Speech segmentation and structural generalisation can occur simultaneously.
Generalisations are powerful and apply to novel material not previously experienced.
Statistical computations can explain performance for both tasks.
Algebraic computations distinct from statistics aren’t needed to describe performance.