Ellis, Rod: Understanding Second Language Acquisition. Oxford: Oxford University Press, 1985. (327 pages) ISBN 0 19 437081 x (out of print).
Understanding Second Language Acquisition by Rod Ellis provides an overview of what was known about second language acquisition (SLA) by (around) 1984. The author points out that the discipline was still in its infancy, so no single SLA theory could be presented as received opinion. The book was written with two types of readers in mind: students and teachers. The book assumes a basic knowledge of linguistics (for example, linguistic terms such as idiolect and minimal pairs are not explained in the glossary) but is otherwise very accessible.
Ellis explains in the introduction why it is important for language teachers to become familiar with SLA theory:
All teachers have a theory of language learning. That is, they act in accordance with a set of principles about the way language learners behave. This theory, however, may not be explicit. In many cases the teacher's views about language learning will be covert and will only be implicit in what he does. (…)
Greater consciousness of the complex process of language learning will not guarantee more effective teaching—arguably our state of knowledge is insufficient to warrant firm pedagogical applications—but it will stimulate critical thought, challenge old principles, and maybe suggest a few new ones. A conscious understanding of SLA is a basis for modifying and improving teaching.
Since the book is more than thirty years old and obviously outdated, I will not discuss the whole book but instead list the questions and areas where a lot of uncertainty existed in the early 1980s and where more research was needed. This list of open research questions can then be compared with more recent accounts of SLA to see what progress has been made since the early 1980s.
- There had been little empirical research on
whether and under what circumstances learners transfer the realizations of a given function in their L1 to their use of the L2(page 39).
- There had been few studies on free variation in interlanguage that tried
to look at form-function relationships, that is to investigate which forms are used to express which meanings.(Page 85)
- The effect of individual learner differences on the route of SLA had not yet been seriously investigated (page 99).
- There was still disagreement (or
a lack of agreement) on whether children are better language learners than adults (page 104-105).
- Age did not appear to alter the route of acquisition, but studies had equated accuracy order with acquisition order (page 105).
In general the evidence linking cerebral dominance and age differences in learners is not clear.(Page 108.)
- Empirical research had not yet addressed the hypothesis that field dependence may be most facilitative in naturalistic SLA, while field independence may lead to greater success in classroom SLA. (Page 114-445.)
- Research results on the effect of cognitive style on proficiency had not been conclusive; cognitive style did not appear to be an important factor in SLA. (Page 115.)
There has been no research into the effect of cognitive style on route of acquisition.(Page 116.)
- There was no general agreement on the definitions of “motivation” or “attitude” (page 117).
- While it was clear that motivation is an important factor in SLA, it was not clear exactly how it affected learning. In other words, there is a correlation, but the direction was not clear. (Page 119.)
- Research results (except Rossier's PhD dissertation, 1976) gave only partial support to the hypothesis that extroverted learners learn a foreign language more rapidly. (Page 120.)
- Little was known about
the relationship between motherese and the route of development(i.e. route of acquisition; page 131).
- SLA research had not progressed as far as L1 research with regards to the effects of input and interaction (page 132).
- The evidence on the correlation between quantity of input and rate of learning was mixed; more research was needed (page 160).
The detailed workings of intake analysis […] have not yet been describedand SLA researchers had limited understanding of them (page 173).
- Some empirical research on communication strategies in SLA had led to results that were suggestive rather than definitive. Researchers increasingly turned to introspective methods. (Pages 183, 188.)
- The Critical Period Hypothesis appeared untenable (page 201).
- It was not clear how research could solve the problem of discovering
which cognitive processes need to be re-established in a second language and which transferred(page 202). (See V. Cook, 1985.)
- There had been
relatively few studies of how linguistic universals affect interlanguage(page 203). See E. Kellerman, 1984, for a review.
- By the mid 1980s,
the evidence in support of the Universal Hypothesis [was considered] inconclusive(page 2013). This was due to both theoretical and methodological problems.
- There had been relatively little research into the effect of classroom instruction (as opposed to naturalistic SLA or mixed SLA) on the route of SLA. As a consequence, conclusions could only be tentative.
- There was insufficient evidence to make a choice between the non-interface position (Krashen), the interface position and the variability position (page 241).
- It was not very clear what implications for language teaching could be drawn from SLA research (page 242).
- SLA research was still needed into the discourse features that facilitate the rate of development (page 260).
- It was not clear to what extent some left-brain functions could be
localised, i.e. creative language use (including syntax
and semantic processing) and
the motor operations involved in speaking and writing. See the detailed reviewed in Psycholinguistics: A Second Language Perspective by Evelyn Marcussen Hatch (Rowley, Mass: Newbury House, 1983).
In general […] claims about localized functions need to be treated circumspectly.(Page 272-273.)
- There were contradictions between classical studies on the neurolinguistic correlates of specific language functions. (Page 274.) See Genesee, 1982.
- Until the early 1980s, SLA research focused mostly on grammar, and neglected vocabulary and pragmatic development. Exceptions: Meara, 1980; Sociolinguistics ans Second Language Acquisition (Wolfson & Judd, 1983). (Page 288.)
The following points highlight issues related to SLA research methods:
[D]oubts [had] been expressed about the validity of examining human behaviour in experimental conditions (e.g. Cicourel et al. 1974), on the grounds that experimental tasks are not likely to elicit natural or normal behaviour.(Page 87)
- Some researchers pointed out a danger in SLA research, namely that certain specific linguistic phenomena may be an artefact of the elicitation instrument, for example, that the “natural” order found through morpheme studies may be an artifact of an elicitation instrument such as the Bilingual Syntax Measure. (Page 90) (See especially K. Porter: “A cross-sectional study of morpheme acquisition in first language users”, Language Learning, 27 (1977): 47-62.)
- There was no clearly defined effect of personality on SLA. This was partly due to problems of defining and measuring various personality traits. (Page 121-122.)
- Regarding research on the “good language learner”:
a common refrain in the research literature that the tests chosen to measure a particular concept may not have been valid(page 123) and
a more qualitative approach based on research and introspection may first be necessary in order to identify the relevant hypotheses(page 123).
- Many studies of foreigner talk had failed to provide baseline data, i.e. data on the speech between native speakers who perform the same tasks as the language learners (page 133). See also M. Long: “Input, Interaction and Second Language Acquisition” (1981).
- P. Lightbown (1983) pointed out that the statistical methods used to support the correlation between frequency of exposure and acquisition can be misleading (page 156).
- In studies into the effect of formal instruction on SLA
difficult to separate out the effects of instruction and exposure(page 225-226).
Krashen does not refer to any studies which have directly compared methods based on formal grammar teaching of one kind or another and methods based on providing opportunities for authentic communication.(Page 233)
A recurring issue in SLA research is the validity of different data types (see, for instance, Naiman 1974, Adams 1978, Burt and Dulay 1980 and Wode et al. 1979).(Page 289.)
Most studies cited by Rod Ellis are about the learning or acquisition of English as a second language or as a foreign language. The studies that focus on the acquisition of other languages are all about languages from the same language family, i.e. the Indo-European languages, typically Germanic or Romance languages. As a consequence, the acquisition of many language features is neglected in SLA research, which is a serious limitation in the scope of this type of research. Examples of neglected features include tone in tonal languages (e.g. Standard Chinese), vowel harmony in languages such as Finnish and Turkish, case systems that are more elaborate than in German (e.g. in Lithuanian, Finnish or Hungarian), or different writing systems (Chinese, Arabic, Korean Hangul, Cyrillic, etc.). In addition, the studies cited by Rod Ellis have very little to say about the inference of the first language when that language does not belong to the Indo-European family. Perhaps this type of research was still rare in the early 1980s. SLA research has gained importance outside the Western world, so a more recent overview of SLA would be able to include findings on a wider range of languages and language features.
Other aspects of foreign language learning that were not discussed in the book are third language acquisition, potential negative effects of multilingualism on previously acquired languages (including L1), the re-acquisition of (partially) forgotten languages, and autonomous language learning in general. Some of these gaps can be explained by the fact that early SLA research focused on language learning in classrooms.
Rod Ellis has written several other books on second language acquisition, including Second Language Acquisition (Oxford University Press, 1997, 160 pages), The Study of Second Language Acquisition (Oxford University Press, second edition, 2008, 1,176 pages) and the second edition of Understanding Second Language Acquisition (Oxford University Press, 2015, 376 pages; the book has a companion website).