This strand of work uses artificial language learning methods to explore different theories about how we children lean to avoid overgeneralization errors: When children learn their mother tongue, they sometimes make errors such as “he giggled me!”, “she fell me over!” or “carry me teddy”. Learning to avoid these mistakes in English (and many other languages of the world) involves learning that some verbs can appear in a two-person sentence (e.
Languages have multiple ways of saying things. In some cases, choosing between alternative forms (e.g., dinner vs. supper; cool vs. awesome) depends on aspects of our identity: Among adults, males and females often have difference preferences, and so do younger and older speakers, or speakers of different dialects. But how early on do children learn that word choice/pronunciation may depend on such cues?
Experiments with collaborators Elizabeth Wonnacott, Kenny Smith, and Helen Brown address this question by teaching 5-year-olds an artificial language where word choice depends on whether the speaker is male or female.
Spelling is a difficult task, particularly in orthographies where letters and sounds do not have one-to-one correspondence. In English, for example, vowel sounds can be spelled in as many as five different ways. In line with evidence that we do not just memorize difficult spellings, my work has been investigating spellers’ sensitivity to untaught spelling rules and patterns.
-Can young spellers pick up on untaught orthographic conventions such as “gz is an illegal spelling of a frequent word-final sound combination in English; *bagz” from simple text exposure?