The study also highlights the constructs of current linguistic theory, arguing for distinctive features and the notion 'onset' and against some of the claims of Optimality Theory and Usage-based accounts.
The importance of Henk Zeevat's new monograph cannot be overstated. [...] I recommend it to anyone who combines interests in language, logic, and computation [...]. David Beaver, University of Texas at Austin
EDITOR: Fethi Mansouri TITLE: Second Language Acquisition Research SUBTITLE: Theory-Construction and Testing PUBLISHER: Cambridge Scholars Publishing YEAR: 2007
Zhisheng Wen (Edward), Hong Kong Shue Yan University
This edited volume contains major papers delivered at the ''5th International Symposium on Processability, Bilingualism and Second Language Acquisition'' that was held at Deakin University’s Melbourne Campus, 26-28 September 2005. Despite its ambitious title, it mainly focuses on promoting and testing Manfred Pienemann’s Processing Theory (PT) as applied to a number of typologically different languages (e.g. English, Japanese, Swedish and Chinese).
Chapter 1 is the conventional introductory chapter written by the editor. Mansouri begins with a brief account of historical trends and developments of SLA research with respect to developments in morpheme order studies, thus setting the scene for bringing Pienemann’s PT within the broader context of SLA research. The second section further articulates how various chapters of the book contribute to the two key themes running through the book, i.e. theory construction and testing of PT, and empirical studies in bilingualism and speech processing that are intended to provide further evidence for the major claims stipulated by PT.
Chapter 2 provides an overview of PT written by its key proponent Manfred Pienemann himself. Pienemann begins by acknowledging the sources of inspiration for PT (i.e. Levelt’s speech production model and Bresnan’s lexical functional grammar) and reiterating its four basic premises pertaining to language processing or the grammatical processor (p. 13): being (1) largely automatic and not consciously controlled; (2) incremental and cumulative; (3) linear in its output; (4) subserved by a short-term memory store. The unique feature of PT, as claimed by Pienemann, is that it possesses the capacity to predict developmental trajectories for any second language, which are in turn indexed by key developmental stages (p. 14). In other words, there is a hierarchy in the process ability of various grammatical structures. Between these hierarchies, there is also a certain degree of flexibility for the shape of L2 grammar that leaves space for finer-grained hypotheses to be formulated by researchers (the Hypothesis Space and Development Dynamics). In the second section of the chapter, Pienemann summarizes the major empirical evidence in PT-based research, which culminates in the ultimate goal of accommodating typologically different languages.
Following the two introductory chapters, Chapter 3 by Kawaguchi presents the first empirical study in this book, which aims to provide evidence (collected from a longitudinal case study of an 18-year-old Australian girl learning Japanese as a second language) to support the Unmarked Alignment Hypothesis and the Lexical Mapping Hypothesis recently postulated within the tenets of Pienemann’s PT. Despite some minor exceptions that need to be further explored, the overall patterns of data converge to corroborate predictions made by the these two specific hypotheses of PT in that canonical mapping indeed preceded non-canonical structures and that the variety of non-canonical structures increased in time.
In Chapter 4, Håkansson and Norrby focus on the “Steadiness Hypothesis” of PT which claims that “the level of grammatical process ability is assumed to be steady across different tasks as long as these tasks are based on the same skill type in language production” (p. 81). To test this hypothesis, the authors drew on written (composition and translation tasks) and oral (elicited communicative tasks) data from 20 students learning Swedish either as a foreign language (with nine from the Melbourne Group) or as an L2 (11 from the Malmo Group). Results from both data sets supported the PT hierarchy despite their differences in modality (written vs. oral), thus confirming the “Steady Hypothesis”. In other words, there is an implicational order of structures and no gaps in the hierarchy of process ability.
With a view to complementing previous PT studies in their emphasis on the predictive power of the theory for language-specific acquisition stages, Mansouri and Håkansson’s study described in Chapter 5 builds up a strong case and provides preliminary evidence for incorporating intra-stage developmental sequences as an additional explanatory tool for the Hypothesis Space. Citing data from Arabic as a second language (ASL), with a focus on definiteness markers, the authors argue that zero (null) and reduced marking in ASL can be accounted for in terms of processing requirements and typological features (form-function mappings). This study provides an additional perspective to the Hypothesis Space as an explanatory module.
In a slightly different manner, Chapter 6, by Keßler, reports on a feasibility study of the computer-assisted procedure of “Rapid Profile” as a tool for conducting online assessment of EFL learner’s language development. The study seeks justification for that claim that Rapid Profile offers PT researchers a reliable and valid means of diagnosing EFL-development in formal L2 classroom settings. Indeed, Keßler’s study reveals an inter-rater reliability of 85.7 per cent, thus rendering the Rapid Profile a feasible diagnostic tool for online-assessment. In addition, this data collection procedure has significant implications for EFL teachers in that these language learning profiles captured by the procedure can facilitate a smooth transition from the early start to a more target language like final state of acquisition in classroom settings.
Chapter 7 by Zhang tests the Topic Hypothesis in the case of acquiring Chinese as a second language. The Topic Hypothesis predicts that learners go through three stages in their acquisition of L2 syntax, “beginning with a canonical order and progressing toward a non-linear order of sentence structure” (p. 147). The key thrust of the Topic Hypothesis lies in its assumption that the latter order deviates from the linearity principle of mapping between argument, functional and constituent structures. The language data collected by Zhang from three English native speakers at an Australian university largely support the principle of the Topic Hypothesis, showing the successive disassociation between topic and subject elements, and between grammatical functions and sequential positions in a sentence.
Chapters 8 to 11 are organized around the second theme of the book, i.e. bilingualism and speech processing within the PT paradigm. In Chapter 8, Itani-Adams looks into the lexical and grammatical development patterns of a Japanese-English bilingual child (age 1;11 to 4;10) who was raised in a “one parent one language” home environment since birth. Data analysis focuses on two aspects of the relationship between lexicon and grammar as found in both languages: on the one hand, verb stems and verbal morphemes; on the other, verbs and the semantic function of arguments. Overall, the data suggest that, regardless of different input languages, nouns are bootstrapped by the bilingual child into each language while verbs showed different patterns of development that were similar to monolingual children, thus indicating that lexical learning is a language-specific operation. That is to say, Japanese and English each developed in a separate but a parallel manner in this bilingual child, thus providing evidence in support of the Separate Development Hypothesis for first language acquisition.
By adding the underlying cognitive mechanism of short-term memory (STM) store into the picture, Suarez and Goh’s study in Chapter 9 investigates the codification process among bilinguals with different levels of English/Chinese dominance. In line with the modality model of STM, the authors conducted two experiments and manipulated the phonological and visual features of words and examined their influence on the degree of semantic proactive interference (PI) in a short-term cued recall task. The results of both experiments suggest that the codification process of these bilinguals depends largely on their dominant language. Chinese dominants seem not to use phonological nor visual strategies to memorize in a cued recall task, while mixed dominants recode phonologically and store the information in a phonological code. Equally interesting is the finding that English dominants who were less proficient in Chinese also seemed to experience more severe visual and phonological interference, while Chinese dominant bilinguals did not show any evidence of this influence on PI, indicating that they have a very integrated phonological, visual and semantic memory system.
Turning away from bilingual to multilingual processing, Van den Noort, Bosch and Hugdahl’s study in Chapter 10 explores the processing of relative clauses among 20 multilingual learners who were all native Dutch speakers (L1), are fluent in German (L2), and started learning Norwegian (L3). Previously, L1 research has shown that subject relatives are easier to comprehend than object relatives. The authors test the hypothesis that object relatives cause a greater working memory load. Ten participants started their free acquisition of Norwegian (L3) in the 6 months prior to the study (early stage L3 learners), whereas the other ten started their acquisition of Norwegian more than 3 years before (more advanced L3 learners). Participants were administered a relative clause task in all three languages, a reading span task (verbal working memory) in Norwegian (L3) and a number ordering task. The results show that differences in subject and object relatives can only be found among participants who are in an advanced stage of third language acquisition. In addition, both verbal and non-verbal working memory task scores were not correlated with the total comprehension score on the relative clause task in all three languages, which is in line with the Separate Sentence Interpretation Resource theory of working memory.
The final chapter of the book, Chapter 11 by Kim and Kwon, proposes a Parallel Developmental Sequences (PDS) model for SLA that integrates three separate L2 developmental modules: procedural developmental sequence, syntactic developmental sequence, and morphological developmental sequence. Emulating the logic of the connectionist Parallel Distributed Processing (PDP) model in cognitive psychology, the PDS model proposes that L2 development follows a series of parallel developmental stages sequentially: with one of the three mechanisms (i.e. the procedural, the syntactic, and the unificational mechanism) being activated first, and then interconnected, and finally synchronized with each other before moving to the next state of the PDS (p. 249). The PDS model is put to the test by empirical research on sentence construction tests through a real-time experimental method. Responses from 136 Korean college students were collected and analyzed, culminating in an implicational scale vis-à-vis their developmental stages in terms of PDS. The results showed that the implicational scale reached a satisfactory level of .83, thus rendering the PDS model a viable framework for gauging interlanguage development of L2 learners.
To sum up, this edited volume has significantly broadened the theoretical bases and the application areas of Pienemann’s Processability Theory (PT) within the SLA research arena. First, on the theoretical front, the book does not just further recapitulate the major tenets of PT (Chapter 2) -- which is important and necessary for readers who are new to the theory -- but also successfully extends the existing PT paradigm to include a conceptual framework that serves to account for intra-stage developmental sequences (Chapter 5) and also a more ambitious PDS model (Chapter 11) that integrates multiple parallel procedures in L2 interlanguage development. On the applied side, the book also adds further dynamics to the PT paradigm by reinforcing the computer-assisted data collection procedure of Rapid Profile (Chapter 6), which will have significant implications for L2 classroom research and practice.
The most important contribution of this edited volume, as I see it, is its initial efforts to incorporate the construct of short-term memory (Chapter 9) and working memory (Chapter 10) as a postulated cognitive mechanism subserving the processing of grammatical structures by L2 learners. Despite the unimpressive findings as reported in both studies with the working memory tasks (verbal and non-verbal) in language processing, there are ample reasons to believe that if working memory had been operationalized differently and tested by different measures, the results could be quite different (cf. Wen, 2012). In other words, it is still quite likely that the limited capacity of working memory should have significant impact on the processing of L2 grammatical structures. Further PT-based studies can follow on this line of development to demystify the role of working memory in L2 processing.
Pienemann clearly indicates in his introduction to PT in Chapter 2 that PT as a theory draws heavily on Levelt’s speech production model. However, it should be noted that Levelt’s (1989) model is mainly based on L1 speech production, which tends to be largely automatic (though see Hartsuiker & Barkhuysen, 2006 for a different view). When it comes to L2 learning then, it is quite likely that L2 learners need to draw on more working memory resources in processing L2 grammatical structures as opposed to L1 processing (McLaughlin, 1995). That is to say, the basic premises of PT as stated by Pienemann (p. 13) may need to be re-evaluated and augmented in light of such limited capacity of working memory and the consequences they may in turn bring to bear on the processing of grammatical structures in L2 (Payne & Ross, 2005). More empirical studies following this line will prove to be critical for PT to make its foray into a more viable framework accounting for cross-linguistic influences among L2 learners in processing typologically different languages.
Hartsuiker, R.J., & P. N. Barkhuysen (2006). Language production and working memory: The case of subject-verb agreement. Language and Cognitive Processes, 21, 181-204.
Levelt, W. J. M. (1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press.
McLaughlin, B. 1995. Aptitude from an information processing perspective. Language Testing, 11: 364-381.
Payne, J. S. and Ross, B. (2005). Working memory, synchronous CMC, L2 oral proficiency development. Language Learning and Technology, 9(3), 35-54.
Wen, Z. (2012). Working memory and second language learning. International Journal of Applied Linguistics, 22: 1–22. doi: 10.1111/j.1473-4192.2011.00290.x
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Zhisheng Wen (Edward) is an Assistant Professor in the Department of
English at Hong Kong Shue Yan University. His current research foci are
theoretical and methodological issues surrounding “Working memory as
foreign language aptitude” in SLA. Dr. Wen’s papers have appeared in many
academic journals and his monograph “Working memory and second language
learning” will be published by Multilingual Matters. He is the recipient of
several international research awards and grants, including the recent 2012
'Language Learning' Roundtable Conference Grant which enabled him to
convene an International Roundtable on “Memory and SLA” in Hong Kong (June
2012). Following the Roundtable, he is now working on an edited volume and
a special journal issue on working memory and SLA.