Learning Japanese as a Second Language: A Processability Perspect ...

suggestions for deciding which L2 aspects to focus on. However, CAH also had limitations and sometimes made inappropriate predictions (e.g., Duskava, 1969). It also occasionally underpredicted the transfer: Not all areas of similarity between L1 and L2 led to immediate positive transfer (Hyltenstam, 1977). Research carried out by various linguists additionally found that many errors in learning were not related to the L1–L2 contrast (e.g., Lococo, 1975).

2.1.2. Error Analysis and the Nativist View
of Language Acquisition

Challenges to CAH led to consideration of the possibility that L2 learners follow a universal path in acquiring a target language. This is similar to children’s L1 learning, which follows a universal path, as shown in Brown’s (1973) study. Chomsky’s view of L1 learning (1959, 1965) was very influential at that time for explaining the universal path. Rejecting the idea posited by the behaviourists, Chomsky claimed that L1 learning is a product of rule formation but not of habit formation. This is evident from the fact that the child learning L1 is able to both comprehend and produce utterances he/she has never heard before. This clearly indicates that language learning is not merely an imitation of input, as behaviourism had assumed. From this fact, Chomsky claimed that the ability to acquire language is innate to humans, and that children possess the ability to discover language rules from input gained from the linguistic environment. Chomsky assumed that children are equipped with what he called a language acquisition device (LAD): After being exposed to the input, children start to use language automatically through the discovery and expansion of the language rules. However, when children have not fully mastered the rules or limitation of the rules, they display ‘developmental errors’ (Brown, 1973). One example of this is the use of English past tense morpheme -ed even for irregular verbs such as goand eat. This kind of error is due to overgeneralisation of the English grammatical rule of encoding PAST.

Influenced by Chomsky’s assumptions about L1 acquisition, some SLA researchers also claimed that L2 learning, like L1 learning, is not the result