Hypothesis Testing in Second‐language Acquisition Theory

Learner hypotheses may be classified logically into two kinds depending on the relationship of the hypothesis to the data needed to test it. Type‐N hypotheses require “negative evidence” (of which explicit correction is a special type) for testing. Type‐P hypotheses can be tested on the basis of “positive data” (comprehensible input, for example) alone. The difference between hypotheses is related to the intuitive concepts of “strictness” and “looseness”. Answers to theoretical questions about the place of input in a formal L2 acquisition model as well as to empirical questions about the necessity of correction are dependent on this hypothesis type distinction.