type influenced strength of second-order associations. For one group, Lomby/Nerky represented tails again (has-a-property relation); for another, Lomby/Nerky represented cities near the aliens (located-near relation). Both groups formed warm-Nerky/cool-Lomby associations (ps<.001), with no significant differences depending on relation type (p=.123). In Experiment 3 we tested if participants could learn second- and third-order relations using a similar task structure, while adding an additional concept-concept pairing: emotions-movie genres, movie genres-alien species, and alien species-colors. Participants formed associations between abstract concepts and colors through second-order relations (movie genres–colors), but evidence did not support forming associations through third-order relations (emotions–colors). It is possible these results demonstrate limitations on forming higher-order color-concept associations. Alternatively, additional exposure may be needed for third-order associations to emerge. Together, these findings demonstrate that people use indirect co-occurrences to form color-concept associations, even from concept-concept relations varying in relatedness. This work furthers our understanding of how people form color-associations for concepts without directly observable colors, permitting future investigations into association formation for more abstract concepts.

Acknowledgements: This work was supported in part by the National Science Foundation [BCS-1945303].