Dynamics of first language attrition across the lifespan

Abstract

This special issue explores the relevance of Dynamic Systems Theory (henceforth DST) to research in first language attrition (henceforth L1 attrition). While research in L1 attrition has traditionally drawn on paradigms in other areas of linguistic enquiry, e.g. approaches current in L2 acquisition studies, important developments across a range of disciplines have paved the way for a fundamentally different conceptualisation of language development. In this special issue, DST guides our enquiry into L1 attrition, which we view as one specific dynamic process of change in language development.

DST was first linked to L1 attrition by Herdina and Jessner (2002), and then further developed by de Bot, who suggested that DST provides ‘a fresh and promising perspective on language attrition as a normal part of language development and not as a lamentable loss of what once was’ (2004, p. 233). De Bot argued that a DST-based model of language development should be applicable to language attrition and that the theoretical underpinning of L1 attrition research would be strengthened by adopting this perspective (see also Ecke, 2004). However, de Bot also pointed to methodological challenges implicit in applying a DST perspective to L1 attrition. Here, one of the most salient challenges is actually testing DST, a feat yet to be accomplished. Nevertheless, eight years on, the purpose of this special issue is to present for the first time empirical research which has adopted the DST framework to study L1 attrition across the lifespan.

In this introduction, two propositions related to the relevance of DST in L1 attrition research are discussed.

L1 attrition as such can be interpreted as evidence against the view that once an L1 is acquired ‘in full’ (i.e. throughout childhood and up to the onset of puberty), it displays invariance which is not evidenced in a second language (henceforth L2, i.e. a language acquired post-puberty). Instead, it is proposed here that L1 attrition /as such/ is evidence for the view that the L1 develops dynamically across the lifespan.

L1 attrition itself is dynamic. As it unfolds, often in the context of a life-changing event such as migration to a new country and attendant L2 acquisition, it does so dynamically in a process of constant change and in interaction with the developing L2 and environment.

We define and contextualise key concepts (L1 attrition and DST) in order to relate them to our two propositions. We further discuss important features of DST in relation to the contributions of this special issue, as well as some methodological challenges relevant for studying the dynamics of L1 attrition across the lifespan.

L1 attrition

We delimit L1 attrition to non-pathological, non-age related changes in the L1 of a late bilingual, which often entail reduced proficiency and loss of structural features previously acquired (see Köpke & Schmid, 2004; Schmid, 2011). While L1 attrition may potentially occur for reasons not involving another language, e.g. the ‘desert island situation’ (Sharwood Smith & van Buren (1991, p. 22), it usually unfolds alongside the development of an L2, for example after relocation to a new country where the ambient language is acquired. Such relocation may involve either being entirely cut-off from the L1, as often happens with adopted children (Isurin, 2000), or at least a reduction in contact with the L1.

Moreover, L1 attrition contrasts with language change processes at the group or societal level, such as intergenerational language shift, although it may be a precursor to, and share certain similarities with, such macro processes. Accordingly, a conceptual and terminological distinction between micro-level L1 attrition and macro-level language shift under the cover term of language loss has been drawn (de Bot, 1998).

Lowered proficiency in the L1, or linguistic behaviour that differs from ‘typical speakers’ (Cook, 2003, p. 2) of the L1, are generally considered to be symptomatic of L1 attrition. However, the conception of L1 attrition as loss (i.e. in contrast to ‘change’) has been questioned. For example, Cook (2003) argues for a multi-competence perspective when discussing attrition in bilinguals, which includes acknowledging and valuing bilinguals’ proficiency in more languages than one, as well as the systematic differences between bi- and monolinguals in their L1 as a result of the formers’ bilingualism (Kaufman, 1991). This approach is congruent with a DST perspective as outlined below.

Although L1 attrition may be encountered at any stage across the lifespan (see below), a crucial question for L1 attrition research according to the definition offered here is whether it is possible for a mature L1 to change upon acquisition of an L2 (but whether the change in the L1 is caused by L2 acquisition is indeed another question). Within certain theoretical perspectives, such as the generative paradigm and approaches based on the critical period hypothesis, L1 attrition is something that is not expected (e.g. Montrul, 2008; Seliger, 1996). However, findings in line with those presented in this special issue indicate that a mature L1 is indeed susceptible to L1 attrition (see e.g. de Leeuw, 2009; de Leeuw, Schmid & Mennen, 2010; Jarvis, 2003; Lubińska, 2011; Opitz, 2010; Opitz, 2011; Tsimpli, Sorace, Heycock, Filiaci & Bouba, 2003). For example, in a particularly revealing case, L1 speakers who underwent attrition were perceived to be non-native speakers of their L1 (de Leeuw, et al., 2010). Such results suggest that the categorical distinction that is often drawn between the L1 and the L2 is questionable. Moreover, concerning the relative vulnerability of different linguistic sub-systems, certain parts of the lexicon have been found to be more vulnerable than certain parts of morpho-syntax (Köpke & Nespoulous, 2001); or, in Universal Grammar (UG) terms, syntax-related interfaces are vulnerable, while narrow syntax is impervious toL1 attrition (Montrul, 2004; Sorace, 2000; see also Gürel, 2008).

If it is accepted that a fully acquired L1 can undergo L1 attrition, as defined here, the next obvious question is whether the changes in the L1 (i.e. L1 attrition) affect only the processing of language (or, in UG terms, performance), leaving the underlying knowledge structures (UG: competence) intact, or whether the knowledge itself is affected. With the methodologies currently available, this question cannot be answered unanimously, whichever way it is framed, since knowledge structures are not amenable to direct observation, and accessing them involves processing. While generative approaches tend to argue the former view (knowledge remains unchanged; only processing is affected), psycholinguistic and dynamic approaches do not preclude the latter (the knowledge structures themselves may change, or indeed be ‘lost’). DST theorists might also argue that language structure and language use are inextricably linked, contra generative assertions of unassailable L1 competence (e.g. Larsen-Freeman & Cameron, 2008).

One way of approaching the problem of how to distinguish knowledge from processing within L1 attrition research (which aides us in our definition of L1 attrition, and therefore also in determining whether L1 attrition is indeed possible) is research related to the savings paradigm (Ebbinghaus, 1885/1913). Within the context of L1 attrition, the objective here is to examine the impact of training on L1 proficiency by measuring the speed with which bilinguals relearn supposedly lost aspects of their L1, and then comparing this speed to the learning speed of individuals who encounter these aspects for the first time (de Bot & Stoessel, 2000). As discussed by Esther de Leeuw, Ineke Mennen and James M. Scobbie in this issue, if the ‘relearners’ master the material more quickly than the ‘learners’, it may be deduced that the former were able to build on knowledge that had become inaccessible, but was in fact there. However, the actual nature of linguistic knowledge is likely to continue to defy description, not just because it cannot be accessed directly, but also because the probability of finding two groups, one of ‘relearners’ and another of first-time ‘learners’, with exactly the same proficiency level is very low.

Whether findings revealing L1 attrition reflect the changes to the underlying knowledge or the processing of that knowledge (or both), research into mature speakers’ L1 development has shown that L1 proficiency is susceptible to systematic change post-puberty alongside the growth of an L2. Findings revealing L1 attrition, as well as continual or renewed growth, extend the window for L1 development further down the lifespan into adulthood, prompting the inclusion of a lifespan perspective in this special issue. Indeed, Schmid (2009) argues that

[o]f all situations of bilingual development, first language attrition is arguably the one for which a lifespan perspective is most crucial, since it is conditioned by at least one major life event (most commonly emigration) and furthermore is a slow process where changes typically are witnessed only after years or decades. (Schmid, 2009, p. 171)

The contributions in this special issue address the need for a lifespan perspective on L1 attrition occurring post-puberty. We contend that findings indicating L1 attrition can be interpreted as evidence against the view that a fully acquired L1 reaches a static stage post-puberty. Instead we propose that the L1 develops dynamically across the lifespan and that L1 attrition as such is evidence for such development. Moreover, we propose that L1 attrition itself is a dynamic process, as discussed in the following section.

Dynamic Systems Theory

DST can be viewed as an analytical model which serves to describe the behaviour of physical phenomena over time (e.g. Thagard, 2005). Such phenomena are hypothesised to display properties such as non-linearity, complexity, and unpredictability in their development, while DST may also help to explain the state of physical phenomena at a specific moment in time. For example:

[…] meteorologists make their predictions and explanations by considering a large number of variables, including temperature, humidity, and air pressure at various locations. They incorporate these variables into mathematical equations that describe how the weather system changes over time. Meteorologists treat weather as a dynamic system, that is, as a system whose changes over time can be characterized by a set of equations that show how current values of variables depend mathematically on previous values of those variables. (Thagard, 2005, p. 198)

Within this paradigm, language development is considered to be comparable to physical phenomena, such as the weather. More specifically in relation to language, DST was initially adopted in contrast to the so-called ‘computational approach’ in which cognition is likened to the workings of a computer. This arguably reductionist approach assumes that ‘representations are static structures of discrete symbols’ (Thagard, 2005, p. 1). Within linguistics, theories have emerged which share characteristics with the computational approach by viewing language, particularly the L1, as static post-puberty. Likewise, generativist approaches emphasise innate language acquisition abilities with binary features and fixed, predetermined grammars (‘Five Graces Group’ et al., 2009, p. 14). Such perspectives are challenged by results indicating L1 attrition.

However, L1 attrition research itself has to some extent adopted a static view in its methodological approach. The guest contribution by Monika Schmid, Barbara Köpke and Kees de Bot sketches shortcomings of current research into L1 attrition conducted from a static systems framework. According to the authors, such research not only neglects the potential for language to develop over time, it also relies on falsely aggregating seemingly similar bilinguals into groups in order to conduct group-level statistical analyses at a single point in time. The guest contributors suggest that studying L1 attrition in isolation neglects the complexities of individuals’ overall development in both or all of their languages (e.g. see also Conny Opitz’s contribution). Their suggestions are in line with other calls to no longer interpret non-linear, emergent behaviour as ‘noise’ in the system (Herdina & Jessner, 2002, p.78). Instead, according to their proposal and in line with the propositions here, L1 attrition itself should be considered dynamic. As it unfolds across the lifespan, it does so dynamically in a process of constant change in interaction with other elements such as the developing L2 and the surrounding environment.

Schmid, Köpke and de Bot also suggest a number of ways in which to address dynamic change in L1 attrition research methodologically. In relation to their call for longitudinal micro-investigations, it is particularly worth bearing in mind that a methodological problem which arises in longitudinal investigations is that the repeated testing of variables within a bilingual migrant group may disturb the ‘natural course of the process it [the test] hoped to track down’ (Jaspaert & Kroon, 1989, p. 81). A solution to this challenge is to conduct post hoc longitudinal studies. Unfortunately, recordings which would enable a post hoc analysis are next to non-existent, i.e. it is generally not the case that migrants unknowingly record their conversations after migration, nor that their conversations are unknowingly recorded by researchers. Future longitudinal studies may, however, examine migrant communities in such a way that the participants are unaware of the test.

Within DST the development of the L1 is seen as emergent from ongoing human social interactions across the lifespan. However, L1 development as a dynamic system is nonetheless thought to be highly, or sensitively (de Bot, 2007; de Bot, Lowie, & Verspoor, 2007), dependent on its initial state (i.e. the L1, as a dynamic system, is dependent on its initial state, but may nonetheless undergo L1 attrition). In relation to this, the remaining contributions in this special issue examine interpersonal and intrapersonal variation in L1 attrition. To start, Esther de Leeuw, Ineke Mennen and James M. Scobbie test the fruitfulness of a maturational constraints framework in comparison with that of a dynamic systems framework in explaining interpersonal and intrapersonal variation in L1 attrition. They argue that the former theory conceptualises the L1 more statically than DST, whilst the latter incorporates dependency on an initial state, as well as social and linguistic factors. The authors conclude on the basis of their results that DST ultimately offers more as a theoretical contribution to L1 development as it is more likely to explain the evidenced high degree of variability in L1 attrition.

Dynamic systems are usually considered to be complex, in that they consist of numerous diverse (nested subsystems of) elements which are interconnected and distinct (de Bot, 2007; de Bot, et al., 2007). A consequence of the interconnectedness of the various elements of a dynamic system is that change at one level, or in one subsystem, will lead to changes in another. The full interconnectedness of systems means that isolating single variables as explanatory may be unproductive because the impact of one element within the system or subsystems on other elements and (sub-)systems runs the risk of being ignored (de Bot, 2007). The interactions between the system’s elements are thought to self-organise collectively and change over time, although not necessarily linearly (de Bot, 2007; Larsen-Freeman & Cameron, 2008; Smith & Thelen, 2003). Furthermore, dynamic systems are considered to interact with the environment, or be open (Larsen-Freeman & Cameron, 2008).

In line with these properties, Conny Opitz shows that the effects of acquiring an L2 and maintaining an L1 (i.e. subsystems nested within a larger language system) may lead to the attrition of other languages within a multilingual’s repertoire. She emphasises that L1 attrition is a feature of multilingualism and that it should be studied as part of an interconnected system. Her qualitative data, in particular the two case studies, also provide empirical support for the Dynamic Model of Multilingualism (DMM, Herdina & Jessner, 2002).

Mirela Cherciov’s research of this special issue is based on the expectation that L1 attrition will be more severe among those late bilinguals who have a negative attitude towards their L1, i.e. those who do not identify with the L1 country any more, and as a result have a tendency to use the L1 of that environment less frequently. On the other hand, a positive attitude towards the L1 and its associated environment is expected to promote maintenance of the L1. Furthermore, she argues that some methodologies are more contingent than others at tapping into this relationship between L1 attrition and attitude towards the environment. Her findings suggest that face-to-face interviews, rather than questionnaires, are more likely to capture the impact of a person’s attitude towards their native language, and hence the likelihood of L1 attrition. Her qualitative methodological approach is consistent with DST in that it calls attention to complex developments which would potentially be sidelined in a quantitative study examining attitude.

It is fortuitous that we are able to include a longitudinal case study in this special issue. Peter Ecke and Christopher J. Hall investigate language attrition in a multilingual speaker over the course of 10 years, arguing that temporary changes in the environment affect the non-linearity of language attrition. As evidenced in their study, changes in the environment may accumulate to cause qualitative change over time with the next developmental change depending on the previous one. Indeed, they found that the speaker’s L1 (German) gained stability after an early period suggestive of L1 attrition, while the late learned, but frequently used, L4 (Spanish) displayed evidence of increasing instability in the lexical system. Similarly, it has been suggested that the first years after migration have the greatest influence on the extent to which an individual undergoes L1 attrition (de Bot & Clyne, 1994).

In their epilogue, Philip Herdina and Ulrike Jessner discuss the importance of L1 attrition research such as it is presented in this special issue for expanding our understanding of the development of bi- and multilingual language systems. Research on attrition as part of bi- and multilingual development gives rise to new research questions, and these, they argue, are best studied from a DST perspective. Thus, DST offers important insights for research on multilingualism in general, and for L1 attrition research in particular, while promoting the development of new research methodology. The contributions in this volume are a first step towards the empirical confirmation of models of multilingualism adopting a DST perspective such as DMM (Herdina & Jessner, 2002).

Taken as a whole, we interpret the findings from the contributions presented in this special issue as support for the proposition that L1 attrition is evidence for the capacity of an L1 to develop dynamically across the lifespan. Moreover, the findings support the proposition that L1 attrition itself is dynamic, unfolding in a process of constant change and in interaction with the developing L2 and environment. Given the great potential of adopting a dynamic perspective to L1 attrition, and research in bi- and multilingualism more generally, we hope that the contributions collected in this special issue are the first of many to study L1 attrition from this perspective.

Footnotes

Acknowledgements

We thank all of the authors of the contributions to this special issue for their patience and reliability. We also thank Monika Schmid for her original suggestion to organise a special issue on L1 attrition, and Li Wei for his support in taking the venture further. Finally we express our thanks to the colleagues who acted as anonymous reviewers for the present issue, and whose criticism has improved the quality of this collection.

Author biographies

Esther de Leeuw is a Lecturer in Linguistics at Queen Mary,University of London. Her research most often employs phonetic techniques to investigate language variation and change, especially in the speech of late bilinguals. She studied in Germany and the Netherlands before obtaining her PhD in Speech and Hearing Sciences from Queen Margaret University, Edinburgh in 2009.

Conny Opitz teaches Russian as a foreign language and linguistics at Trinity College Dublin. She was awarded her PhD in Applied Linguistics by Trinity College Dublin in 2011. Taking a dynamic, multi-competence view of bilingualism, her PhD research was an extensive study of the parallel processes of L1 attrition and L2 acquisition in German adult bilinguals in Ireland and focused on proficiency data vis-à-vis a range of cognitive and sociolinguistic predictors. Her research interests include the application of dynamic systems theory to multilingual development and use.

Dorota Lubińska teaches Swedish as a foreign language at KTH Royal Institute of Technology, Sweden. In 2011 she obtained her PhD in Bilingualism from Stockholm University, Sweden. Her thesis was concerned with the adult L1 attrition of highly-educated long-term Polish immigrants in Sweden. In January 2014 she will take up a position as a Senior Lecturer in Swedish as a Second Language at Stockholm University.

References

Cook

(2003). Introduction: The changing L1 in the L2 user’s mind. In Cook

(Ed.), Effects of the second language on the first (pp. 1–18). Clevedon, UK: Multilingual Matters.

De Bot

(1998). The psycholinguistics of language loss. In Extra

Verhoeven

(Eds.), Bilingualism and migration (pp. 345–361). Berlin, Germany: Mouton de Gruyter.

De Bot

(2004). Introduction: Special issue on language attrition. International Journal of Bilingualism, 8(3), 233–237.

De Bot

(2007). Dynamic Systems Theory, lifespan development and language attrition. In Köpke

Schmid

M. S.

Keijzer

Dostert

(Eds.), Language attrition. Theoretical perspectives (Vol. 33,pp. 53–68). Amsterdam, NL: John Benjamins.

De Bot

Clyne

(1994). A 16-year longitudinal study of language attrition in Dutch immigrants in Australia. Journal of Multilingual and Multicultural Development, 15(1), 17–28.

De Bot

Lowie

Verspoor

(2007). A Dynamic Systems Theory approach to second language acquisition. Bilingualism: Language and Cognition, 10(01), 7–21.

De Bot

Stoessel

(2000). In search of yesterday’s words: Reactivating a long-forgotten language. Applied Linguistics, 21(3), 333–353.

De Leeuw

(2009). When your native language sounds foreign: A phonetic investigation into first language attrition. Unpublished PhD thesis, Queen Margaret University, Edinburgh, UK.

De Leeuw

Schmid

Mennen

(2010). The effects of contact on native language pronunciation in an L2 migrant context. Bilingualism: Language and Cognition, 13, 33–40.

10.

Ebbinghaus

(1885/1913). Memory: A contribution to experimental psychology ( Ruger

H. A.

Bussenius

C. E.

, Trans.). New York, USA: Teachers College, Columbia University.

11.

Ecke

(2004). Language attrition and forgetting: A cross-disciplinary review. International Journal of Bilingualism, 8(3), 321–354.

12.

‘Five Graces Group’, Beckner

Blythe

Bybee

Christiansen

M. H.

Croft

Ellis

N. C.

Holland

Schoenemann

Larsen-Freeman

(2009). Language is a complex adaptive system: position paper. Language Learning, 59, 1–26.

13.

Gürel

(2008). Review article: Research on first language attrition of morphosyntax in adult bilinguals. Second Language Research, 24(3), 431–449.

14.

Herdina

Jessner

(2002). A dynamic model of multilingualism. Perspectives of change in psycholinguistics. Clevedon, UK: Multilingual Matters.

15.

Isurin

(2000). "Deserted island" or a child’s first language forgetting. Bilingualism: Language and Cognition, 3(2), 151–166.

16.

Jarvis

(2003). Probing the effects of the L2 on the L1. In Cook

(Ed.), Effects of the second language on the first (pp. 81–102). Clevedon, UK: Multilingual Matters.

17.

Jaspaert

Kroon

(1989). Social determinants of language loss. ITL Review of Applied Linguistics, 83, 75–98.

18.

Kaufman

D. H.

(1991). First language attrition as a creative interplay between two languages. Unpublished PhD thesis, State University of New York at Stony Brook.

19.

Köpke

Nespoulous

J.-L.

(2001). First language attrition in production skills and metalinguistic abilities in German-English and German-French bilinguals. In Andersen

R. W.

Hulsen

Strating

Yağmur

(Eds.), Sociolinguistic and psycholinguistic perspectives on maintenance and loss of minority languages (pp. 221-234). Münster, Germany: Waxman.

20.

Köpke

Schmid

(2004). Language attrition: The next phase. In Schmid

M. S.

Köpke

Keijzer

Weilemar

(Eds.), First language attrition: Interdisciplinary perspectives on methodological issues (pp. 1–46). Amsterdam, NL: John Benjamins.

21.

Larsen-Freeman

Cameron

(2008). Complex systems and applied linguistics. Oxford: Oxford University Press.

22.

Lubińska

(2011). Förstaspråksattrition hos vuxna: Exemplet polsktalande i Sverige [Adult first language attrition: The case of Polish speakers in Sweden]. Disserations in Bilingualism 21. Stockholm: Centre for Research on Bilingualism. Stockholm University.

23.

Montrul

(2004). Subject and object expression in Spanish heritage speakers: A case of morphosyntactic convergence. Bilingualism: Language and Cognition, 7(2), 125–143.

24.

Montrul

(2008). Incomplete acquisition in bilingualism. Re-examining the age factor. Amsterdam, NL: John Benjamins.

25.

Opitz

(2010). L1 attrition and L2 acquisition: Global language proficiency and language dominance in adult bilinguals. Eurosla Yearbook 10, 248–281.

26.

Opitz

(2011). First language attrition and second language acquisition in a second language environment. Unpublished PhD thesis, Centre for Language and Communication Studies, Trinity College Dublin, Dublin.

27.

Schmid

(2009). L1 attrition across the lifespan. In de Bot

Schrauf

R. W.

(Eds.), Language development over the lifespan (pp. 171-188). New York, USA: Routledge.

28.

Schmid

(2011). Language attrition. Cambridge: Cambridge University Press.

29.

Seliger

H. W.

(1996). Primary language attrition in the context of bilingualism. In Ritchie

W. C.

Bhatia

T. K.

(Eds.), Handbook of second language acquisition (pp. 605-627). London, UK: Academic Press.

30.

Sharwood Smith

van Buren

(1991). First language attrition and the parameter setting model. In Seliger

H. W.

Vago

R. M.

(Eds.), First language attrition (pp. 17-30). Cambridge, USA: Cambridge University Press.

31.

Smith

L. B.

Thelen

(2003). Development as a dynamic system. Trends in Cognitive Sciences, 7(8), 343–348.

32.

Sorace

(2000). Differential effects of attrition in the L1 syntax of near-native L2 speakers. In Howell

C. S.

Fish

S. A.

Keith-Lucas

(Eds.), Proceedings of the 24th annual Boston University conference on language development (Vol. 2, pp. 719–725). Somerville, MA, USA: Cascadilla Press.

33.

Thagard

(2005). Mind: Introduction to cognitive science. Cambridge, MA: MIT Press.

34.

Tsimpli

Sorace

Heycock

Filiaci

Bouba

(2003). Subjects in L1 attrition: Evidence from Greek and Italian near-native speakers of English. In Beachely

Brown

Colin

(Eds.), Proceedings of the 27th Boston University conference on language development (pp. 787–797). Somerville, MA, USA: Cascadilla Press.