of the University of Minnesota in a number of ways. Two University of Minnesota Graduate School Grants in Aid of Research supported our research, one awarded to Elaine Tarone in 2002 and another in 2003 to Martha Bigelow, who also won a single-semester leave to gather data and immerse herself in the community. In 2004 Elaine Tarone and Kit Hansen were given a College of Liberal Arts Graduate Research Partnership Program grant. All these grants funded Kimberly Johnson, Larry Davis, Mike Hinrichs, and Becky Uran Markman, who helped us with research, transcription, and data analysis. Nora Wildgen White is the artist who drew pictures we used to gather the data (see Figure 3.2). The University of Minnesota has also supported our travel to numerous conferences to present versions of the papers upon which this book is based.
We are deeply grateful to Jenefer Philp, whose original study simultaneously anchored and inspired us. We wish to acknowledge Bonnie Swierzbin and Bob delMas, who were co-authors with us on papers published with the data reported in this book. We benefited greatly from feedback on earlier drafts of segments of this book by Michael Graves, Abukar Ali, Anne Lazaraton, Jim Lantolf, Lourdes Ortega, Merrill Swain, Henry Widdowson, Jill Watson, and George Yule. Mike E. Anderson provided weekly encouragement on getting this project to completion. Finally, we would like to thank our editor, Cristina Whitecross, with whom it has been such a pleasure to work.
Abbreviations
ACT Adaptive Control Theory
EI elicited imitation
ESF European Science Foundation
ESL English as second language
IL interlanguage
L1 first language
L2 second language
LARC Language Acquisition Research Centre
LESLLA Low-Educated Second Language and Literacy Acquisition
NLLSD Native Language Literacy Screening Device
NNS non-native speaker
NS native speaker
SLA second language acquisition
SPEAK Speaking Proficiency English Assessment Kit
TLU target-like use
ZISA Zweitspracherwerb Italienischer und Spanischer Arbeiter
ZPD Zone of Proximal Development
Introduction: Questioning assumptions in SLA research
A prominent goal of research on second-language acquisition (SLA) is to identify universal cognitive processes involved in acquiring ‘second languages’,1 or L2s. Yet there has been very little SLA research to date on the cognitive processes of illiterate or low-literate adult L2 learners.2 Almost all the adult learners studied in SLA research have been literate, in the sense that they have been able to decode printed or embodied text. Typically, they have even been college students, such as undergraduates in foreign-language programs, graduate students in intensive English programs, or international teaching assistants. These L2 learners have all been initiated into the social practice of print literacy, which is an essential skill that affords them access to, and power in, the academic world. It is a skill providing access to many of academia’s other literacies as well, including media and digital literacies.
But can these literate L2 learners be assumed to be representative of all L2 learners? Can we base an SLA theory of universal cognitive processes on data drawn only from literate learners? What about L2 learners who do not participate in the social practice of print literacy at all? These learners are sometimes referred to as ‘preliterate’ (Robson 1982). Such learners clearly exist in large numbers throughout the world, but we know next to nothing about their processes of oral second language acquisition. Because illiterate and low print literate L2 learners rarely if ever set foot in the social world of academia in which SLA researchers operate, they have been left out of the SLA database. For example, at least since 1990, ours has been the only study published in the TESOL Quarterly that documents the SLA processes of post-critical period L2 learners with low print literacy levels (Bigelow, delMas, Hansen, and Tarone 2006). To leave these learners out of SLA research is both to deny their existence or relevance and to deny them any educational benefits that might accrue for pedagogy, from our improved understanding of the way they may learn L2s differently from literate learners.
This omission is also important for theory generalizability. Theories need to account for major accepted findings in the field if they are to be viable (Long 1990). Of course, if those accepted findings are drawn only from Population Y, and not from Population Z, then, as far as we know, our theory applies only to Population Y. If we want to know whether our theories apply to Population Z, then we will need to test them with data and findings from Population Z. This is the situation we currently face in the field of SLA. Virtually all of our findings on SLA are drawn from Population Y: a group of highly literate learners. We have almost no findings on the SLA processes of members of Population Z: low-literate and illiterate adult learners.
This omission restricts the usefulness and practical applicability of the entire SLA research enterprise. Illiterate and low-literate adults learn second languages all the time. As early as 1970, Hill stated that it was common for unschooled and illiterate individuals in remote places of the world to learn second languages. In 1980 and thereafter, some learners moved from those remote places of the world and into US cities. Large numbers of illiterate Hmong immigrants to the USA had an urgent need to learn English as an L2. Yet, when ESL teachers have asked SLA scholars for research-based advice, the scholars have had little to say to them that pertained directly to the SLA of illiterate learners. There was no research on the SLA processes of illiterate adults.
The cost of disregarding social context in SLA theory-building
The omission of low-literate or illiterate L2 learners from the SLA database can be viewed as an outcome of a more general theoretical problem in SLA research – namely, a general neglect of the social dimension in the process of SLA. In modeling L2 learner competence as an (undefined) abstraction, neutral in terms of its variable realization in different modes of use, SLA researchers have taken the position that social context simply does not matter (cf. Long 1998). Tarone (2000b, 2007, forthcoming) points out that SLA researchers have allowed their investigation of such crucial constructs as L2 learners’ ‘abstract competence’, L2 ‘input’, ‘output’, and even ‘context/setting’ to be restricted to the laboratory-like environment of the academic world, a world that is more conducive to psycholinguistic than sociolinguistic thinking. She argues that major adjustments are needed if these concepts are to make any sense in the socially embedded experiences of L2 speakers in their own worlds. And she cites studies showing that social factors in fact influence the cognitive processes of L2 learners. For example, according to Bondevik (1996), salesmen in a Minnesota electronics store did not provide corrective feedback in the way ESL teachers on the UCLA campus did in Long’s study (1980); Bondevik’s findings raise questions about Long’s claim that corrective feedback behavior is ‘universal’ (cf. Tarone 2007, forthcoming).
A very serious outcome of SLA researchers’ construction of SLA as an abstract cognitive process, universal and unaffected by social context, is that it has led to a general failure to study the process of SLA as it is routinely engaged in by a whole range of populations of L2 learners in a range of social contexts outside schools and academia. As a consequence, we know very little about the process of SLA of such learners as: completely illiterate but bi- or multilingual learners in newly industrialized countries; unschooled (but possibly functionally literate) L2 learners
An encouraging recent development has been the ‘Low-Educated Second Language and Literacy Acquisition (LESLLA) publications: Van de Craats, Kurvers, and Young-Scholten (2006) and Faux (2007).
