LINGUIST List 28.2081

Wed May 03 2017

Review: Lang Acquisition; Ling Theories; Psycholing: Keßler, Lenzing, Liebner (2016)

Editor for this issue: Clare Harshey <clarelinguistlist.org>


Date: 08-Nov-2016
From: David Alfter <david.alftersvenska.gu.se>
Subject: Developing, Modelling and Assessing Second Languages
E-mail this message to a friend

Discuss this message


Book announced at http://linguistlist.org/issues/27/27-2842.html

EDITOR: Jörg-U. Keßler
EDITOR: Anke Lenzing
EDITOR: Mathias Liebner
TITLE: Developing, Modelling and Assessing Second Languages
SERIES TITLE: Processability Approaches to Language Acquisition Research & Teaching 5
PUBLISHER: John Benjamins
YEAR: 2016

REVIEWER: David Alfter, University of Gothenburg

Reviews Editor: Helen Aristar-Dry

SUMMARY

Part I: Theory Development

Anke Lenzing: Modelling and assessing second languages. 30 years onward. (ix–xiv).

In the introduction, Lenzing briefly sketches the joint history of, on the one hand, research on developmental stages in second language acquisition (SLA), focusing on Pienemann’s (1998) Processability Theory (PT), and, on the other hand, the application of this research in language teaching and testing. Hyltenstam & Pienemann’s (1985) edited volume Modelling and Assessing Second Languages Acquisition is taken as a starting point. The title of the reviewed volume refers to this collection. While being theoretically grounded, PT has from the beginning had strong bonds to applied linguistics. Lenzing provides a list of topics and research questions, both theoretical and applied ones, that have been addressed under the PT framework, e.g. the initial state of the L2 grammatical system, trajectories for typologically different languages, L1 transfer, the acquisition of case, language impairment, assessment/linguistic profiling, textbook analysis, and classroom studies. Since the lexical functional grammar (LFG) update (Bresnan 2001), PT has tried to theoretically explain more phenomena, foremost argument structure and discourse functions (Pienemann 2005). Thus, Lenzing argues, the theory has proven to be fruitful while it also continues to develop. This is why the editors of the reviewed volume “have not sought to impose theoretical agreement on contributors” (p. xi).

Anke Lenzing: The development of argument structure in the initial L2 mental grammatical system (3–33)

In the first contribution in Part I, Theory Development, Lenzing investigates “The development of argument structure in the initial L2 mental grammatical system”. She argues for the Multiple Constraints Hypothesis (MCH), which she put forward in Lenzing (2013). MCH, which is based on Bresnan (2001) and Pienemann (2005), claims that the L2 development is constrained in all three three levels of grammatical representation: the functional (f) level, the argument (a) level, and the constituent (c) level. Since grammar, according to LFG, is lexically driven, problems with argument structure in the initial state are attributed to the lack of syntactic features in the lexical entries for verbs. Thus, learners at this stage have to rely on direct mapping, based on universal hierarchies: the most prominent semantic role (agent) is mapped onto the most prominent function (subject), which is mapped onto the most prominent NP constituent (the first one).

In a cross-sectional and longitudinal study, oral English L2 data were collected from 24 German school children twice, the first time after one year or English classes and the second time after two years. Since the study focuses on absolute beginners, the data set, especially from the first occasion, is very small: the 24 pupils in total produces 8 questions and 49 statements at this occasion. In the second occasion, the corresponding numbers are somewhat better: 11 and 131. These utterances are analysed qualitatively for both traditional PT level and argument structure. Results show that most learners are still at PT level 1 and have not annotated their lexical entries for syntactic features after one year of studies. Thus, Lenzing argues that the initial L2 mental grammar is “highly constrained”. After two years of studies, however, most learners have reached PT level 2 or 3; moreover, during the second year of studies, syntactic features are inserted in the lexical entries for the verbs.

The study is methodologically and theoretically solid. Moreover, it is well-written. Focusing on absolute beginners is difficult but also important if we want to understand what SLA actually is. The qualitative approach is praiseworthy, but the use of “statistics” is problematic: e.g. it is claimed that 50% is “far more” than 27% even though only 8 tokens are analysed. This might mislead the reader to assume that the results can be generalized.

Satomi Kawaguchi: Question constructions, argument mapping, and vocabulary development in English L2 by Japanese speakers: A cross-sectional study (35–63)

Just like Lenzing, Kawaguchi studies the relation between traditional PT stages and argument structure. She also relates these developments to vocabulary size. The study is cross-sectional, with vocabulary size as independent variable. Nine participants are chosen, three with a vocabulary size of around 4,000 words; three with a vocabulary size of around 7,000 words; and three with a vocabulary size of around 11,000 words. With translation tasks, she elicits structures that need non-default mapping, both lexical (intransitive unaccusative verbs and transitive psych verbs) and structural (passives and causatives) non-direct mapping. There are clear trends: the greater vocabulary, the better grammatical processing. However, the great divide goes between the three learners with a very small vocabulary and the rest of the learners. From a vocabulary size of 6,000 words onwards, the highest developmental stage, with non-default mapping (marked alignment), is reached. Lexical non-default mapping seemed more difficult than structural non-default mapping; the former caused problems for both the mid and small size vocabulary groups, whereas the latter caused problems only for the small vocabulary size group.

Kawaguchi argues that this research provides insights that can be used in language teaching in order to make more learners reach an advanced L2 English level, but how this the results can be used in teaching is not made clear. She also argues that translation tasks can be used more in PT research, which is normally based on oral production; translation tasks are easier than oral elicitation for language teachers wanting to test their students.

Gisela Håkansson: Processability Theory and language development in children with Specific Language Impairment (65–78)

With traditional PT analysis, Håkansson conducts a re-analysis of data elicited from 10 Swedish children with specific language impairment (SLI). SLI children have been reported to have problems with different phenomena in different languages, and several explanations for these problems have been put forward. By looking at the ten children at an individual level, and by treating them as language learners, having reached different developmental stages, we can, Håkansson claims, understand what underlies different problems at the surface structure. One important point is that, e.g., tense morphology in English and Swedish respectively rely on different processes; that is why English children with SLI have problems with tense whereas Swedish ones do not. The main point of the study is that the PT hierarchy seems to hold for children with SLI, and thus can be used to compare these children cross-linguistically. This requires that children with the diagnosis are considered language acquirers and that they, thus, are investigated at an individual level.

The article is well written and manages to unify contradictory theories concerning SLI children's development using PT. It is not clear to us why this article is placed in the theory part of the book; it is rather a good example of how the PT can be applied.

Manfred Pienemann, Anke Lenzing & Jörg-U. Kessler: Testing the Developmental Moderated Transfer Hypothesis: The initial state and the role of the L2 in L3 acquisition (79–98)

In the last contribution of part 1, the theoretical debate becomes intense. The Developmental Moderated Transfer Hypothesis, one of the original hypotheses within PT, states that L1–L2 transfer, either positive or negative, will never take place before the learner is ready to process the structure to be transferred. This should explain why German speakers learning Swedish and vice versa have problems with the verb second (V2) rule, despite the same rule applying to both languages (Håkansson 2002). In the article, the authors first refute arguments from Bohnacker’s (2006) and Bardel and Falk’s (2007) attempts to show that this phenomenon is to be explained by the L2 in both cases being English. The authors are harsh in their criticism of these studies, which are accused of being both methodologically and theoretically weak. Thereafter, they present a replication of Bardel and Falk’s second study: a very small amount of Swedish was taught to seven German speakers. Three of them had learnt other V2 languages before; four of them had not; but all of them had learnt English to a high level. They memorized 40 Swedish words and then they had one communicative Swedish class. They were then tested and no one productively produced verb–subject inversion. Nor did they use ‘do’ support with negation, which would be expected from the Bardel and Falk suggestion that L2 always overrides L1 in L3 acquisition. The study is small but well conducted. For anyone having an interest in theories of SLA, it’s a joy to read this article, but the tone can be a bit harsh.

Part II: Theory Assessment

Yanyin Zhang & Bo Liu: The 'tense' issue: Variable past tense marking by advanced end-state Chinese speakers of L2 English

This article investigates the variable past tense -ed marking in Chinese learners of English as L2 even at high proficiency levels. The study concentrates on spoken language, as the Chinese speakers in other studies have been shown to produce past tense -ed in the written mode at a high level but fail to produce past tense -ed in the spoken mode with the same accuracy. The nine participants are interviewed for about 50 minutes each and a quite large number of obligatory contexts for past tense (of different types of verbs) are analysed after a thorough exclusion process. The hypothesis is that speakers who had learned English at top notch universities in China would produce past tense -ed to a higher degree than speakers who did not attend top notch universities.

Based on the suggestion from an anonymous reviewer, the authors also analysed obligatory contexts for plural -s. The idea was to test the hypothesis from Pienemann (1998) that those learners making “bad choices” at an early stage will not develop as well as those making better choices. However, it is not clear how this hypothesis is tested with this data, and even less clear how the hypothesis is corroborated. The seven participants that produced the past tense suffix to a high degree also produced to plural –s to a high degree. Of the two participants that did not produce the past suffix to the same extent, one also had problems with the plural –s; the other one had not. However, all learners were better in producing the plural –s than the past tense suffix. Thus, the conclusion of the authors that the “bad choice” has an effect is far fetched. Furthermore, the division between lexical and phrasal plural –s markings, originating from Pienemann (1998), is, we claim, a misconception, which is also evident from the results in the present study: there is no difference between plural –s marking on nouns preceded by words like “two” or “those” and on nouns not preceded by such words.

We do not know how much the participants have studied outside university, but it is claimed that some have done a considerable amount of self study to attain near-native competence. As all informants had lived outside of China for at least 10 years, and quite some in English speaking countries, we are unsure about the reliability of the conclusion that top notch university students produce past tense -ed with higher frequency and make fewer ''bad choices'' than non top notch university students. Moreover, we do not know how they were selected to the top notch universities – were they better at English before starting to study at the university?

Jana Roos: Acquisition as a gradual process: Second language development in the EFL classroom

The point of departure for this article is three questions: what should be taught, when and how? According to the author, the two first questions can be answered by PT. To answer the third questions, she turns to the theory/methodology of task-based language learning. Using data from an earlier study, she exemplifies how communicative tasks can be used both for assessing the readiness of learners and for giving learners opportunities to practice specifically those structures that they are developmentally ready for. Thus, the article demonstrates in a clear way how PT and task-based learning can benefit from each other in an applied context.

Katharina Hagenfeld: Psychometric approaches to language testing and linguistic profiling - A complementary relationship?

In this chapter, Hagenfeld investigates whether linguistic profiling can help with the shortcomings of the CEFR evaluation method. She claims that CEFR is a broad classification method, whereas PT allows a more fine-grained classification. However, PT only focusses on grammar, while CEFR is focused on what tasks can be handled with language. Furthermore, CEFR lists assessment criteria but it is not a classification tool and does not provide classification tools; CEFR is also not applicable to the very beginner levels, as the first level in CEFR (A1) requires some basic language control which is beyond PT stages 1 and 2, according to a pilot study by Lenzing and Plesser (2010).

The author investigates the correlation between two Linguistic profiling tools and the CEFR scale: Rapid Profiling (RP) (Mackey, Pienemann & Thornton 1991; Pienemann & Mackey 1993; Keßler 2006, 2008) and Autoprofiling (AP) (Lin 2012).

The results show that RP and AP tend to agree on the assigned stage of development and that this assessment is in line with the CEFR assessment. However, the tested PT stage (5) covers three different CEFR levels. The only claim that can be made is that there seems to be a correlation between CEFR levels and PT stages at the higher stages. The author claims that AP is still an early project and needs further refinement; eventually it could be a replacement for RP, which requires a trained human profiler; however, AP only focuses on written production whereas RP also takes into account spoken production

The author points out the tentative nature of the results. The study itself is interesting, but we wonder about its applicability.

Esther Maier, Lea Neubauer, Katharina Ponto, Stefanie Couve de Murville & Kristin Kersten: Assessing linguistic levels of L2 English in primary school programs

In this chapter, the authors compare the outcomes of different study programmes on the development of English skills in children in three different German schools. They look at immersive programs, where teaching is done by native English speakers and exposure to the English language is about 70% or more; and traditional programs where English is taught two hours per week by native speakers of German. Furthermore, they investigate whether certain variables such as sex, prior exposure to the L2 or home language influence the language acquisition process.

The study is well done and interesting. The number of participants is quite high (n=105), which gives more weight to the possible conclusions. The authors devise several tasks for the elicitation of certain linguistic features, all of which are suitable for children. They also test whether the tasks they have chosen are motivating and comprehensible for the children.

Not surprisingly, the authors find that L2 exposure and L2 intensity correlate positively with attained proficiency. They also find that prior experience/exposure to the L2 such as having attended a bilingual preschool has a significant effect. However, it is not clear whether this prior exposure overrides the later immersive programs or whether they are complementary. As the authors used the old version of PT, they quickly ran into a ceiling effect where learners reached the highest level at an early stage; hence, they could not account for further development in the learners.

Jörg-U. Keßler & Mathias Liebner: Diagnosing L2-English in the communicative EFL Classroom: A task-based approach to individual and developmentally moderated focus on form in a meaning-focused setting

In this chapter, the authors address the lack of personalized instruction in heterogenous EFL classrooms. As the range of attained proficiency varies, it would be favorable to give each learner tasks depending on their stage of proficiency. The authors use PT as a framework and Rapid Profiling to determine the PT stages of individual learners. They then use Task-based Language Teaching (e.g. Ellis 2003, Eckerth & Siepmann 2008) to provide learners with tasks suitable for their current level.

The approach is very interesting and of importance if we want to move towards individualized teaching. The authors claim to have tested the approach on grade 8 students, but we missed a results and evaluation section.

Henning Rossa: The cognitive processes elicited by L2 listening test tasks - A validation study

This article focuses on the interface between SLA and language testing, and specifically on the validity of a listening comprehension task in an EFL test. Comprehension is described as a complex cognitive process, and the question is what mental processes test-takers engage in while attempting to solve comprehension tasks and what the relation between these processes and the test outcome is. Extreme cases are sampled from a larger population: 9 good test-takers and 9 ones with bad results; they are all grade 9 students in Germany studying English. The participants were asked to think aloud while solving multiple choice tasks. In total 329 instances of think-aloud data were analysed: did the test-takers recall the necessary informationen needed to solve the question? Though the study does find instances where the test-taker draws the right conclusion from faulty reasoning, and instances where there the test-taker reasons correctly but draws the wrong conclusion, the general conclusion is that the test is valid: the comprehension test does actually test the test-takers ability to recall information and generate appropriate inferences. The study, both qualitative and quantitative in its approach, is well written and carried out in a thorough way.

EVALUATION

The chapters in this book are loosely held together by the theme of PT. They are mostly stand-alone contributions that look at different aspects of PT from different angles. As the editors point out, they want to give the reader an overview of different positions held by researchers in the PT community and stimulate the reader's own thinking. As such, the reader can pick selected chapters or read the chapters out of order. As each article provides sufficient background information, the articles are easily understandable.

Overall, the book presents an interesting selection of articles with different perspectives which give a good overview of recent research in the PT community.

This book is suitable to students, researchers, people wanting to know how PT is developing. However, the reader should have a certain familiarity with PT before reading the book.

REFERENCES

Bardel, C. & Falk, Y. (2007). The role of the second language in third language acquisition: The case of Germanic syntax. Second Language Research, 23(4), 459-84.

Bohnacker, U. (2006). When Swedes begin to learn German: From V2 to V2. Second Language Research, 22, 1-44.

Eckerth, J. & Siepmann, S. (2008). Research on task-based language learning and teaching. Theoretical, methodological and pedagogical perspectives. Frankfurt: Peter Lang.

Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.

Hulstijn, J. H., Ellis, R. & Eskildsen, S. W. (2015). Orders and sequences in the acquisitoin of L2 morphosyntax, 40 years on: An introduction to the special issue. Language Learning, 65(1), 1-5.

Håkansson, G., Pienemann, M. & Sayehli, S. (2002). Transfer and typological proximity in the context of second language processing. Second Language Research, 18(3), 250-273.

Keßler, J-U. (2006). Englischerwerb im Anfangsunterricht diagnostizieren: Linguistische Profilanalysen am Übergang von der Primarstufe in die Sekundarstufe I. Tübingen: Narr.

Keßler, J-U. (2008). Communicative tasks and second language profiling: Linguistic and pedagogical implications. In J. Eckerth & S. Sieckmann (Eds.), Task-based language learning and teaching. Theoretical, methodological, and pedagogical perspectives (pp. 291-310). Frankfurt: Peter Lang.

Lenzing, A. (2013). The development of the grammatical system in early second language acquisition - The Multiple Constraints Hypothesis. Amsterdam: John Benjamins.

Lenzing, A. & Plesser, A. (2010). Challenging the scope-precision dilemma in language testing: The common European framework and linguistic profiling. Paper presented at the 10th International Symposium of Processability Approaches to Language Acquisition (PALA). University of Western Sydney, Australia, 19-21 September.

Lin, B. J. (2012). Is automatic linguistic profiling feasible in an ESL context? Unpublished PhD dissertation. University of Newcastle.

Mackey, A., Pienemann, M. & Thornton, I. (1991). Rapid Profile: A second language screening procedure. Language and Language Education, 1(1), 61-82.

Pienemann, M. (2005). Cross-linguistic aspects of Processability Theory. Amsterdam: John Benjamins.

Pienemann, M. (1998). Language processing and second language development.
Processability Theory. Amsterdam: John Benjamins.

Pienemann, M. (1984). Psychological constraints on the teachability of languages. Studies in second language acquisition, 6(02), 186-214.

Hyltenstam, K. & Pienemann, M. (Eds.) (1985). Modelling and assessing second language acquisition. Clevedon: Multilingual Matters.

Pienemann, M. & Mackey, A. (1993). An empirical study of children's ESL development and Rapid Profile. In P. McKay (Ed.), ESL development: Language and literacy in schools, Vol 2: Documents on bandscale development and language acquisition (pp. 115-259). Canberra: National Languages & Literacy Institute of Australia and Commonwealth of Australia.


ABOUT THE REVIEWER

David Alfter is a PhD students in Computational Linguistics at the University of Gothenburg. His research focuses on modeling learners in CALL systems.

Anders Agebjörn is a PhD student in Swedish as a second language at the University of Gothenburg. His interests are grammar and second language acquisition.

Page Updated: 03-May-2017