![]()
IIIThe teaching and learning of foreign languages
![]()
Kumiko Sakoda
5 Errors and learning strategies by learners of Japanese as a second language
1 Introduction
The purposes of this chapter include: (1) discussing learnersâ errors and (2) claiming that making errors is a learning strategy used by language learners. Analyzing learnersâ errors and correct usages helps us understand how their language systems and their rules operate and how they differ from the ones that teachers teach and from those of native speakers. In the 1960s and 1970s, error analysis, which attempted to classify the errors made by learners, flourished as a result of the criticism of contrastive analysis, which focused on the systematic comparison of a pair of languages for the purpose of identifying their structural differences and similarities. However, error analysis was also problematic for some reasons, and its heyday is long gone. However, it is obvious that no one can acquire a target language without making errors; even good language learners commit errors. Likewise, even the best teachers and the best method (if it exists) cannot go without learners producing errors.
Errors hold significant meaning for researchers and teachers. Error analysis in the 1960s and 1970s was problematic because of its methodological problems and scope limitations. This chapter affirms the correctness of the view that the learnersâ errors are worthy of study in their own right, not just as a degenerate form of the target system, by presenting two empirical acquisition studies of Japanese as a second language (L2). One is a study on acquisition of the demonstratives KO-, SO- and A-, whereas the other concerns the particles NI and DE. Analysis of three years of longitudinal spoken data revealed interlanguage, a dynamic linguistic system developed by an L2 learner who is approaching the target language but has not become fully proficient. Learners are developing their own language systems, including errors, and one of the characteristics is that they make chunks and their chunking reflects their own rules and systematicity, which represent learning strategies.
To begin with, âerrorsâ and âlearning strategiesâ need to be defined. According to Corder (1967), an âerrorâ is a deviation in learner language which results from lack of knowledge of the correct rule (Ellis 2008). Lennonâs (1991) definition holds an error to be a linguistic form or combination of forms that, in the same context and under similar conditions of production, would, in all likelihood, not be produced by the speakerâs native speaker counterparts. James (1998) defines an error as an instance of language that is unintentionally deviant and is not self-corrigible by its author. This chapter tentatively adopts Jamesâ definition. However, to define the term âerrorsâ is quite problematic, which we will discuss later.
The notion of âlearning strategiesâ differs among researchers. Oxford (1990) and OâMalley and Chamot (1990), for example, classified learning strategies into some groups, such as cognitive, meta-cognitive, social, memory-related, compensatory and affective strategies. According to Selinker (1972), fossilization, which refers to the loss of progress in L2 acquisition despite regular exposure to and interaction with the target language, is caused by five central processes, and a learning strategy is one of them. Selinker describes a strategy of L2 learning as a tendency on the part of learners to reduce the target language to a simpler system. Also, Ellis (2008: 970) states, âA learning strategy is a device or procedure used by learners to develop their interlanguage. Learning strategies account for how learners acquire and automatize L2 knowledge.â We adopt Ellisâs definition in this chapter.
The chapter is organized as follows: Section 2 illustrates a historical overview of error analysis both in English as a Second/Foreign Language (ESL/EFL) and Japanese as a Second/Foreign Language (JSL/JFL) studies. Section 3 discusses problems of error analysis by focusing on the definition of error analysis, the limitation of scope, and methodological and theoretical problems. Section 4 introduces two acquisition studies of JSL: a study of Japanese demonstratives KO-, SO- and A- and a study of Japanese particles NI and DE. Analyzing learnersâ errors and correct use investigates learnersâ developing Japanese language systems. Section 5 concludes the chapter and suggests tha errors should not be regarded as signs of inhibition, but simply as an evidence of the learnerâs strategies of learning, followed by a look at future directions in SLA research.
2 Overview of Error Analysis
This section illustrates an overview of error analysis in ESL/EFL and JSL/JFL studies. The section introduces some of the main findings of the studies.
2.1 Error analysis in ESL/EFL research: Historical trends
In the 1950s the audio-lingual approach to foreign language education was prevalent. In the heyday of the audio-lingual approach, it was thought that learning would be more difficult if the differences between the native and target language were great whereas it would be easier if the differences were small. These assumptions were based on the idea of contrastive analysis, which involves analyzing two languages for the purpose of observing where two languages differ and then using those differences as the basis for predicting errors and for developing foreign-language teaching curriculum. In other words, contrastive analysis, which shares its ideas with the behaviorist approach represented by Skinner (1957), is grounded on the assumption that language is a set of habits: (1) the habits established in oneâs first language (L1) might help or facilitate second-language (L2) learning (i.e., positive transfer); but (2) they might interfere with L2 learning (i.e., negative transfer) and produce errors. However, the results from contrastive analysis were not always as had been expected. Instead, shared errors among learners of different first languages were found to be common.
By the early 1960s, Chomskyâs (1957, 1959, 1965) formulation of an innate rulegoverned system evolving toward the full adult grammar was received with enthusiasm by linguists and psychologists, and eventually by SLA researchers and foreignlanguage educators in the early 1970s. In the field of L1 acquisition, to begin with, Brown and Bellugi (1964) stressed the rule-governed nature of language acquisition as opposed to behaviorist theory, which holds that childrenâs speech is not rulegoverned but is shaped by external contingencies and reinforced by caretakersâ approval. While acknowledging the significant influence of environmental or parental interactions (e.g., mothers modify their speech to their children by simplifying, repeating, and paraphrasing), Brown and Bellugi emphasized that the process of language acquisition could not be explained by the behaviorist stimulus-response-reinforcement system alone. Rather, these researchers believed that the childâs ability to engage in inductive processing would aid language acquisition. They paid particular attention to inductive processes characterizing the childâs acquisition of syntactic structures. In analyzing toddlersâ language acquisition, Brown and Bellugi thus concluded that mother-child interaction, which is a cycle of imitations, reductions, and expansions, would help the childâs inductive processing of the latent structure and rules of the target language. In other words, Brown and Bellugi suggest that the childâs innate ability to formulate hypotheses about the rules underlying language might overshadow the importance of environmental factors. In this formulation, therefore, because language acquisition/learning is a developmental process that involves reorganization of knowledge, errors are considered a sign of progress (Davis, Ovando, and Minami 2013; Minami 2002, 2015; Minami and Ovando 2003).16
The impact of the above-described L1 research on L2 research was immense. Researchers turned their attention to errors and collected learnersâ errors (Dulay and Burt 1972). Following Brownâs (1973) L1 acquisition study, which reported remarkable similarity among his three subject children in the order in which fourteen functor/ grammatical morphemes (e.g., present progressive -ing, plural âs) were acquired, Dulay and Burt (1974) conducted a series of L2 studies to examine the question of whether developmental sequences, regardless of the L1 or the L2, could be identified in terms of a common/invariant sequence of acquisition for functor morphemes in English. Replacing contrastive analysis, error analysis thus enjoyed considerable popularity in the late 1960s and 1970s. Consequently, many studies were conducted using error analysis (e.g., Corder 1967, 1971, 1981; Chamot 1978; Richards 1974).
Error analysis emphasized the importance of understandin...
