Answers to theoretical questions about the place of input in a formal second language acquisition model are dependent on a distinction between two kinds of learner hypotheses. Type-N hypotheses require "negative evidence" for testing, while Type-P hypotheses are tested on the basis of "positive data" alone. (Author/CB)

Examines error gravity research design and its theoretical assumptions. Results indicate that investigators have only skimmed the surface of the process of error evaluation, which is shaped by extralinguistic factors. The article concludes that researchers should reconceptualize error gravity research and reassess earlier studies to confirm or…