This section deals with the methodology of error analyses, particularly as it is applied to the analysis of the adult learner’s syntax in a second language. Corder presents a model based on a distinction between an idiosyncratic dialect (the learner’s personal, unstable, developing grammar) and a social dialect (the target language which is the dialect of a social group). He illustrates the possible realtioships between the learner’s idiosyncratic dialect and the target social dialect, and compares these with a number of other X dialect/social dialect relationships, such as that evidence in a piece of poetry, in the speech of a young child, or in the speech of an aphasic. Corder attributes greater importance to the relationship mother tongue/learner’s dialect than others might; it does however acknowledge that there are other variables involved.

Richards, while acknowledging the influence of the mother tongue on the learner’s language, documents a number of other common features of the learner’s dialect, often ignored in discussions of learner’s errors. These are referred to as intralingual and developmental errors and reflect the general characteristics of rule learning, such as overgeneralization, incomplete application of target language rules, failure to learn conditions under which rules apply, and the development of false concepts. Some of these are seen to be reinforced by common teaching procedures. Jain sees the learner’s language as manifesting a general learning strategy to simplify the syntax of the language he is learning. He suggests that the motivation to add new rules to one’s idiosyncratic dialect may decline, once a degree of proficiency has been achieved for the language to function adequately as an operational tool, abd illustrates the concept of overgeneralization as a learning strategy. Jain is concerned with reconstructing the learner’s interlanguage, when hypothesis testing has largely stopped.