Abstract
AbstractGrammars often employ the tool of repetition (reduplication) in word formation. But repetition cannot be allowed to occur accidentally. Cross‐linguistically, when word formation producesunintendedrepetition, construction‐specific strategies to avoid this repetition develop. These strategies include allomorphy, haplology and sheer avoidance of the offending construction. This entry concentrates on several aspects of repetition avoidance (also known as ‘the repeated morph constraint’) that have not been central to previous treatments. First, it reveals how repetition avoidance is relevant to the root–affix asymmetry, showcasing phenomena from Modern Hebrew that draw a parallel between templates and affixes. Second, repetition avoidance is shown to motivate the recruitment of ‘interfixes’ – featurally empty or not – that intervene between the repeated entities. Third, the phenomenon of repetition avoidance is argued to be active both in the phonology and prior to it, which makes the case for pre‐phonological optimization. Finally, a listener‐oriented model is put forth that describes the possible scenarios for the emergence of repetition avoidance.