AUTHOR: David C. Wood TITLE: Formulaic Language and Second Language Speech Fluency SUBTITLE: Background, Evidence and Classroom Application PUBLISHER: Continuum International Publishing Group YEAR: 2010
Eleonora Luzi, Department of Linguistics, Roma Tre University
The aim of Wood's monograph -- as it is clearly stated in the Introduction -- is to fill a gap in the research on fluency, in particular between research on its temporal correlates and research on formulaic language. Their relationship is then investigated in this volume, illustrating in Part I the background on the theoretical constructs of fluency, formulaic language, cognitive processing and other social and cultural factors, presenting in Part II a research study, and discussing in Part III classroom applications.
Part I. The first chapter deals with fluency. Wood provides a sketch of the several meanings of fluency: beside the general meaning 'fluent' acquires in language teaching as a synonym of 'native-like', fluency refers to different aspects depending on whether it is employed in SLA studies (where it is usually opposed to accuracy), or in proficiency testing studies (where we have witnessed the attempt to isolate fluency from other proficiency parameters and to define it on the basis of temporal variables). In empirical research on fluency, temporal variables (i.e. speech rate) have been analyzed since the early 1980s, but only from the mid 1980s has qualitative discussion been added, relating quantitative results to strategy employment, to formulaicity and to transfer. From the survey Wood carries out in this chapter, the most common temporal variables investigated are: i) amount of speech; ii) rate of speech; iii) repair phenomena; iv) pause phenomena; and v) length of fluent runs. Chapter 1 ends with some considerations about what pause location and run length may indicate about cognitive processing in production. In particular, the distribution of pauses in production can reveal a transfer of procedural knowledge from the L1 and the reliance on declarative knowledge, or, on the contrary, automatized procedural skill to buy time in order to plan the following sequence.
Chapter 2 presents a survey of the literature on the nature and function of formulaic sequences, often addressed as fundamental in acquisition, processing and production. Wood adopts and extends Wray & Perkins' (2000) definition of 'formula', considering it a form-meaning mapping, posited midway between lexicon and syntax, stored as a single lexical unit, but also derivable from syntactic rules. Wood illustrates, moreover, the criteria presented in Wray (2002, 2008) for identifying formulas; that is, the structure, compositionality, fixedness, invariability, intonational criterion and fluency criteria. The last part of chapter 2 highlights the fundamental role of formulaic sequences in fluency. Formulaic language helps fluency through 'routines', fixed utterances without internal variation, and through 'patterns', utterances with fillable slots; both working as 'islands of reliability' (Tomasello, 2003) once they have been automatized, enhancing fluency, reducing planning, processing and encoding.
Chapter 3 provides a general and summarized theoretical background, with insight into cognitive theories. Concepts like declarative and procedural knowledge are introduced, illustrating the distinction between long-term, short-term and working memory and how they are linked respectively to automatized, fast and effortless processing, and to controlled, slow and effortful processing. Automatic processing makes certain memory nodes activate every time a given stimulus occurs: regular and repeated activation of these nodes as a consequence of the occurrence of the type of stimulus, leads to learning and automating that process, enhancing fluency. Furthermore some cognitive theories, such as instance theory (Logan, 1988), restructuring (Cheng, 1985; McLaughlin, 1990), connectionism (Ellis, 2002), and phonological memory (Baddeley, 2000) are illustrated, highlighting - despite the specific differences - the agreement on the speeding up with time and practice of the production of formulaic sequences through psycholinguistic processes. Even though the existence of formulas and their usefulness is then no longer in question, what still remains unclear is how formulas are retrieved: Wood suggests that retrieval and storage could depend on a combination of processes including frequency, pragmatic salience, repetition and practice.
In chapter 4 a short panorama of social and cultural factors that can influence L2 production is offered. They are anxiety, as a stable trait of the character or as a temporal state that varies over time, self-efficacy, connected with motivation, voice, cultural fluency and typology.
Part II. Chapter 5 illustrates the design of the research. It is a longitudinal analysis of 11 English learners. The hypotheses tested in the research deal with the increase over time of 1) the rate of production; 2) the amount of production time spent speaking as opposed to pausing; 3) the length of runs between pauses; and 4) the frequency of formulaic sequences in runs between pauses. The subjects' L1s belong to 3 different linguistic types: inflecting (Spanish), isolating (Chinese) and agglutinative (Japanese). Subjects were selected according to the result of a placement test (intermediate level) and they were shown 3 short films twice each, equally distributed, and they were asked to retell them. In quantitative analysis, temporal variables were quantified according to the following ratios and rates: phonation/time ratio, speech rate, articulation rate, mean length of run and formula/run ratio. Formulas were selected by 3 expert native speakers and only those with 66% agreement were included.
Chapter 6 reports the results of statistical analyses for each temporal variable. Quantitative results suggest that participants improved significantly over 6 months in fluency and in the number of formulaic sequences used: one-way ANOVA tests showed significance in the change over time of all variables except for one subject. Correlations for earlier and later viewings of the same film were not systematically higher than correlations between scores based on different films, suggesting that there was no effect of film prompt on temporal variables. Two-way ANOVA tests showed no significance of L1 effect. The research hypotheses seem to be supported by statistical results, since subjects recorded an increase of articulation and speed rates, a decrease in the amount of hesitation and an increased mean of length of runs and of use of formulas.
Chapter 7 presents the qualitative side of the analysis. Given the great variability in the narrative skeleton of the retellings, in the discourse strategies and in the linguistic choices, the initial idea of a one-by-one comparison (i.e. first viewing vs. second viewing) was then abandoned. The final analysis was two-fold: one qualitative analysis on narrative moves (e.g. setting the scene, conclusion) and another one focusing on formulas. The second analysis reveals the ways in which subjects make use of formulas to increase their fluency, for example using self-task formulas (e.g. 'I know', 'I think'), repeating the formula in a run, using multiple formulas to extend a run, using more marking formulas (e.g. 'at the beginning', 'this is the end of the story') and finally relying on one repeated formulas (e.g. 'then', 'next').
Chapter 8 summarizes the quantitative and qualitative results and offers clear conclusions reconnecting findings with the initial hypotheses. The need for a more extended corpus of real-communication data research to support the findings the one hand, and for more strictly controlled experimental research to highlight the cognitive implications on the other, is stated.
Part III, made up of just one chapter (chapter 9) proposes some classroom applications. There is a scientific gap in pedagogical research on how training on formulaic sequences can have positive effects on fluency. Language teaching research has been carried out only recently on learning vocabulary by exploiting collocations, chunks, and lexical phrases, but no research has been dedicated to the relation between fluency and formulaicity. Wood in this chapter suggests that instructors focus on several principles while preparing class materials, like input authenticity, interaction, production, feedback and attention to formulaic sequences. He also surveys and proposes some activities that, focusing on formula learning, can enhance and then advance fluency, like the 'mingle jigsaw' (Wood, 1998), the '4/3/2 activity' (Nation, 1989), and the Marketplace and Messenger, or class photo and family tree tasks (Gatbonton & Segalowitz, 1988).
Wood's monograph is a great piece of research, addressed in form and in content both to Second Language Acquisition and Language Testing experts, and to non-experts. The clear statement of his goals in the Introduction and the reasonable organization of the chapters help the expert reader to appreciate his methodological choices, and the non-expert reader to understand the fundamental motives behind SLA research. Clearly written, and therefore accessible to everyone, Part I offers a well-balanced review of literature in which new trends are introduced. The research gaps, highlighted in Part I, are then translated into feasible research questions and hypotheses, tested and supported by quantitative analysis, and enriched by qualitative considerations. In the final part, probably the 'icing on the cake' of the entire monograph, Wood offers real classroom application proposals: unlike most SLA research where very short space is usually dedicated to classroom application, Wood lists in detail and explains activities suitable for enhancing and improving formula learning and fluency.
Additional positive attributes are the well-organized structure of the monograph, the reasoned selection of the subjects on L1 typology, the innovative introduction to fluency research of the Formula/Run ratio as a variable relating fluency and formulaicity, and broadly speaking the scientific innovation underlying this connection, and the use of real-time data with a reduced degree of elicitation.
One limitation of the monograph is the lack of a reference to a wider linguistic framework in which all of the cited approaches can be embedded, for example, Constructionism (Masini, 2007; Luzi, 2010). Constructionism is based on the idea of construction as a form-function mapping, a more general theoretical construct in which chunks, collocations, routines, patterns, and lexical phrases can be included. Numerous relevant aspects, illustrated in Wood's monograph, have been considered in depth within this general framework -- in which several linguistic and acquisitional functionalist approaches converge -- like the role of formulas in L1 acquisition (Goldberg, 2006) or the evolutionary side of the linguistic research (Givón, 2009). A reference to this new linguistic strand would have given the entire research project a wider scientific scope, strengthening the connection not only with SLA research, but also with linguistic, cognitive and evolutionary research.
Nevertheless, ''Formulaic Language and Second Language Speech Fluency'' can be considered a good example of SLA research and a step forward toward the identification of the cognitive processes underlying language acquisition.
Baddeley, Alan D. 2000. The episodic buffer: a new component of working memory? Trends in Cognitive Science, 4. 417-423.
Cheng, Patricia. 1985. Restructuring versus automaticity: Alternative accounts of skill acquisition. Psychological Review 92. 414-423.
Ellis, Nick C. 2002. Frequency effects in language processing. Studies in Second Language Acquisition 24. 143-188.
Gatbonton, Elizabeth & Segalowitz, Norman. 1988. Creative automatization: Principles for promoting fluency within a communicative framework. TESOL Quarterly 22 (3). 473-492.
Givón, Talmy. 2009. The Genesis of Syntactic Complexity. Amsterdam, Philadelphia: John Benjamins Publishing Company.
Goldberg, Adele E. 2006. Constructions at Work. The Nature of Generalization in Language. Oxford: Oxford University Press.
Logan, Gordon D. 1988. Towards an instance theory of automatization. Psychological Review 95. 492-527.
Luzi, Eleonora. 2010. L'apprendimento di costruzioni complesse in italiano L2. PhD Dissertation: Roma Tre University.
Masini, Francesca. 2007. Parole sintagmatiche in italiano. PhD Dissertation, Roma Tre University.
Nation, Paul. 1989. Improving speaking fluency. System 17(3). 377-384.
Tomasello, Michael. 2003. Constructing a Language. A Usage-Based Theory of Language Acquisition. Cambridge (MA): Harvard University Press.
Wood, David. 1998. Making the grade: An interactive course in English for Academic purposes. Toronto: Prentice Hall Allyn and Bacon.
Wray, Alison. 2002. Formulaic language and the lexicon. Cambridge: CUP.
Wray, Alison. 2008. Formulaic language: Pushing the boundaries: Oxford: OUP.
Wray, Alison & Perkins, Michael R. 2000. The functions of formulaic language: An integrated model. Language and Communication 20. 1-28.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Eleonora Luzi graduated in Languages and Linguistics at the University of
Roma Tre, and is now concluding her PhD in Applied Linguistics at the same
university with a dissertation on the acquisition of Complex Constructions
in L2 Italian. Her research interests are Second Language Acquisition, L2
Italian, acquisition of syntax, and assessment and testing.