Abstract
<p>Four experiments investigated whether previously observed differences in the perceived blurriness of words versus nonwords reflect early perceptual processes or post-perceptual cognitive factors such as memory. Building on Lupyan’s report that familiar words literally appear less blurry than scrambled letter strings objectively equally blurred, we designed a forced-choice paradigm featuring an immediate “perceptual” block, where both stimuli remained visible during the judgment, and a “memory” block, where one stimulus disappeared before participants responded. Across Experiments 1–3 using words as targets, nonwords were consistently more likely than words to be judged as blurrier than the target, with this effect generally stronger in the memory condition. Experiment 3 eliminated a same word confound present in Experiments 1 and 2, suggesting that the effect was not solely driven by stimulus repetition. However, Experiment 4, in which the target was always a nonword, failed to produce a statistically reliable lexicality effect in either block. Combined analyses across Experiments 3 and 4 indicated that the clarity advantage for word over nonword samples was stronger when words were targets than when they were samples which could be due to an upward vertical attention bias for meaningful stimuli. Together, these findings suggest that familiarity-related sharpening effects are highly sensitive to contextual and task-related factors, emphasizing the need for caution when interpreting such effects as evidence of direct top-down influences on low-level visual perception. All four experiments also demonstrated a sharpening effect of memory for both words and nonwords, consistent with previous research.</p>