Statistical learning and reading: Visual regularities support affix detection in developing readers

The present study investigated the role of visual statistical learning in how developing readers learn to identify morphemes within words. A total of 121 children (55 girls, aged 6-11 years, M = 8.82, SD = 1.30) were recruited in Trieste, Italy, between January and June 2019. They were familiarized with pseudo-letter strings containing affix-like chunks, which could be identified only by their statistical properties. After passively observing the stimuli, children were more likely to attribute previously unseen strings to the familiarization lexicon if they contained a chunk, regardless of its position within the string. Results indicate that children can acquire morpho-orthographic knowledge through visual regularities from printed input. This ability was not modulated by age nor reading fluency, suggesting an early-maturing learning mechanism. The findings emphasize the importance of incorporating this fundamental, language-agnostic mechanism into morphology and reading acquisition theories.