Interlanguage is the speech or other output of second language (L2) learners or the internal mental system that underlies this output. The term was introduced by Larry Selinker (Reference Selinker1972) in an article that was a founding document in the field of SLA. In this book, 10 scholars trace our progress in the study of interlanguage from 1972 up to the present time.
In the Introduction, Han and Tarone note that Selinker’s article did not contain a theory of interlanguage but rather presented a series of foundational questions and possible answers. These themes included what kind of data is best, the influence of the first language, fossilization, overgeneralization, and how best to explain and understand descriptive findings. Only three of these themes can be discussed here. The first is the question of what kind of data should be used to study interlanguage. Selinker (Reference Selinker1972) was clear on the matter, stating that the only relevant data were contained in “‘attempted meaningful’ performance in a second language” (p. 210). Of course, a great many interlanguage studies have used other kinds of data, including grammaticality judgments and nonsense words. In Chapter 7, Gass and Polio observe that the use of these and other data types continues, but they note that Selinker’s article problematized grammaticality judgments and nonsense words so that researchers are now more likely to use these two types only along with additional data sources.
A second question Selinker raised was the matter of fossilization, which Han defines in Chapter 3 as “features [of interlanguage] that are impermeable to … instruction or exposure” (p. 52). Han makes two particularly important points: (1) Although Selinker conceived of fossilization as a general phenomenon affecting ultimate overall attainment in SLA, we now look at fossilization in terms of specific interlanguage systems, such as the article system; and (2) an understanding of fossilization will require an account of how interlanguage is mentally represented and processed. Han suggests that first language transfer is one source of fossilization and sketches a processing account based on Levelt’s (Reference Levelt1989) well-known production model for native speaker speech, where messages are initially framed within a mental module called the Conceptualizer. Han claims that for L2 speakers “the operations within the Conceptualizer are carried out and framed in the L1 [first language], resulting in an L1-packaged message. This message will then be encoded using L2 grammatical and phonological means…hence L1 thinking for L2 speaking” (p. 65).
The over-arching question that Selinker raised was whether interlanguage is systematic, a revolutionary question at the time because many scholars considered learners’ speech to be a hodge-podge of transferred forms and errors. All of the authors review studies showing that system can, indeed, be found in interlanguage. But, as Selinker insisted, we must go beyond description to explanation, which can only be couched in terms of some theory. A continuing tension in the book is whether this theory should be universal grammar (UG), as represented in the chapters by Montrul and VanPatten, or the more recent emergentist approaches, in which system is not motivated by innate knowledge but rather is constructed from input by general cognitive learning mechanisms, as in dynamic systems theory, discussed by Larsen-Freeman in Chapter 9. In Chapter 8, Ortega recounts how several of these rival approaches have described and explained the generally agreed upon facts of negative acquisition, including UG, Labovian variation theory, and several compatible emergentist approaches, including dynamic systems theory and construction grammar, which is associated with the school of cognitive linguistics. In regard to variation theory, it seems to me that it, too, is compatible with the emergentist accounts because it relates interlanguage regularities to variable input data from the speech community, not to innate structures. Indeed, Ortega says that variation in interlanguage (such as the fact that Schumann’s subject Alberto used both no and don’t for a long time to negate propositions in a systematic but nonnative way) is a “Trojan horse,” meaning that it must be accounted for but is presently ignored by some theories, including UG and Pienemann’s (2011) processing account.
Interlanguage: Forty Years Later is a valuable and important book for SLA scholars and graduate students because it puts us in touch with our history and because it provides state-of-the-art accounts of the most basic issues within our discipline, which were first raised by Selinker in the beginning.
