Regularization is a phenomenon in language acquisition and language development whereby irregular forms in morphology and syntax are often replaced by a regular one. For example geese are often regularized as Gooses.
Learners exposed to inconsistent input appear to change the language as they learn it, making it more regular. Children will almost always regularize inconsistent forms. Adult learners, however, will only regularize the most complex inconsistencies (Kam et al., 2009) An example would be a study done by Singleton and Newport in 1994, the study was based on a child named Simon, who was profoundly deaf and received no other ASL input as he attended a school where no ASL was signed in addition he has no contact with other children who knew ASL. So his only form of ASL contact was his parents who have both learned ASL in their mid to late teens. This made his scenario very rare and unique because while Simon is a native signer of ASL, his parents are non native signers. However even so, his parents provided him with his only linguistic model of ASL. Simon’s and his parents’ production of a particular class of morphemes in ASL (morphemes of movement) was studied. An example of morphemes of movement can be seen in the figure below.
The parents were found to use these morphemes inconsistently, producing them around 70% of the time in obligatory contexts. In contrast, Simon used these forms very consistently, around 90% of the time in obligatory contexts. This level of consistency was equivalent to that seen in the productions of control children who had been exposed to native ASL. So it was hypothesized that Simon reorganized the inputs so that his own linguistic system was rule-governed. We will refer to this process as ‘regularization’. A form of refularization would be overregularization.
One reason adults and children differ in their language acquisition abilities is that they also differ in other cognitive capacities: for instance, the relatively poor memory and/or processing abilities of children may make them more likely to over-regularize inconsistent input.