The history of linguistics is filled with discoveries of generalizations, and their subsequent rejection when exceptions to linguistic rules are uncovered. Using the learnability perspective, this interdisciplinary monograph makes a convincing, cognitively grounded case for exceptions, arguing that productive linguistic rules do not need to be exceptionless to be acquired and subsequently exist in a speaker's grammar. The study is based on the acquisition of the Icelandic case-marking system, which for half a century has frequently been the exception to the rule in formal case theory. Building upon experimental data from 148 children aged 2–13, as well as incremental child language corpus analyses, Nowenstein addresses four distinct, but related topics in language acquisition research: (morpho)syntactic bootstrapping, rule formation and productivity, variation and change in acquisition and the role of memory in language learning. The book uncovers important insights into the many unanswered questions within the field of language acquisition.