A matter of timing: when does learning to read start to impact on nonlinguistic object recognition?

Mirror invariance (i.e., processing mirror images like b and d as equivalent percepts) is an original property of the visual system entrenched by evolution. This property collides with learning a script with mirrored symbols like the Latin alphabet which requires mirror discrimination. Therefore, examining the impact of literacy on mirror-image processing is one of the most interesting ways to investigate the consequences of literacy on the evolutionary older cognitive system of visual object recognition. In the present study, we investigated when, during reading development, mirror discrimination becomes automatic in object recognition and whether the impact of literacy extends to other orientation-contrasts, considering explicit vs. implicit, automatic processing of orientation. Geometric shapes were presented to preliterate preschoolers and first-grade beginning readers on two same-different matching tasks differing by criterion: orientation-based vs. shape-based (orientation-independent) tasks. On orientation-based judgments, first-graders presented and overall advantage over preschoolers, and preschoolers showed a stronger difficulty with mirror discrimination. Thus, regarding explicit orientation processing, the impact of literacy was stronger for (but not restricted to) mirror images. In contrasts, on shape-based judgments of geometric shapes, the two groups differed only in mirrored trials. Whereas preschoolers were as able to perform shape-based judgments of identical as of mirrored pairs, first-graders exhibited a strong mirror cost: they were slower on mirrored trials, and even slower than preschoolers. This spillover effect of literacy on orientation-invariant object recognition is thus specific to mirror images. It begins to emerge with letter knowledge before literacy instruction and continues to develop along with reading skills