Linguistik online 8, 1/01

On the Functions of L2 Speech Production and Related Cognitive Processes for the Acquisition of L2 Speech Competence[1]

Research Project of the German Science Foundation (DFG)

Olaf Bärenfänger, Sabine Beyer, Karin Aguado & Jan Stevener (University of Bielefeld)

[Last change: 12-13-2002]

1 Introduction

The learners' oral output plays an important role in their acquisition of L2 competence. For instance, it is the basis for corrective feedback, it allows for syntactic processing, and it promotes the automatisation of speech production processes. Moreover, oral output is a precondition for the active testing of hypotheses and the acquisition of discourse competence. Up to now, research on the acquisition of foreign language competence has predominantly investigated the structure and the functions of the input. In contrast, we consider oral output central to L2 acquisition. We are therefore trying to identify further functions of the output and to investigate the complex interdependencies between the learners' linguistic competence and their performance. Although other (mainly socio-cognitive) factors also affect oral speech production, it essentially results from a small number of interdependent general cognitive processes, namely attention, monitoring, and automatisation. The aim of our project is to concentrate on these three factors and integrate them into a unified model of L2 speech production and acquisition.


2 The research context

2.1 Speech production models

Levelt's (1989) model of speech production is probably the most influential one in psycholinguistics. It rests upon a firm empirical basis which mainly consists of speech-error data from adult L1 speakers. A direct transfer to multilingual contexts hence seems, at least, uncertain. The three main components of this model, namely conceptualizer, formulator, and articulator, process language in an unidirectional, incremental way. Extending Levelt's model to bilingual speech production, de Bot (1992) assumes that there are two separate language-specific formulators and shared representations in the mental lexicon. Due to the static nature of both models, it has been questioned whether they are able to account for the typical variation in the performance of L2 learners and for the characteristics of multilingual speech production processes (cf. Poulisse/Bongaerts 1994; Williams/Hammarberg 1998). As both models show a low degree of flexibility, they can neither explain the reorganisation of the mental lexicon during the processes of SLA (de Bot/Woutersen/Weltens 1995; de Bot/Cox/Ralston/Schaufeli/Weltens 1995; Dufour/Kroll/Sholl, under revision) nor the simultaneous use and interaction of both languages within lexico-grammatical formulation. Finally, it is a matter of dispute if unidirectional processing and modularity should be discarded (Ellis 1999) e.g. in favour of interactive approaches which propose parallel processing and different levels of activation. It seems that such models would better be able to seize interdependencies between the processing of both languages and the variation within L2 speech production.

2.2 Attention

Attention is generally regarded a necessary precondition for successful learning. While Schmidt's Noticing Hypothesis (1990; 1995) conceptualises conscious attention as indispensable for turning input into intake, Tomlin/Villa (1994) break down the construct of attention. On the basis of psychological theories of perception they distinguish between alertness, orientation, and detection, and claim that detection in SLA may occur without awareness or even without alertness and orientation. Robinson (1995) points out that the relationship between attention and awareness in both approaches is still a critical issue and Simard/Wong (2001) recently have argued that Tomlin and Villa's model may not be generalisable to the context of SLA. In addition to previous research which has mainly dealt with characteristics of attention directed towards the input, there is recently an increasing number of studies investigating attention directed towards the output (e.g. Tarone 1983, Ellis 1987, Crookes 1989, Foster and Skehan 1996, Kormos 2000, Bygate 2001, Samuda 2001, Skehan 2001, Swain and Lapkin 2001). As we consider L2 production a central source of L2 competence (see Introduction), we do not subordinate output to input but regard both as equivalent. One pivotal part of our research design is therefore to manipulate the focus of attention directed towards the output, thus intending to clarify the complex relationship between different attention processes and the arising of L2 competence. Taking into account a close connection between attention, monitoring, and automatisation, we strive ultimately for developing an attention-based model of L2 speech production and acquisition.

2.3 Monitoring

Error detection and subsequent self-repairs are only possible if there is a specific control device. Speakers evaluate their output (both in L1 and L2) in terms of form, content, and situation which requires a specific type of attention process: monitoring. Apart from its corrective function in ongoing speech, monitoring also plays a crucial role in language acquisition (cf. Clark 1982). Although most researchers would concede monitoring processes in speech production, there is only little agreement upon their nature. Particularly the frequency and the distribution of self-repairs are suited to gain detailed insights into monitoring processes. Hesitation phenomena such as filled and unfilled pauses, repetitions, and drawls as well as prosodic features may yield supplementary insights into planning processes. Furthermore, it may be worthwhile to focus on the position and timing of cut-offs, false starts etc. This will lead to a deeper understanding of the underlying processing units and the rules according to which speakers repair their production errors (Nooteboom 1980; Blackmer/Mitton 1991).

2.4 Automatisation / automaticity

In cognitive psychology automatised processes are generally characterised as the quick and invariant performance of a large number of related tasks. They thus help save cognitive resources which can then be used for tasks requiring more attention. Automaticity has mainly been described in terms of such features as high speed, low error rate, lack of effort required, absence of general capacity use, and difficulty in controlling the process (see de Keyser 2001 for a comprehensive overview). It appears plausible to assume that these general features of automaticity also be reflected in the oral output of language learners. In consequence, several quantitative measures for L2-specific automaticity (temporal variables like speech and articulation rate, phonation/time ratio, mean length of runs, pauses, stretches of words or hesitations) are proposed as well as qualitative measures (self-repairs, recurrence of linguistic items, uniformity of syntactic structures within automatised sequences, idiomaticity etc.) (Bärenfänger 2002).


3 Research interests and aims

The major aim of our project is to foster the convergence of the different lines of research described above and to construct a unified L2-specific model of speech production and acquisition. As we conceive attention a pivotal factor in L2 oral production and acquisition, we need to investigate this variable in detail as well as the dependent cognitive factors monitoring and automatisation — both of which are characterised by attention processes or their absence respectively. Consequently, we address the following research questions:

  1. In which ways does the focus of attention affect the oral output regarding monitoring and automatisation? E.g. can monitoring be actively directed or does it operate autonomously? Are there means to suppress automaticity in oral L2 production thereby enhancing creative speaking? And, in the case of low monitoring, are there attention resources available for other cognitive tasks?
  2. Do changes in monitoring and automatisation during the course of L2 acquisition allow to trace back changes in attention? E.g. does a higher degree of automatisation in advanced learners' speech permit monitoring to operate on higher levels of speech production? And is there a typical chronology for the acquisition of automaticity across the various linguistic dimensions?

4 Methods

Oral L2 speech production is such a complex research issue that the application of a multimethod research design seems to be the most promising approach. In our study, speech production data of 16 learners of GFL are elicited and analysed by a number of different methods (triangulation; cf. Aguado/Riemer 2000). In order to guarantee the compatibility of these methods and to assess their efficiency, each method is evaluated by means of a tencriteria catalogue (cf. Bärenfänger/Stevener 2001).

The collection and analysis of the primary data consist of two major parts. In the framework of a cross-sectional psycholinguistic experiment, we systematically investigate the questions raised in the previous section under no. 1, i.e. the impact of attention upon L2 production regarding automatisation and monitoring. For this purpose, we manipulate the learners' attention within a 3 x 3 design. In each of the nine sessions, the participants solve two speech production tasks in one experimental condition. In the first task, they first read a text and then reproduce it. In the second task they are asked to describe a cartoon. Immediately after each task, the participants report their difficulties. All speech production is recorded by video, audio, and DAT. The first experimental factor directs the learners' attention towards their oral production by instructing them to focus a) on its form or b) on its content; as a control device, the learners' focus of attention c) was not directed actively. The second factor varies the social situation, i.e. learners are exposed to an interaction a) with a higher status native speaker, b) with a peer native speaker, or c) with a fellow-learner. Monitoring and automatisation are analysed in terms of the nine resulting attention conditions. By the manipulation of the first factor, we expect more monitoring regarding correctness in the case of focus on form. In the case of focus on content, we assume more monitoring regarding appropriateness. In contrast to no manipulation of the focus of attention, the former two manipulations will yield more intensive monitoring and less automatisation. As for the second factor, we assume that language learners plan their L2 production depending on the social rank of their interlocutor. The higher the rank, the more planning activity can be observed. Thus, there will be fewer automatised speech sequences and more intensive monitoring.

Based on the findings of the cross-sectional study an 11-months longitudinal study deals with the questions raised under no. 2 in the previous section. It employs a standardised interview technique to reveal changes in the learners' attention (i.e. changes in automatisation and monitoring) during the L2 acquisition process. Each of the three sessions at the beginning, in the middle, and at the end of the study lasts 20 — 50 minutes. After a short warm-up, the interviewer asks increasingly difficult questions, thus bringing the learners to the limits of their proficiency. These quasi-natural data additionally allow an analysis of progress in terms of fluency (as a consequence of automatisation in phonetics, prosody, syntax, and lexis) and permit a reliable assessment of the achieved L2 proficiency. In addition to the primary data collected in the cross-sectional and longitudinal study, secondary data such as the learners' written performance, their individual learning biography, and their self-assessments regarding their language learning progress are taken into consideration.

The large amount of quantitative and qualitative data must be integrated into a holistic interpretation. First of all, the digital audio data must be transcribed. The resulting transcripts are to be analysed by means of an L2 specific discourse analysis (cf. Henrici 1995) and to be categorized according to correctness and situational adequacy. Temporal analyses of the digital audio data will yield insights into the development of fluency and may help to identify automaticity as well as monitoring. Particularly for this purpose, we hope to refine already existing measures. Finally, the analysis of the number and quality of semantic units (propositions) will reveal semantic aspects of oral speech production and its complexity.


5 Conclusion

Recently, automatisation and monitoring are increasingly regarded as fundamental aspects of L2 learning. Our integrated, attention-based approach is intended to construct an empirically grounded L2-specific model of speech production and acquisition. It furthermore elucidates two factors considered central to successful learning, namely interaction and focus. By controlling these factors, classroom instruction will, for instance, be able to systematically enhance the development of L2 automaticity which constitutes a precondition for fluent L2 production.



1 This research project was made possible by a grant of the German Science Foundation (DFG) to Gert Henrici and Karin Aguado (HE 1326/2-1). [back]



Aguado, Karin / Riemer, Claudia (2001): "Triangulation: Chancen und Grenzen mehrmethodischer empirischer Forschung" [= Triangulation: chances and limitations of multimethod empirical research]. In: Aguado, Karin / Riemer, Claudia (eds.): Wege und Ziele. Zur Theorie, Empirie und Praxis des Deutschen als Fremdsprache (und anderer Fremdsprachen) Festschrift für Gert Henrici zum 60. Geburtstag. Baltmannsweiler: 245-257.

Bärenfänger, Olaf / Stevener, Jan (2001): "Datenerhebungsverfahren und ihre Evaluation: Ein Kriterienkatalog" [= Data collection methods and their evaluation: a criteria catalogue]. Fremdsprachen Lehren und Lernen 30: 13-27.

Bärenfänger, Olaf (2002): "Automatisierung der mündlichen L2-Produktion: Methodische Überlegungen" [= Automatisation of oral L2 production: methodological issues]. In: Börner, Wolfgang / Vogel, Klaus (eds.): Grammatik und Fremdsprachenerwerb. Kognitive, psycholinguistische und erwerbstheoretische Perspektiven. Tübingen: 119-142.

Blackmer, Elizabeth R. / Mitton, Janet (1991): "Theories of monitoring and the timing of repairs in spontaneous speech". Cognition 39: 173-194.

Bygate, Martin (2001): "Effects of task repetition on the structure and control of oral language". In: Bygate, Martin / Skehan, Peter / Swain, Merrill (eds.): Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. Harlow: 99-118.

Bygate, Martin / Skehan, Peter / Swain, Merrill (eds.) (2001): Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. Harlow.

Clark, Eve V. (1982): "Language change during language acquisition". In: Lamb, Michael E. / Brown Ann L. (eds.): Advances in Developmental Psychology (vol. 2). Hillsdale, NJ: 173-197.

Crookes, Graham (1989): "Planning and interlanguage variation". Studies in Second Language Acquisition 11: 367-383.

de Bot, Kees (1992): "A bilingual production model: Levelt's 'speaking' model adapted". Applied Linguistics 13: 1-24.

de Bot, Kees / Cox, Albert / Ralston, Steven / Schaufeli, Anneli / Weltens, Bert (1995): "Lexical processing in bilinguals". Second Language Research 11: 1-19.

de Bot, Kees / Woutersen, Mirjam / Weltens, Bert (1995): "The bilingual lexicon: Modality effects in processing". Journal of Psycholinguistic Research 24: 289-298.

de Keyser, Robert (2001): "Automaticity and automatisation". In: Robinson, Peter (ed.): Cognition and Second Language Instruction. New York: 125-151.

Dufour, Robert / Kroll, Judy / Sholl, A. (under revision): "Individual differences in second language processing: working memory capacity and translation performance".

Ellis, Nick (1999): "Cognitive approaches to SLA". Annual Review of Applied Linguistics 19: 22-42.

Ellis, Rod (1987): "Interlanguage variability in narrative discourse: style shifting in the use of the past tense". Studies in Second Language Acquisition 9: 12-20.

Foster, Pauline / Skehan, Peter (1996): "The influence of planning on performance in task-based learning". Studies in Second Language Acquisition 18: 299-324.

Henrici, Gert (1995): Spracherwerb durch Interaktion? Eine Einführung in die fremdsprachenerwerbsspezifische Diskursanalyse [= Language acquisition through interaction? An introduction to foreign language specific discourse analysis]. Baltmannsweiler.

Kormos, Judit (1999): "Monitoring and self-repair in L2". Language Learning 49: 303-342.

Kormos, Judit (2000): "The role of attention in monitoring second language speech production". Language Learning 50: 343-384.

Levelt, Willem (1989): Speaking: From Intention to Articulation. Cambridge, MA.

Nooteboom, Sieb (1980): "Speaking and unspeaking: detection and correction of phonological and lexical errors in spontaneous speech". In: Fromkin, Victoria (ed.): Errors in Linguistic Performance: Slips of the Tongue, Ear, Pen, and Hand. New York: 87-95.

Poulisse, Nanda / Bongaerts, Theo (1994): "First language use in second language production". Applied Linguistics 15: 36-57.

Robinson, Peter (1995): "Attention, memory, and the 'noticing hypothesis'". Language Learning 45: 283-331.

Samuda, Virginia (2001): "Guiding relationships between form and meaning during task performance: the role of the teacher". In: Bygate, Martin / Skehan, Peter / Swain, Merrill (eds.): Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. Harlow: 119-140.

Schmidt, Richard (1990): "The role of consciousness in second language learning". Applied Linguistics 11: 129-158.

Schmidt, Richard (ed.). (1995): Attention and Awareness in Foreign Language Learning. University of Hawai'i at Manoa.

Simard, Daphnée / Wong, Wynne (2001): "Alertness, orientation, and detection: The conceptualisation of attentional functions in SLA". Studies in Second Language Acquisition 23: 103-124.

Skehan, Peter (2001): "Tasks and language performance assessment". In: Bygate, Martin / Skehan, Peter / Swain, Merrill (eds.): Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. Harlow: 167-185.

Swain, Merrill / Lapkin, Sharon (2001): "Focus on form through collaborative dialogue: exploring task effects". In: Bygate, Martin / Skehan, Peter / Swain, Merrill (eds.): Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. Harlow: 99-118.

Tarone, Elaine (1983): "On the variability of interlanguage systems". Applied Linguistics 4: 143-163.

Tomlin, Russel S. / Villa, Victor (1994): "Attention in cognitive science and SLA". Studies in Second Language Acquisition 16: 185-204.

Williams, Sarah / Hammarberg, Björn (1998): "Language switches in L3 production: implications for a polyglot speaking model". Applied Linguistics 19: 295-333.



DFG-Projekt ‚Mündliche L2-Produktion'
University of Bielefeld
German as a Foreign Language
Department of Linguistics and Literary Studies
P.O. Box 10 01 31
D-33501 Bielefeld

fon: +49 (0)521 106-3629


URL: http://www.uni-bielefeld .de/lili/projekte/L2-pro

Linguistik online 8, 1/01

ISSN 1615-3014