Randi Starrfelt¹ (), Anders Petersen¹, Peter Laursen², Erling Nørkær¹; ¹University of Copenhagen, Denmark, ²Bispebjerg Hospital, Copenhagen, Denmark

Visual processing of faces and words rely on a number of shared low-level processes, but the degree to which higher order visual and memory processes are shared or independent between these stimulus categories is debated. There are suggestions that the two interact and even compete during development, so that learning to read impacts the (cerebral) organization of the face recognition system. Whether there is a systematic relationship between face and word processing on the behavioral level in literate adults, and if so, in what direction, remains unclear. Here, we take advantage of unique data from a representative sample of Danish adults aged 41-71 (n=694) tested with the Cognitive Function Scanner test battery. Data on immediate and delayed (1h) face and word recognition, as well as a test of general visual ability were collected as part of a larger study of cognitive abilities. In pre-registered analyses, we test the relationship between face and word processing and its direction, and whether this potential relationship may be explained by general visual ability. In addition, we explore whether the relationship between word and face processing changes with recognition delay. The relationship between d’ for words and faces was significant at both immediate and delayed recognition, in a positive direction. Adding scores from the general visual ability test to the regression did not abolish this relationship. Performance for both faces and words decrease with recognition delay, but more so for words than faces. The results suggest that word and face processing do rely on some shared higher order processes, and contradict the hypothesis of a negative relationship (competition) between performance with faces and words. Interestingly, the positive relationship between faces and words does not seem to be explained by a common underlying visual factor