The Impact of Visual Imagery Absence on Fine-Grained Memory for Object Colour and Size: An Aphantasia Study

Visual memory and visual imagery are closely intertwined cognitive functions, yet whether visual memory necessarily deteriorates when visual imagery is impaired remains unresolved. Aphantasia, the inability to voluntarily generate mental images, provides a compelling model to address this issue. However, previous reports of preserved or subtle deficits in aphantasia may have stemmed from tasks with low cognitive demands, coarse measures, and memory content lacking fine-grained demands, thereby limiting sensitivity to genuine impairments. Using high-load paradigms and continuous precision measures with mixture modelling, we tested colour memory, a domain requiring fine-grained representation, to examine whether these conditions would reveal pronounced impairments in aphantasia. Experiment 1 (20 aphantasic, 20 controls) tested colour memory for six sets of 20 objects (15 meaningful, 5 meaningless). Meaningless objects were included to further minimise verbal labelling. All object colours were sampled from the Hue–Saturation–Value (HSV) space, a perceptually uniform colour space. Recall was assessed using a continuous colour wheel, and performance was analysed with a computational mixture model that decomposed response errors into retrieval success (whether a colour was retrieved) and precision (how accurately it was remembered). Results revealed that aphantasic individuals showed markedly larger errors than controls, driven by reduced retrieval success rather than precision. Moreover, error magnitude and retrieval success correlated with several imagery dimensions (e.g., object imagery, spontaneous use of imagery), whereas retrieval precision was not. Experiment 2 (18 aphantasic, 20 controls) further tested whether deficits extend beyond colour using an analogous size-memory task with a slider. No group differences were observed in size error, retrieval success, or precision, and no correlations were observed between memory and several imagery-related aspects. Taken together, this study provides the first evidence that visual imagery is indispensable for retrieval success but not for precision when accessing visual memory traces for object colour, and that this indispensability may not extend to all object features.