Wen, Zhisheng. (2014).Theorizing and measuring working memory in first and second language research. Language Teaching, 47/2, 173-190.

Language Teachinghttp://journals.cambridge.org/LTA

Additional services for Language Teaching:

Email alerts: Click hereSubscriptions: Click hereCommercial reprints: Click hereTerms of use : Click here

Theorizing and measuring working memory in rst andsecond language research

Zhisheng Wen

Language Teaching / Volume 47 / Issue 02 / April 2014, pp 174 - 190DOI: 10.1017/S0261444813000517, Published online: 27 February 2014

Link to this article: http://journals.cambridge.org/abstract_S0261444813000517

How to cite this article:Zhisheng Wen (2014). Theorizing and measuring working memory in rst and second languageresearch . Language Teaching, 47, pp 174-190 doi:10.1017/S0261444813000517

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/LTA, IP address: 42.98.133.22 on 28 Feb 2014

http://journals.cambridge.org Downloaded: 28 Feb 2014 IP address: 42.98.133.22

Lang. Teach. (2014), 47.2, 174–190 c© Cambridge University Press 2014doi:10.1017/S0261444813000517

Research Timeline

Theorizing and measuring working memory in first and secondlanguage research

Zhisheng Wen Hong Kong Shue Yan University, Hong Kong, [email protected]

Working memory (WM) generally refers to the human ability to temporarily maintain andmanipulate a limited amount of information in immediate consciousness when carrying outcomplex cognitive tasks such as problem-solving and language comprehension. Though muchcontroversy has surrounded the WM concept since its inception by Baddeley & Hitch (1974),an increasing number of cognitive psychologists have accepted WM as a multi-componentsystem comprising both domain-specific storage mechanisms and domain-general executivefunctions (Miyake & Shah 1999; Baddeley 2012; Williams 2012). Such a FRACTIONATED viewof this cognitive construct manifests itself clearly in distinct strands of WM-language research,where two contrasting research paradigms have emerged (Wen 2012).

WM researchers following the European tradition have sought to establish the critical roleplayed by the phonological component of WM (henceforth PWM, which embraces a passivesound-based store and an active sub-vocal rehearsal process) in vocabulary acquisition andgrammar development (e.g., Gathercole & Baddeley 1993; Gathercole, Baddeley & Papagno1998; Baddeley 2003). To measure WM, these researchers usually administer a simple storage-only memory span task such as the digit span, word span or nonword repetition span task, andadopt participants’ total recall score as their WM capacity (Gathercole et al. 1994; Gathercole2006). In contrast, many cognitive psychologists based in North America have tended to em-phasize the executive functions associated with the WM concept (henceforth EWM) and focuson teasing out the implications of its attention-regulating mechanisms (e.g., information updat-ing, task-switching, and inhibitory control; Miyake & Friedman 2012) for language learningand processing. The WM measure they choose is usually a complex memory span task thatuses the storage and processing functions of WM, such as the reading span task designed byDaneman & Carpenter (1980). Extensive research following this paradigm has also pointedto significant effects of WM on selective facets of language activities such as comprehensionprocesses, syntactic processing, and speech production (Miyake & Friedman 1998). Takentogether, both research camps of WM have accumulated increased evidence for a close linkbetween WM and L1 learning (Gathercole & Baddeley 1993; Baddeley 2003; Cowan 2011).

A growing number of second language acquisition (SLA) researchers are adopting thesewell-defined research paradigms of WM–language in cognitive psychology and increasinglysubscribe to the view that WM should play an equal if not more important role in thelearning and processing of a foreign or second language (L2). The rationale behind thisassumption came mainly from the perceived fundamental difference between first language

http://journals.cambridge.org


Z H I S H E N G W E N : T H E O R I Z I N G A N D M E A S U R I N G W O R K I N G M E M O R Y 175

acquisition (L1A) and L2 learning. Viewed from the information-processing perspective, L1Aunequivocally occurs early in life and proceeds automatically, as reflected in children pickingup their mother tongue relatively easily. L2 learning, in contrast, usually comes at a laterstage in life and relies on more controlled language processing and thus presumably placesmuch greater demands on such cognitive resources as WM and/or attention (Harrington1992; McLaughlin 1995).

An alternative approach to linking WM functions with the process of SLA can be deducedfrom the basic tenet of the connectionist account of language acquisition. Interpreted from thisperspective, a central task of language acquisition (be it L1A or SLA) is to acquire, store, anduse a functioning repertoire of all sorts of linguistic sequences ranging from lexis or multi-wordformulae to morpho-syntactic constructions (Ellis 1996, 2012). In the process of chunkingthese linguistic sequences, WM, and particularly its PWM component, serves to sustain andrehearse (albeit sub-vocally) their phonological material in immediate consciousness whilelong-term representations are being constructed.

Such intriguing portrayals of the WM–SLA association have motivated an increasingnumber of empirical studies exploring the potential effects of WM in various aspects of SLA(Juffs & Harrington 2011; Wen forthcoming). Nevertheless, due to the elusive nature of theWM construct and the daunting array of assessment procedures in cognitive psychology,many of these WM–SLA studies are subject to limitations and caveats in terms of researchdesign and methodology, thus posing great challenges to many SLA researchers (Juffs 2006;Gass & Lee 2011).

To resolve these theoretical and methodological issues surrounding the WM constructand its assessment, the SLA field requires a more principled way of conceptualizing andoperationalizing the construct, one that is conducive to future research. In pursuit of that goal,I have proposed a conceptual framework for WM that integrates the general consensus aroundWM theories in cognitive psychology into existing SLA research paradigms (Wen 2012,forthcoming). The present timeline aims to document and highlight selected seminal researchfrom both fields that has contributed to the formulation of this INTEGRATED perspectivefor theorizing and measuring the WM construct in L1 and L2 research. These strands oftheoretical and empirical research are organized and coded into the following four categories,which constitute the broad themes underlying this timeline:

A Conception of WM in cognitive psychologyA1 Theorizing or conceptualizing PWMA2 Theorizing or conceptualizing EWM

B Assessment of WM in cognitive psychologyB1 Designing and constructing assessment procedures for testing PWMB2 Designing and constructing assessment procedures for testing EWM

C Theorizing and measuring WM as it relates to L1C1 Theorizing and measuring PWM as it relates to L1C2 Theorizing and measuring EWM as it relates to L1

D Theorizing and measuring WM as it relates to L2D1 Theorizing and measuring PWM as it relates to L2D2 Theorizing and measuring EWM as it relates to L2

More importantly, when these lines of WM–language research from both cognitivepsychology and SLA are put into perspective, they seem not just to converge on the



1 7 6 R E S E A R C H T I M E L I N E

FRACTIONATED nature of WM but to further corroborate the SEPARATE and DISTINCTIVE rolesof PWM and EWM in specific language-learning activities (Wen forthcoming). Emergingpatterns of this PWM–EWM dichotomy seem to suggest that the PWM component,with its associated cognitive mechanisms, is most closely related to the ACQUISITION andDEVELOPMENTAL aspects of language-learning domains such as vocabulary and formulaacquisition and grammar development. In contrast, the EWM component, with its embeddedexecutive functions, is particularly relevant to the PROCESSING and real-time PERFORMANCE

aspects of language-learning activities, such as certain demanding cognitive processes involvedin language comprehension and production (Kail & Hall 2001; Engel de Abreu & Gathercole2012; Szmalec, Brysbaert & Duyck 2013). This proposition and other associated hypothesesregarding the different effects of PWM and EWM, as they relate to specific L1A andL2 learning domains and activities, are encapsulated in the PHONOLOGICAL/EXECUTIVE

HYPOTHESIS (the P/E model; Wen forthcoming). To incorporate this demarcation betweenPWM and EWM into this research timeline, each of its four broad themes is furthercategorized into two subordinate themes.

To sum up, once PWM and EWM, the two key components of WM for language, arepinned down and their associated functions/mechanisms aligned with specific L1A andL2 learning domains, NOVEL, SPECIFIC, and TESTABLE hypotheses regarding their intricaterelationships can be formulated in a more effective manner. In light of these proposals forthe WM–language nexus, future research can set out to either support or refute hypothesesoutlined in the P/E model and to further tease apart all their possible theoretical andpedagogical ramifications.

Acknowledgements

The author thanks Graeme Porte, Peter Skehan, and two anonymous reviewers for Language

Teaching for constructive comments on earlier drafts of this research timeline.

References

Baddeley, A. D. (2012). Working memory: Theories, models and controversies. Annual Review of Psychology63, 1–30.

Engel de Abreu, P. M. J. & S. E. Gathercole (2012). Executive and phonological processes in secondlanguage acquisition. Journal of Educational Psychology 104, 974–986.

Kail, R. & L. K. Hall (2001). Distinguishing short-term memory from working memory. Memory &Cognition 29, 1–9.

Wen, Z. (2012). Working memory and second language learning. International Journal of Applied Linguistics22, 1–22.

ZHISHENG WEN is currently an Assistant Professor at Hong Kong Shue Yan University. He has lectured,researched and published in applied linguistics and psycholinguistics. His current research foci are issuessurrounding working memory and language aptitude in SLA. He is a recipient of the 2012 LanguageLearning Roundtable Conference Grant and convened the international Roundtable on ‘Memory andSLA’ in Hong Kong in June 2012. His research monograph, Working memory and second language learning,and an edited volume, Working memory in second language acquisition and processing (co-edited with MailceMota and Arthur McNeill), are both to be published by Multilingual Matters.



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

177

YEAR REFERENCES ANNOTATION THEME

1887 Jacobs, J. (1887). Experiments on‘prehension’. Mind 12.45, 75−79.

Jacobs conducted the very first systematic experimental work on the holdingcapacity of short-term memory (STM). More importantly, he devised a techniquecalled the ‘digit span’ test, in which participants were asked to recall a series ofrandom digits. He found that most people could remember only between five andnine digits from random lists presented to them. The digit span has since become astandard measure of STM.

B1

1890 James, W. (1890). Principles of psychology

(Vol. 1). London: Macmillan.In this early work, James proposed a clear distinction between ‘primary’ and‘secondary’ memory. According to James, primary memory stores a limited amountof information temporarily, while secondary memory stores knowledge more or lesspermanently.

A

1956 Miller, G. (1956). The magical numberseven, plus or minus two: Some limitson our capacity for processinginformation. Psychological Review 63,81−97.

Miller conducted many experiments in Bell Laboratories showing that a person canconsistently manage to hold only around seven items (the ‘magical number seven’)readily accessible to consciousness (i.e., immediate memory). This quantification ofshort-term memory capacity has become synonymous with the concept itself and iswidely known outside the field of psychology (cf. COWAN1 2001).

A, B

1958and1959

Brown, J. (1958). Some tests of thedecay theory of immediate memory.Quarterly Journal of Experimental Psychology

10, 12−21.Peterson, L. R. & M. J. Peterson (1959).Short-term retention of individualverbal items. Journal of Experimental

Psychology 58, 193−198.

These two papers by Brown and Peterson & Peterson demonstrated anotheraspect of the limited characteristics of short-term memory: that information can beheld in short-term memory only for a few seconds and decays gradually unlessrehearsed.

A

1960 Miller, G., E. Galanter & K. H.Pribram (1960). Plans and the structure of

behavior. New York: Holt.

The term ‘working memory’ (WM) is generally believed to have been first mentionedin this book rather than in the seminal paper by BADDELEY & HITCH (1974), thoughit was described here only very briefly and was used by the authors to refer to thetemporary retention of plans and goals in human behavior (COWAN 2011).

A



17

8R

ES

EA

RC

HT

IME

LIN

E


1966 Baddeley, A. D. (1966a). Short-term memoryfor word sequences as a function of acoustic,semantic and formal similarity. Quarterly Journal

of Experimental Psychology 18, 362−365.

Baddeley, A. D. (1966b). The influence ofacoustic and semantic similarity on long-termmemory for word sequences. Quarterly Journal of

Experimental Psychology 18, 302−309.

Results from these two studies led Baddeley to conclude that there are twoseparate storage systems: a short-term memory (STM), which mainly processesphonological information, and a long-term memory (LTM), which is morerelated to semantics. These two empirical studies provide the basis for the dualmodel of memory storage, a precursor to the seminal WM model (BADDELEY &HITCH 1974).

A

1968 Atkinson, R. C. & R. M. Shiffrin (1968).Human memory: A proposed system and itscontrol processes. In K. W. Spence & J. T.Spence (eds.), The psychology of learning and

motivation (vol. 2). New York: Academic Press,89−195.

Atkinson & Shiffrin’s influential ‘modal model’ of information processingproposed that information from multiple modes (visual, audio, spatial, etc.) firstpasses through a parallel series of sensory registers, before entering alimited-capacity STM in which information is subject to decay. In this model, theSTM is responsible not just for encoding and elaborating information but alsofor feeding it into and out of LTM, thus interpreting STM as a gateway to LTM(MIYAKE & SHAH 1999).

A

1974 Baddeley, A. D. & G. Hitch (1974). Workingmemory. In G. A. Bower (ed.), The psychology of

learning and motivation (vol. 8). New York:Academic Press, 47−90.

Baddeley & Hitch presented what is probably the most frequently cited WMtheoretical model in this timeline. It has three components: an articulatory loopfor temporary storage of sound-based data, a visual-spatial sketchpad, whichprocesses visual and spatial information, and a central controlling executive thatcoordinates the activities of the other two components. This tripartite model,simple as it may seem, has become the standard model of WM.

A

1980 Daneman, M. & P. A. Carpenter (1980).Individual differences in working memory andreading. Journal of Verbal Learning and Verbal

Behavior 19, 450−466.

Daneman & Carpenter put forward the ‘reading span task’ as a measure ofWM. In its standard format, participants are requested to read aloud a set ofsentences individually. The number of sentences in each set keeps increasing(usually from two to six). After reading each sentence, the participant also has toevaluate whether it is true or false and, at the same time, remember its finalwords. At the end of each set, the participants are required to recall all the finalwords in the order of their appearance. This type of sentence–word recall task isbelieved to tap both the storage and processing functions of WM and is known asa complex memory span task (as opposed to the simple digit span task).

B2



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

179


1989 Turner, M. L. & R. W. Engle (1989). Is workingmemory task dependent? Journal of Memory and

Language 28, 127−154.

Turner & Engle devised another version of the complex memory span task: theoperation span task. In this operation-word recall task, participants are requestedto first work out a simple mathematical operation (e.g., 2∗3+4 = 10? bird) whileretaining the final word (in this case, ‘bird’) for later recall. This type of complexmemory span task differs from the reading span task in that it taps generalaspects of WM (as opposed to language-specific aspects of the reading span task).This task has become the favored complex memory span task among WMresearchers based in North America.

B2

1991 Daneman, M. (1991). Working memory as apredictor of verbal fluency. Journal of

Psycholinguistic Research 20, 445−464.

Daneman formally introduced a spoken version of the reading span task: thespeaking span task. This task can be considered as a reverse version of thereading span task. Participants are presented with increasingly longer sets ofunrelated words that they need to read silently and keep in mind. At the end ofeach set of words, participants are required to produce aloud a meaningful andgrammatically correct sentence for each individual word in the order of theirpresentation.

B2

1992 Harrington, M. (1992). Working memorycapacity as a constraint on L2 development. InR. J. Harris (ed.), Cognitive processing in bilinguals.Amsterdam: North Holland, 123−135.

Harrington, M. & M. Sawyer (1992). L2working memory capacity and L2 reading skill.Studies in Second Language Acquisition 14, 25−38.

Harrington argued that L1 learning has access to language-specificmechanisms (e.g., universal grammar), while SLA depends more on generallearning mechanisms, so it is likely that WM plays a more important role inconstraining SLA than in L1 learning. Harrington & Sawyer presented anempirical study demonstrating a close link between WM (as measured by theparticipants’ L2 reading span task) and L2 learning (as measured by theirTOEFL subsection scores).

D2

1992 Just, M. A. & P. A. Carpenter (1992). Acapacity theory of comprehension: Individualdifferences in working memory. Psychological

Review 99, 122−149.

Just & Carpenter proposed a limited capacity theory of WM as a constrainton language comprehension. They postulated that the dual functions of storageand processing in language comprehension draw on the same pool of WMresources. The authors thus concluded that individual differences in WMcapacity can account for qualitative and quantitative differences betweencollege-age adults in several aspects of language comprehension, such asresolving ambiguities in sentence processing (cf. WATERS & CAPLAN 1996).

A2



18

0R

ES

EA

RC

HT

IME

LIN

E


1993 Gathercole, S. & A. D. Baddeley (1993).Working memory and language. Hove, UK:Lawrence Erlbaum.

In this widely-cited monograph, Gathercole & Baddeley provided a comprehensiveoverview of the major WM components presented in BADDELEY & HITCH (1974) asthey relate to various aspects of L1 learning activities (including vocabulary learning,reading comprehension, learning to read and speech production). This monographthus presents the most comprehensive and authoritative compendium ofWM–language relationships associated with the European tradition.

C

1994 Gathercole, S. E., C. S. Willis, A. D.Baddeley & H. Emslie (1994). Thechildren’s test of nonword repetition: Atest of phonological working memory.Memory 2, 103−127.

Gathercole et al. provided empirical evidence that supports the nonword repetitionspan task as a measure of the phonological component of WM (i.e., the phonologicalloop). They found positive correlations between nonword repetition span scores (in 600children aged four to nine) and their performance in vocabulary and readingcomprehension skills, thus signaling an intimate link between WM and languagelearning.

B1

1995 McLaughlin, B. (1995). Aptitude from aninformation processing perspective.Language Testing 11, 364−381.

This paper by McLaughlin was probably the first to present a theoretical argumentfor adopting WM as a central component of foreign language aptitude (WEN &SKEHAN 2011). Framed within the classical information processing model, the authorpostulated that, unlike first language acquisition (L1A), which is usually effortless, L2learning is dominated by the need for effort and controlled language processingactivities. The author therefore speculated that cognitive resources may exert greaterinfluence in SLA than in L1A.

D2

1996 Ellis, N. C. (1996). Sequencing in SLA:phonological memory, chunking andpoints of order. Studies in Second Language

Acquisition 18, 91−126.

Ellis, N. C. & S. G. Sinclair (1996).Working memory in the acquisition ofvocabulary and syntax: Putting languagein good order. The Quarterly Journal of

Experimental Psychology 49A.1, 234−250.

Ellis provided a cogent argument for interpreting the language acquisition processfrom a connectionist perspective of language acquisition, in which language acquisitionis seen largely as a process of learning an adequate amount of linguistic sequences at alllevels (ranging from lexis or words, collocations and formulae to morpho-syntaxconstructions, etc.). Most relevantly, though, the authors claimed that in the process ofacquisition of these linguistic sequences, the phonological component of WM (i.e.,PWM), given its associated mechanisms (a phonologically based short-term storage andan articulatory rehearsal process), plays a fundamental role in facilitating andconsolidating long-term representations of these chunks. Laboratory experiments byEllis & Sinclair further corroborated this proposal (cf. ELLIS 2012).

C, D



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

181


1996 Daneman, M. & P. M. Merikle (1996). Workingmemory and language comprehension: Ameta-analysis. Psychonomic Bulletin and Review 3,422−433.

In this meta-analysis of 77 studies exploring WM and reading comprehension,Daneman & Merikle provided solid evidence that complex memory spantasks that are purported to tap both the storage and processing functions of WM(such as the reading span task and the operation span task) perform better thanthe simple memory span tasks (e.g., storage-only memory tasks such as randomdigit recall and the word span task).

B2

1996 Waters, G. S. & D. Caplan (1996). Themeasurement of verbal working memory andits relation to reading comprehension. Quarterly

Journal of Experimental Psychology 49A, 51−74.

Waters & Caplan called for refinement of the widely used reading span task(DANEMAN & CARPENTER 1980). They argued that poor performance in thereading span task reflects a low ability to rehearse words rather than a limitedstorage capacity (JUFFS 2006). Through carefully designed experiments, theauthors demonstrated that the sentence–word task should also be consideredalong with a sentence processing component (a judgment accuracy measure anda reaction time score) and a recall component. This approach to assessing WM –in terms of a composite score rather than just a recall score – is also widelyadopted in SLA studies (e.g., WALTER 2004).

B2

1998 Baddeley, A. D., S. E. Gathercole & C.Papagno (1998). The phonological loop as alanguage learning device. Psychological Review

105, 158−173.

Baddeley et al. comprehensively reviewed previous empirical studies andcharacterized the phonological loop component of WM as a ‘language learningdevice’, given its important role in acquiring the new phonological forms of asecond or foreign language. The results and findings of these empirical studiesprovide ample evidence for an intimate link between the phonological loop andvocabulary acquisition and grammar development.

A

1998 Miyake, A. & N. P. Friedman (1998). Individualdifferences in second language proficiency:Working memory as language aptitude. In A. F.Healey & L. J. Bourne (eds.), Foreign language

learning: Psycholinguistic studies on training and

retention. Mahwah, NJ: Lawrence Erlbaum,339−364.

In this article, frequently cited in current SLA studies, Miyake & Friedmanproposed viewing WM as language aptitude. They pointed out that, to a largeextent, the three WM components (BADDELEY & HITCH 1974) actuallycorrespond to the three components of foreign language aptitude (SKEHAN 1998).In the empirical study, the authors were able to demonstrate through a complexstatistical technique (path analysis) the intricate relationships betweenparticipants’ L1 WM, L2 WM, and their reading comprehension. The authorsspeculated that, at least in high proficiency language learners, L1 and L2 WMmay draw on a similar pool of cognitive resources.

D2



18

2R

ES

EA

RC

HT

IME

LIN

E


1998 Skehan, P. (1998). A cognitive approach to language

learning. Oxford: Oxford University Press.In this widely cited monograph, WM and other aptitude components are alignedwith major cognitive processes (e.g., noticing, pattern identification, andlexicalization) that are embedded within key developmental stages of SLA (e.g.input, central processing, and output). Seen in this light, WM can be regarded asa central component of language aptitude that affects various aspects of SLA(WEN & SKEHAN 2011).

D

1998 Atkins, P. W. B. & A. D. Baddeley (1998).Working memory and distributed vocabularylearning. Applied Psycholinguistics 19, 537−552.

This study by Atkins & Baddeley is one of the few empirical studies conductedin Europe that directly dealt with the relationship between phonological WMand L2 vocabulary learning. The results suggested that participants’ phonologicalmemory span is correlated with their rate of L2 vocabulary learning, thuscorroborating the intimate link between WM and L2 vocabulary learning.

D1

1999 Caplan, D. & G. S. Waters (1999). Verbalworking memory and sentence comprehension.Behavioral and Brain Sciences 22, 77−126.

Caplan & Waters presented evidence that the WM system contains differentspecialized modules involved in sentence comprehension at two different points.First, there is the on-line sentence processing stage (the ‘interpretive processes’),which involves assigning syntactic structure. Then there is the off-line sentenceprocessing stage (the ‘post-interpretive processes’), which is involved in thedetermination of the meaning of the discourse. The authors further argued thatconventional WM span tasks (such as the reading span task) only measure theinterpretive processes but not the post-interpretive processes.

C2

1999 Miyake, A. & P. Shah (eds.) (1999). Models of

working memory: Mechanisms of active maintenance

and executive control. New York: CambridgeUniversity Press.

This volume edited by Miyake & Shah offers a comprehensive overview ofmajor theoretical models of WM in cognitive psychology (representing mostimportant WM research groups in Europe and North America). All thecontributing authors provided their perspectives on WM by answering eightspecified questions about the nature, structure, and functions of WM. Mostsignificantly, despite apparent differences in research focus, these WM modelshave much common ground, thus allowing the authors to arrive at some ‘unifiedtheories of WM’ that are particularly relevant to language research (WEN &SKEHAN 2011). First, WM comprises multiple components/mechanisms; second,WM is limited in capacity (in terms of time and span); and third, WM is somesort of gateway to long-term memory (LTM).

A



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

183


2000 Baddeley, A. D. (2000).The episodic buffer: Anew component of working memory? Trends in

Cognitive Sciences 4, 417−423.

In this updated model of WM, Baddeley added a fourth component, the‘episodic buffer’, to the original tripartite model (c.f. BADDELEY & HITCH 1974).The episodic buffer functions as a passive storage component of the centralexecutive and processes information from LTM and other WM components.With the episodic buffer separated out from the central executive, Baddeley’sWM model shows similarities with Cowan’s WM framework (COWAN 2011).

A

2001 Cowan, N. (2001). The magical number 4 inshort-term memory: A reconsideration ofmental storage capacity. Behavioral and Brain

Sciences 24, 97−185.

Cowan presented a cogent argument that only around three to five (i.e. 4±1)chunks of information are stored in STM, rather than the original estimate ofseven by MILLER (1956). According to the author, this is the (reduced) number ofunits of information that are activated from long-term memory: in other words,they constitute the ‘focus of attention’.

A

2001 Whitney, P., P. Arnett, A. Driver & D. Budd(2001). Measuring central executivefunctioning: What’s in a reading span? Brain &

Cognition 45, 1−14.

Whitney et al. administered the reading span task (DANEMAN & CARPENTER

1980) to 80 undergraduates with a view to determining the contributions madeby specific mental operations. Hierarchical regression results suggest that, inaddition to storage and processing speed, two other factors – manipulationcapacity and susceptibility to interference – also contribute to this complexmemory span task. The authors call for more studies into the nature of thereading span task as a viable measure of the central executive component of WM.

B2

2002 Engle, R. W. (2002). Working memory capacityas executive attention. Current Directions in

Psychological Science 11, 19−23.

Engle presented a comprehensive evaluation of the theoretical stance of theWM construct adopted by North America-based WM research groups. In thisconception, WM is synonymous with executive control or executive attentionand is domain-general in nature (not specific to, for example, language). In linewith this view of WM, the author calls for the adoption of a domain-general spantask (such as the operation span task) for measuring WM.

A, B

2003 Baddeley, A. D. (2003). Working memory andlanguage: An overview. Journal of Communication

Disorders 36,189−208.

In this widely-cited review paper, Baddeley provided a detailed account of eachof the four components of WM as conceived in his domain-specific model(BADDELEY & HITCH 1974; BADDELEY 2000), and their implications for nativeand second language acquisition, as well as for normal language processing andlanguage disorders.

C, D



18

4R

ES

EA

RC

HT

IME

LIN

E


2003 Chein, J. M., S. M. Ravizza & J. A. Fiez (2003).Using neuroimaging to evaluate models ofworking memory and their implications forlanguage processing. Journal of Neurolinguistics 16,315−339.

Chein et al. provided neuroimaging evidence that serves to compare andcontrast Baddeley’s multiple-component WM model and Cowan’sembedded-processes model. They concluded that Cowan’s model offersadditional insights into the contributions of Broca’s area and the left inferiorparietal cortex to WM that can account for a wide range of findings previouslymissing from Baddeley’s standard model (BADDELEY & HITCH 1974).

A, C

2004 Walter, C. (2004). Transfer of readingcomprehension skills to L2 is linked to mentalrepresentations of text and to L2 workingmemory. Applied Linguistics 25, 315−319.

Following the suggestions of WATERS & CAPLAN (1996), Walter administered thereading span task (DANEMAN & CARPENTER 1980) to two groups of Frenchlearners of English at two proficiency levels (upper-intermediate andlower-intermediate) and correlated their composite WM scores (logicalityjudgment accuracy, reaction time, and recall score) with their comprehensionskills. The results seemed to indicate that upper-intermediate learners succeededin moving on from L1 to L2 mental representations while theirlower-intermediate counterparts failed to do so. Most relevantly, the study foundthat such an advancement of comprehension abilities from L1 to L2 ismodulated by participants’ WM.

D2

2005 Conway, A., M. Kane, M. Bunting, D.Hambrick, O. Wilhelm & R. Engel (2005).Working memory span tasks: A methodologicalreview and user’s guide. Psychonomic Bulletin and

Review 12, 769−786.

Conway et al. offered a comprehensive and authoritative evaluation of themost widely used WM measures (including counting/number span, operationspan, and reading span tasks) and provided practical suggestions for how tomeasure WM reliably and validly (e.g., optimal administration and scoringprocedures). The paper also provided practical suggestions for statistical andmethodological techniques for use with WM span tasks (e.g., in terms of latentvariable analysis and extreme-groups design).

B

2006 Gathercole, S. (2006). Nonword repetition andword learning: The nature of the relationship.Applied Psycholinguistics 27, 513−543.

In this keynote article from a special issue, Gathercole presented a theoreticalaccount of how the cognitive processes underlying the nonword repetition spantask are aligned with those involved in the acquisition of the phonological formsof new words. Based on this analysis, the author proposed an intimate linkbetween phonological WM and vocabulary learning, also arguing that thisphonological memory ability is not only particularly important at the early stagesof language acquisition but is also needed for word learning throughout life.

A1, C1



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

185


2006 Juffs, A. (2006). Working memory, secondlanguage acquisition and low-educated secondlanguage and literacy learners. LOT Occasional

Papers: Netherlands Graduate School of Linguistics,89−104.

Juffs articulated the theoretical and measurement difficulties involved in testingWM capacity in low-educated L2 and literacy acquisition populations. Fortesting this group, the author proposed a simple memory span task (e.g., a wordspan or a digit span task) to be administered first in their L1, augmented with anonword repetition span task (GASS & LEE 2011).

D

2006 O’Brien, I., N. Segalowitz, J. Collentine & B.Freed (2006). Phonological memory andlexical, narrative, and grammatical skills insecond language oral production by adultlearners. Applied Psycholinguistics 27, 377–402.

O’Brien, I., N. Segalowitz, J. Collentine & B.Freed (2007). Phonological memory predicts L2oral fluency gains in adults. Studies in Second

Language Acquisition 29, 557−582.

These two longitudinal studies reported by O’Brien et al. explored the role ofPWM (indexed by the nonword recognition span task) in L2 speech development(including lexical, narrative, and grammatical performance) in a group of lowproficiency English speakers learning L2 Spanish. The results of regressionanalyses pointed to the significant but distinct role of PWM at differentdevelopmental stages of L2 proficiency (e.g., narrative development at earlierstages vs. grammatical competence at later stages).

D1

2007 Conway, A. R. A., C. Jarrold, M. J. Kane, A.Miyake & J. N. Towse (eds.) (2007). Variation in

working memory. New York: Oxford UniversityPress.

Conway et al. presented the second (following MIYAKE & SHAH 1999)comprehensive and authoritative collection of WM theories, with a particularfocus on identifying the sources of individual differences in WM resources. Thisvolume is another important contribution to the conception of WM espoused bythe North America-based research groups. Overall, the book provides greatinspiration for specification of the executive component of WM (EWM). Inparticular, the introductory chapter by Kane, Conway, Hambrick & Engle iswidely cited by L1 and L2 researchers.

A

2008 Dehn, M. J. (2008). Working memory and academic

learning: Assessment and intervention. Hoboken, NJ:Wiley.

Feeling that previous WM models in cognitive psychology (such as those listed inMIYAKE & SHAH 1999) often fail to describe how WM encodes information intoLTM and thus do not offer an account of the actual work conducted by WM,Dehn proposed an integrated model of WM that contains STM componentsand functions, WM components and operations, and activated items in LTM. Inparticular, the author distinguished between conscious WM functions and thesubconscious operations that WM also performs on its content. The author alsocalled for adopting multiple methods and memory scales to assess WM.

A, B



18

6R

ES

EA

RC

HT

IME

LIN

E


2008 French, L. M. & I. O’Brien (2008).Phonological memory and children’s secondlanguage grammar learning. Applied

Psycholinguistics 29, 463−487.

This longitudinal study by French & O’Brien examined the role ofphonological WM in L2 grammar learning in a group of native French-speakingchildren in a five-month intensive English program. PWM was measured bynon-word repetition span tasks in participants’ L2 (English) and L3 (Arabic). Theresults of the study indicated that PWM significantly predicted grammardevelopment (even when vocabulary knowledge was partialed out). The authorsfound that although participants’ PWM in L2 increased between Time 1 andTime 2 (tested in the first and last months of the program), it did not increase inL3. Given this, it may be problematic to adopt a nonword repetition span taskconstructed on the basis of participants’ L2 in WM-SLA studies (GASS & LEE

2011).

D1

2009 Sunderman G. & J. F. Kroll (2009). Whenstudy-abroad experience fails to deliver: Theinternal resources threshold effect. Applied

Psycholinguistics 30, 79−99.

In this empirical study, Sunderman & Kroll sought to identify the minimumlevel of WM capacity (e.g., as measured by the reading span task) required for L2learners to benefit optimally from their study-abroad experience. The dependentvariables were lexical comprehension and production, as measured by atranslation recognition task and a picture-naming task. The results corroboratethe hypothesis that there exists a certain threshold of WM resources (referred toas ‘threshold effects’) for language learners to take full advantage of their stay.

D2

2011 Cowan, N. (2011). Working memory andattention in language use. In J. Guandouzi, F.Loncke & M. J. Williams (eds.), The handbook of

psycholinguistics and cognitive processes. London:Psychology Press, 75−97.

The relationship between WM and attention in language research has been atricky one. In this chapter, Cowan focused on two issues: first, whether there isone central WM function that uses attention and, second, whether such attentionresources cut across different domains (such as verbal and nonverbal processing);in other words, whether there are specific WM modules for different types ofmaterial. The author suggested that WM for language draws on bothattention-dependent and automatic activation processes (cf. DEHN 2008).

C2



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

187


2011 Gass, S. & J. Lee (2011). Working memorycapacity, inhibitory control, and proficiency ina second language. In M. Schmid & W. Lowie(eds.), From structure to chaos: Twenty years of

modeling bilingualism: In honor of Kees de Bot.Amsterdam: John Benjamins, 59−84.

This paper by Gass & Lee explored the relationship between WM capacity(indexed by L1 & L2 reading span tasks) and inhibition (measured by L1 and L2color-word interference tasks) and further explored how such a relation ismediated as a function of participants’ L2 proficiency. Participants were nativeEnglish speakers recruited from first and third year Spanish classes. Usingstatistical analyses, the authors proposed a theoretical model depicting theinterrelatedness between the three factors (WM, inhibitory control, and L2proficiency). A central claim of the model is that an individual’s L2 proficiency isdecisive for his/her L2 WM performance and the ability to control inhibition.

D2

2011 Juffs, A. & M. Harrington (2011). Aspects ofworking memory in L2 learning. Language

Teaching 44, 137−166.

Juffs & Harrington offered a comprehensive review of theoretical models andissues related to the assessment of WM and their implications for differentaspects of SLA. They explored the relationships between WM and L2 sentenceprocessing, reading, speaking, lexical development, and general proficiency. Theyconcluded that WM is not a unitary concept and is likely to be sensitive to L2learners’ age, the language task under investigation, and the linguistic domain inquestion.

D

2011 Wen, Z. & P. Skehan (2011). A new perspectiveon foreign language aptitude: Building andsupporting a case for ‘working memory aslanguage aptitude’. Ilha Do Desterro: A Journal of

English language, literatures and cultural studies 60,15−44.

This paper proposed that, given the critical role of WM in different aspects ofSLA (including vocabulary, grammar, reading comprehension, speech, andwritten production), it is time to include it as a central component of foreignlanguage aptitude (cf. MCLAUGHLIN 1995; SKEHAN 1998). In particular, thepaper enumerated three criteria for doing this, and gave an account of how thesecriteria had been met so far.

D

2012 Ellis, N. C. (2012). Formulaic language andsecond language acquisition. Annual Review of

Applied Linguistics 32, 17−44.

Ellis, in this review article, critically evaluated his earlier proposal (ELLIS 1996)that language learning is largely the learning of a whole range of formulaicsequences. In particular, the author reviewed studies of individual differences inPWM and EWM and outlined their distinct roles in the acquisition andprocessing of lexis, formulae, and morpho-syntactic constructions.

C, D



18

8R

ES

EA

RC

HT

IME

LIN

E


2012 Mackey, A. (2012). Input, interaction and corrective

feedback in L2 learning. Oxford: OxfordUniversity Press.

In this book-length monograph, Mackey devoted a whole chapter to a review ofthe role of WM in interaction-driven SLA studies. Previous empirical studiesconducted by the author and colleagues seemed to suggest a positive correlationbetween L2 learners’ WM capacity (as indexed by a composite WM Z-scoreaveraged from a simple memory span and a complex span task) and thesuccessful noticing of corrective feedback provided to them by interlocutors. Inother words, L2 learners with a high WM span are likely to benefit more fromsuch interaction activities than their lower WM counterparts.

D2

2012 Martin, K. I. & N. C. Ellis (2012). The roles ofphonological STM and working memory in L2grammar and vocabulary learning. Studies in

Second Language Acquisition 34.3, 379−413.

In this empirical study, Martin & Ellis explored the effects of PWM and EWMon vocabulary and grammar learning in an artificial language. The resultssuggested that PWM and EWM make separate contributions to vocabulary andgrammar learning. The study also found that such effects of PWM and EWMcan be either directly constraining on language learning or indirectly mediatingthese learning processes.

C, D

2012 Miyake, A. & N. P. Friedman (2012). Thenature and organization of individualdifferences in executive functions: Four generalconclusions. Current Directions in Psychological

Science 21.1, 8−14.

This paper by Miyake & Friedman provides a comprehensive andauthoritative account of the cognitive and biological underpinnings of three keyexecutive functions related to WM (information updating, shifting, andinhibition) in cognition, as well as their theoretical ramifications for theory,research, measurement, and practice.

B2

2012 Williams, J. N. (2012). Working memory andSLA. In S. Gass & A. Mackey (eds.), Handbook of

second language acquisition. Oxford:Routledge/Taylor Francis, 427−441.

Williams gives an overview of WM research, including WM models andassessment procedures as they relate to the SLA process. In particular, anin-depth discussion is provided on the application of WM in such cognitiveprocesses as reasoning, category learning, and artificial learning in SLA.Furthermore, the author also highlights two approaches to applying WM trainingin SLA: either enhancing the cognitive abilities themselves or modifying theinstruction to relieve the WM burden during language learning and processing.

D



ZH

ISH

EN

GW

EN

:T

HE

OR

IZIN

GA

ND

ME

AS

UR

ING

WO

RK

ING

ME

MO

RY

189


2013 Coughlin, C. E. & A. Tremblay (2013).Proficiency and working memory basedexplanations for nonnative speakers’ sensitivityto agreement in sentence processing. Applied

Psycholinguistics 34.3, 615–646.

This empirical study by Coughlin & Tremblay examined the roles of WMcapacity (measured in participants’ L2 reading span task) and L2 proficiency(medium- and high-level) in L2 learners’ processing of agreement morphology,i.e., the extent to which they are sensitive to violations of subject–verb agreementrules. The results suggested that L2 learners with higher WM span tend to bemore sensitive to agreement violations and that the performance of highproficiency L2 learners is comparable to that of native speakers.

D2

2013 Kaushanskaya, M. & J. Yoo (2013).Phonological short-term and working memoryin bilinguals’ native and second language.Applied Psycholinguistics 34.5, 1005–1037.

This study provided empirical evidence of the relationship between PWM(measured by the nonword repetition span task) and EWM (measured by thecomplex memory span task) as they relate to L1 and L2 learning among Koreanspeakers learning L2 English. The results point to a stronger cross-linguisticassociation for the EWM span task than for the PWM task. In other words,PWM is specific to language, while EWM is related to more general learning.

C, D

2013 Szmalec, A., M. Brysbaert & W. Duyck (2013).Working memory and (second) languageprocessing. In J. Altarriba & L. Isurin (eds.),Memory, language, and bilingualism: Theoretical and

applied approaches. Cambridge: CambridgeUniversity Press, 74−94.

In this chapter, the authors argued that language acquisition in L1 and L2 issupported by extraction of information from WM in the right sequence, whilelanguage processing (e.g., perception and production) tends to rely more oncontrol of attention. The authors also postulated that the EWM functions are notspecific to language processing. They concluded that, although L2 learners’ WMcapacity as measured by the reading span tasks may not increase, continuous useof such general EWM functions should serve to enhance overall mental flexibility.

D



19

0R

ES

EA

RC

HT

IME

LIN

E


f/c Wen, Z. (forthcoming). Working memory and second

language learning: An integrated framework. Bristol,UK: Multilingual Matters.

Emulating the seminal work of GATHERCOLE & BADDELEY (1993) on WM and L1learning, Wen provides a comprehensive account of the role of WM(conceptualized as consisting of PWM and EWM) in SLA. The book’s majorclaim is that PWM and EWM are the two most relevant components of WM tothe SLA process and that their distinctive effects should be treated separately inspecific SLA domains and cognitive processes (WEN 2012). The authorspeculates that the central feature of this integrated framework for WM in SLA isthat PWM underpins the acquisition and development aspects of vocabulary andgrammar learning, while performance-based aspects of SLA are mainly mediatedby EWM. Incorporated within this Phonological/Executive Hypothesis are alsogeneral guidelines for implementing WM assessment procedures (simple memoryspan task vs. complex memory span task) in tandem with target participants’ age(in L1) and proficiency levels (in L2).

A, B,C, D

1 Authors’ names are shown in small capitals when the study referred to appears in this timeline.


Wen, Zhisheng. (2014).Theorizing and measuring working memory in first and second language research. Language Teaching, 47/2, 173-190.

Documents