A review of language regression in autism spectrum disorder and the role of language theories: Towards explanation

Background & aims The purpose of this study is to summarize characteristics of language regression—a phenomenon most commonly associated with autism spectrum disorder, outline language theories of early word learning, and use them to propose theoretical bases to language regression. Method Using a systematic review of the current literature in language regression and a purposive sampling of language theories, hypotheses for the cause of language regression are discussed. Results The following hypotheses were developed based on the reviewed theories of word learning: (a) Initial first words were not true words, but instead were protowords; (b) fast mapped forms never progressed to truly learned words; (c) initial first words were echoed productions of heard words; (d) the described regression is actually a maintenance of previous developmental levels with increasing chronological age. Conclusions It is possible that individuals with autism spectrum disorder are not “losing” words, but instead are not progressing in their ability to learn and use new words. Early word learning in typically developing children is seemingly similar to early development in individuals with autism spectrum disorder; however, the ability to expand from early “intermediate words” to expansive vocabularies, word combinations, and morphological variations is impeded. Clinical implications for evaluation of word “loss” and treatments are discussed.