Tải bản đầy đủ

Third language acquition in adulthood

Free ebooks ==> www.ebook777.com


Third Language Acquisition in Adulthood


Studies in Bilingualism (SiBil)
The focus of this series is on psycholinguistic and sociolinguistic aspects
of bilingualism. This entails topics such as childhood bilingualism,
psychological models of bilingual language users, language contact and
bilingualism, maintenance and shift of minority languages, and sociopolitical aspects of bilingualism.
For an overview of all books published in this series, please see
http://benjamins.com/catalog/sibil

Editors
Dalila Ayoun

Robert DeKeyser

University of Arizona


University of Maryland

Editorial Board
Kees de Bot

Aneta Pavlenko

Thom Huebner

Suzanne Romaine

Kenneth Hyltenstam

Núria Sebastián-Gallés

Judith F. Kroll

Merrill Swain

Johanne Paradis

G. Richard Tucker

Christina Bratt Paulston

Li Wei

University of Groningen
San José State University
Stockholm University

Temple University
Merton College, Oxford
University of Barcelona

Pennsylvania State University
University of Alberta

Ontario Institute for Studies in Education
Carnegie Mellon University


University of Pittsburgh

University of London

Volume 46
Third Language Acquisition in Adulthood
Edited by Jennifer Cabrelli Amaro, Suzanne Flynn and Jason Rothman

www.ebook777.com


Third Language Acquisition
in Adulthood
Edited by

Jennifer Cabrelli Amaro
University of Florida

Suzanne Flynn
MIT

Jason Rothman
University of Florida and University of Ottawa

John Benjamins Publishing Company
Amsterdamâ•›/â•›Philadelphia


8

TM

The paper used in this publication meets the minimum requirements of
the╯American National Standard for Information Sciences – Permanence
of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data
Third language acquisition in adulthood / edited by Jennifer Cabrelli Amaro, Suzanne
Flynn, Jason Rothman.
p. cm. (Studies in Bilingualism, issn 0928-1533 ; v. 46)
Includes bibliographical references and index.
1. Multilingualism. 2. Adult education. 3. Language acquisition. 4. Language and
languages--Study and teaching. 5. Language transfer (Language learning)
I. Cabrelli Amaro, Jennifer. II. Flynn, Suzanne. III. Rothman, Jason.
P115.T572012
404’.2--dc23
isbn 978 90 272 4187 0 (Hb ; alk. paper)
isbn 978 90 272 7303 1 (Eb)

2012033116

© 2012 – John Benjamins B.V.
No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any
other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

www.ebook777.com


Table of contents
Acknowledgements
Introduction
Third language (L3) acquisition in adulthood
Jennifer Cabrelli Amaro, Suzanne Flynn and Jason Rothman

vii

1

part 1.╇ Theory
L3 morphosyntax in the generative tradition: The initial stages and beyond
María del Pilar García Mayo and Jason Rothman

9

L3 phonology: An understudied domain
Jennifer Cabrelli Amaro

33

The L2 status factor and the declarative/procedural distinction
Camilla Bardel and Ylva Falk

61

Rethinking multilingual processing: From a static to a dynamic approach
Kees de Bot

79

Multilingual lexical operations: Keeping it all together ... and apart
David Singleton

95

L3/Ln acquisition: A view from the outside
Roumyana Slabakova

115

part 2.╇ Empirical studies
Further evidence in support of the Cumulative-Enhancement Model:
CP structure development
Éva Berkes and Suzanne Flynn
Acquisition of L3 German: Do some learners have it easier?
Carol Jaensch

143

165




Third Language Acqusisition in Adulthood

Examining the role of L2 syntactic development in L3 acquisition:
A look at relative clauses
Valeria Kulundary and Alison Gabriele

195

Variation in self-perceived proficiency in two ‘local’ and two foreign
languages among Galician students
Jean-Marc Dewaele

223

Advanced learners’ word choices in French L3
Christina Lindqvist

255

Foreign accentedness in third language acquisition: The case of L3 English
Magdalena Wrembel

281

Index

311

www.ebook777.com


Acknowledgments
The papers included in this volume were presented at the Obermann Center
Summer Seminar Third Language Acquisition: Developing a Research Base held
at the University of Iowa in conjunction with MIT in the summer of 2010. The
workshop was mainly funded by an extremely generous grant awarded by the
Obermann Center for Advanced Studies, and supplemental monies for the
workshop were also provided by MIT. We are extremely grateful for all the support from the Obermann Center, but especially to the then director Jay Semel
and the current director of operations Neda Hatami. Without their generous
support (financial, logistical and more) this workshop would have been impossible. We are also very grateful for the help of the many graduate students, colleagues and friends who helped to make the workshop run so smoothly. We are
indebted to Felipe Amaro, who organized the vast majority of the event logistics
and so much more.
Finally, we would like to acknowledge and thank all of the participants for
sharing their insightful research and the many colleagues who served as peer reviewers for the papers included in this volume. In alphabetical order, these colleagues were: Irma Alarcón, Larissa Aronin, Mariana Bono, José Luis Blas Arroyo,
Walcir Cardoso, Peter Ecke, Claire Foley, Rebecca Foote, Britta Hufeisen,
Michael Iverson, Carol Jaensch, Scott Jarvis, Jaehuyn Jo, Usha Lakshmanan, Diane
Larsen-Freeman, Gillian Lord, Cristóbal Lozano, Gita Martohardjono, Nicole
Marx, Grit Mehlhorn, Terrence Odlin, Roumyana Slabakova, Rex Sprouse,
Whitney Tabor, Marie-Claude Tremblay, Ineke van de Craats and Mary Zampini.


Free ebooks ==> www.ebook777.com

www.ebook777.com


Free ebooks ==> www.ebook777.com

introduction

Third language (L3) acquisition in adulthood*1
Jennifer Cabrelli Amaro1, Suzanne Flynn2 and Jason Rothman1
University of Florida1 and Massachusetts Institute of Technology (MIT)2

As in all other domains of language acquisition, the study of third language (L3) or
subsequent (Ln) acquisition demands converging evidence derived from a wide
range of theoretical frameworks. And, such investigation from multiple theoretical perspectives will inevitably lead to debates within the L3/Ln community of
researchers. Clearly, as in all fields of science, we may not always agree on, for
example, what constitutes evidence or the theoretical implications of such data.
Each paradigm begins with a certain set of assumptions that then leads to the
development of specific hypotheses to be experimentally investigated. In this
context, one hopes that eventually convergence emerges with respect to the hypotheses generated across paradigms.
Extant studies attempt to lessen paradigmatic divides by, for example,
challenging the macro-field to agree on inclusion and exclusion criteria for what
constitutes an L3/Ln. In addition, attempts are made to establish a common set of
factors that need to be consistently controlled for in empirically based studies
(see, e.g. Cabrelli Amaro in press; Falk & Bardel 2010; Hammarberg 2010; Leung
2007; Rothman, Iverson & Judy 2011; Rothman, Cabrelli Amaro & de Bot in press).
In this way, ‘clandestine’ commonalities might be revealed across paradigms and
fields, providing suggestions for future research where empirical gaps are uncovered. Historically, most research in L3 acquisition has focused on the structure of
the mental lexicon, education and sociolinguistics. More recently, the field has
witnessed a sharp increase in the domain of L3/Ln acquisition of morphosyntax.
However, in spite of these recent trends during the last two decades, we believe
that it is fair to say that the linguistic study of L3/Ln acquisition is still in its infancy. For example, as noted by Cabrelli Amaro (this volume), L3/Ln phonology is
notably understudied, making Cabrelli Amaro’s and Wrembel’s contributions to
this volume even more welcome.

*

The authors’ names appear in alphabetical order; each contributed equally to the chapter.


Free ebooks ==> www.ebook777.com


Jennifer Cabrelli Amaro, Suzanne Flynn and Jason Rothman

Instead of summarizing the entirety of what has already been written about the
field in other relevant sources, we refer the reader to the aforementioned L3/Ln
sources. In this chapter, we attempt to highlight a few important points that, we
contend, thread together the articles that comprise this book. To begin, this collection embodies a wide range of approaches to the study of L3/Ln acquisition, ranging from sociolinguistic perspectives elucidated in Dewaele’s article, the Dynamic
Systems approach discussed in de Bot’s article, to a generative approach to L3/Ln
language acquisition assumed in many of the contributions. The language combinations of the learners covered in this volume are impressive, ranging from, for
example, English, Portuguese and Spanish (see García Mayo and Rothman this
volume) to those such as English, German and Japanese (see Jaensch this volume).
These facts alone allow us to easily understand the significance of L3/Ln acquisition
as an important and distinct field of study. Of course, multilingualism has always
been a linguistic reality; however, it is only recently that multilingualism has not
been considered simply an additive extension of bilingualism, but rather a field that
demands to be acknowledged in its own right. Furthermore, from purely linguistic,
sociological, psychological, educational and cognitive viewpoints, it has become
increasingly apparent that the study of L3/Ln acquisition can provide new insights
to many unanswered questions in these general areas of study (see Slabakova this
volume, for how L3 can inform theories of adult L2 acquisition).
In yet another way, results of L3/Ln studies uniquely inform our understanding of what constitutes the initial state and beyond for language acquisition. A
fundamental question that many of the articles in this volume attempt to address
is that of the role of existing linguistic systems in the process of acquisition of subsequent languages. With respect to the L3/Ln acquisition of morphosyntax, there
are currently three existing models concerning the role of other languages known
in the L3 acquisition process, each of which makes certain developmental predictions. As pointed out initially by Flynn, Foley and Vinnitskaya (2004), examining
the role of previous linguistic knowledge in L3/Ln acquisition allows one to appreciate more profoundly the possible dynamic nature of previous language
influence in acquisition more generally. L3 research offers a unique opportunity to
assess the extent to which language-specific properties of the L1 and/or the L2
determine subsequent language development. Empirical results from Flynn et al.
(2004) suggest that language acquisition is a non-redundant process; that is, learners do not appear to redundantly represent certain fundamental properties of a
specific language grammar already available to them from another language already known to the learner. Providing the Cumulative-Enhancement Model
(CEM), Flynn et al. further argue that, in principle, learners have access to all previous linguistic knowledge for multilingual acquisition but that use of previous
knowledge is delimited in specific ways (see Berkes & Flynn this volume). Both of

www.ebook777.com


Free ebooks ==> www.ebook777.com


Introduction:╇ Third language (L3) acquisition in adulthood

the other models, Bardel and Falk’s (2007) L2 status factor and Rothman’s
(2010, 2011) Typological Primacy Model, agree that linguistic acquisition is not a
redundant process; however, they argue that other cognitive and psychological
considerations conspire in the mind of L3/Ln learners such that the deployment of
previous linguistic knowledge manifests itself somewhat differently than is claimed
by the CEM (see Bardel and Falk this volume and García Mayo and Rothman this
volume for more details). Specific details aside, what all of these models and the
many studies that seek to empirically evaluate them do is collectively highlight one
area of commonality that L3/Ln researchers share, even in the face of some very
fundamental disagreements. As it pertains to most of the studies in this volume,
the language experience variables examined are purely linguistic in nature, although they do not necessarily have to be, as highlighted in other articles such as
that of Dewaele (this volume). That said, the volume’s authors all seem to agree
that a point of departure in L3/Ln acquisition that embraces the unique properties
that differentiate L2 from L3/Ln is essential for many interrelated reasons. As noted above, the relevance of this line of argumentation, overtly expressed or not, is at
the core of all the articles included in this volume. Yet, this is just one area of overlap across paradigms that we could highlight in L3/Ln studies.
Before closing this concise introduction to the volume, we feel obliged to offer
some insights, as we see them, into the future of L3/Ln acquisition. Following
Rothman, Cabrelli Amaro and de Bot (in press), we highlight here four separate
areas related to the empirical study of L3/Ln acquisition that we believe transcend
paradigmatic boundaries and are in critical need of standardization/refinement as
this emerging field continues to develop in the near future: (i) determining what
inclusion and exclusion variables should be applied for subject participants in
L3/Ln studies, (ii) resolving issues related to the comparative fallacy applied specifically to L3/Ln, (iii) creating independent measures of proficiency for L3/Ln
acquisition and (iv) increasing focus on the specific contributions of results from
L3/Ln acquisition research for various subfields of linguistic inquiry, from theory
to practice.
The issue of determining what inclusive and exclusive considerations should
be applied for subject selection in L3/Ln studies is a matter of great importance for
many interrelated reasons. In the absence of some consensus on this issue,
comparisons of data across available studies are rendered difficult, if not impossible. For many reasons, it is not inconsequential to ask whether L3 acquisition
should be assessed in chronological terms. For example, should age of acquisition
be a deterministic inclusion factor by which early childhood bilinguals
(e.g. simultaneous bilinguals, heritage speaker bilinguals and child L2 learners)
acquiring an L3 are considered adult L2 learners if the target L3 is the first language
acquired in adulthood? What about the case of an adult who has learned two




Free ebooks ==> www.ebook777.com


Jennifer Cabrelli Amaro, Suzanne Flynn and Jason Rothman

languages beyond his/her first to a level of high proficiency and attempts a sequentially fourth language? Is s/he an L3 learner or an L4 learner? Should proficiency
of previous language acquisition be a deterministic factor? For example, is someone really acquiring a third language if his/her experience with an L2 resulted in a
steady state that, for whatever reason, is far from the target L2? This, of course, has
serious implications for L3/Ln studies that focus on what is available to the learner
at the initial state (see Rothman & Cabrelli Amaro 2010; Rothman 2011 for more
detailed discussion of this issue). Let us also consider the case of an adult who
begins the study of new languages in close succession to each other before significant acquisition of these languages has occurred. Can there therefore be multiple
L2s or L3s in such cases? Clearly, the issues surrounding the characterization of
the language knowledge of an individual in terms of an L1, L2, L3...Ln are
numerous and complex.
With respect to the inherent comparative fallacy of native vs. non-native comparisons (e.g. Bley-Vroman 1983), it is argued that the learner’s L3 is not uniquely
represented in the learner’s mind/brain. That is, the only distinction to be made
with respect to the representation of language is between the acquisition of an L1
versus the representation of all other non-L1s. If one takes seriously several contemporary arguments about the nature and consequences of multiply represented
linguistic systems in the mind/brain of an individual, then this simplistic distinction is very problematic. For example, consider the argument that individuals with
multiple linguistic systems ‘process’ language differently from monolinguals.
Sorace (2011) argues that issues related to necessary linguistic suppression
(inhibitory control) in bilinguals and related issues with respect to memory systems, including executive control/function, render even the most advanced of L2
learners different from monolinguals. If Sorace’s account proves true, then adding
one more language to this process could have even more drastic consequences for
processing. Thus, why would we use monolinguals as the benchmark of comparison for L3 learners? The question then becomes, what should the standard of
comparison be to gauge L3 knowledge, if there should be any external comparison
at all? Perhaps, as suggested by Bley-Vroman (1983) for L2 acquisition, one should
focus more on internal measures of comparison, such as examining the extent to
which learners make relevant distinctions between carefully counterbalanced
properties of grammar. The answer to the question of what controls are the most
appropriate for L3 acquisition depends in large part on individual research
questions. For example, for L3 initial stage studies examining the nature of transfer, a control comparison might necessarily be different from that of L3
development and/or ultimate attainment studies.
If we are looking at transfer effects from previous languages at the very beginning stages, that is, the initial state or just after, then we need to know the actual

www.ebook777.com


Free ebooks ==> www.ebook777.com


Introduction:╇ Third language (L3) acquisition in adulthood

competence L3 subjects have for the same properties in the L1 and L2 under investigation in the L3 study. With evidence of this competence, when we observe L3
performance and argue that it derives from the L2, for example, then we know
what the learner’s actual L2 competence indicates. If, for example, an apparent L1
effect emerges at an early stage of development, it is possible that the learner does
not have the property in the L2 available for transfer. If this is the case, then L2
transfer is in principle possible, but not for this particular learner, since s/he never
acquired the property in the L2. However, failing to test both the L1 and L2 in addition to the L3 would leave us unable to reasonably argue this point. One can
employ these same standards for research in phonology, morphology, etc. In this
way, the rapidly expanding field of L3 acquisition can critically inform the debates
outlined above.
With respect to independent measures of proficiency for multilinguals, no
standardized measures currently exist (see e.g Cruz Ferreira 2010; Rothman &
Iverson 2010 for some suggestions). At a minimum, L3/Ln involves three linguistic
systems at various levels of representation, which in turn have hypothesized
consequences for other non-linguistic domains of cognition. Thus, steps forward
in understanding how to evaluate L3/Ln proficiency are essential, especially given
the increased importance of multilingualism throughout the world.
We would like to also highlight another inherent difference between L2 and
L3 acquisition as it relates to proficiency testing. Because L3/Ln learners are by
definition not native speakers of the L2, one cannot simply presume the level of
representation of linguistic knowledge in the L2 for any given grammatical
domain without precisely evaluating this proficiency using standardized measures. Thus, testing proficiency in the L2 seems to be essential for participation
in any L3 study, or at least one has to be able to report accurately the L2 proficiency level.
To conclude this introduction to what we believe is an impressive collection
of contemporary studies on L3 acquisition in adulthood, we are confident that
the articles in this volume will make important contributions at multiple levels.
As the field of L3/Ln acquisition grows, the number of linguistic domains investigated increases, the language pairings considered broaden, the methodologies
employed advance, and, as a result, new insights about the linguistic and cognitive underpinnings of language become evident as a result of these unique data.
L3/Ln acquisition is an extremely promising area for future study precisely because of all its potential. We look forward to the time when the proverbial envelope is pushed even further. For now, we are hopeful that this collection helps
this effort by providing individual examples of exemplary studies on L3/Ln acquisition, and by delivering a collection which is greater than the sum of its individual parts.




Free ebooks ==> www.ebook777.com


Jennifer Cabrelli Amaro, Suzanne Flynn and Jason Rothman

References
Bardel, C. & Falk, Y. 2007. The role of the second language in third language acquisition: The
case of Germanic syntax. Second Language Research 23: 459–484.
Bley-Vroman, R. 1983. The comparative fallacy in interlanguage studies: The case of systematicity.
Language Learning 33(1): 1–17.
Cabrelli Amaro, J. in press. Methodological issues in L3 phonology. Studies in Hispanic and
Lusophone Linguistics 6(1).
Cook, V. 1992. Evidence for multi-competence. Language Learning 42: 557–591.
Cruz-Ferreira, M. 2010. Multilingualism, language norms and multilingual contexts. In Multilingual Norms, M. Cruz Ferreira (ed), 1–17. Frankfurt: Peter Lang.
Falk, Y. & Bardel, C. 2010. The study of the role of the background languages in third language
acquisition. The state of the art. IRAL, International Review of Applied Linguistics in Teaching 48(2–3) 185–220.
Flynn, S., Foley, C. & Vinnitskaya, I. 2004. The cumulative-enhancement model for language
acquisition. Comparing adults’ and childrens’ patterns of development in first, second and
third language acquisition. International Journal of Multilingualism 1(1): 3–17.
Hammarberg, B. 2010. The languages of the multilingual: Some conceptual and terminological
issues. IRAL, International Review of Applied Linguistics in Language Teaching 48(2–3):
91–104.
Leung, Y-k.I. 2007. Second language (L2) and third language (L3) French article acquisition by
native speakers of Cantonese. International Journal of Multilingualism 4(2): 117–149.
Rothman, J. 2010. On the typological economy of syntactic transfer: Word order and relative
clause attachment preference in L3 Brazilian Portuguese. IRAL, International Review of Applied Linguistics in Language Teaching 48(2–3): 245–273.
Rothman, J. 2011. L3 syntactic transfer selectivity and typological determinacy: The typological
primacy model. Second Language Research 27(1): 107–128.
Rothman, J. & Cabrelli Amaro, J. 2010. What variables condition syntactic transfer?: A look at
the L3 initial state. Second Language Research 26(2): 189–218.
Rothman, J., Cabrelli Amaro, J. & de Bot, K. In press. Third language acquisition. In The Cambridge Handbook of Second Language Acquisition, J. Herschensohn & M. Young-Scholten
(eds). Cambridge: CUP.
Rothman, J. & Iverson, M. 2010. Independent normative assessments for bi/multilingualism,
where art thou? In Multilingual Norms. Cruz-Ferreira (ed), 33–51. Frankfurt: Peter Lang.
Rothman, J., Iverson, M. & Judy, T. 2011. Some notes on the generative study of L3acquisition.
Second Language Research 27(1): 5–19.
Sorace, A. 2011. Pinning down the concept of “interface” in bilingualism. Linguistic Approaches
to Bilingualism 1(1): 1–33.

www.ebook777.com


Free ebooks ==> www.ebook777.com

part 1

Theory


Free ebooks ==> www.ebook777.com

www.ebook777.com


Free ebooks ==> www.ebook777.com

L3 morphosyntax in the generative tradition
The initial stages and beyond*1
María del Pilar García Mayo1 and Jason Rothman2

University of the Basque Country (UPV/EHU)1, University of Florida and
University of Ottawa2
This chapter introduces the nascent field of third language (L3) morphosyntactic
acquisition research that adopts generative theory. We briefly introduce the
main tenets of generative linguistic theory and how it has been applied to
cases of child first (L1) and adult second (L2) acquisition. After justifying why
it is prudent and profitable to treat L3/Ln learners as unique cases of adult
acquisition as opposed to another instance of adult L2 acquisition, we review
selected studies that have shaped the current research program in generative
L3 morphosyntax research. Finally, we offer some insights on how the study of
L3 acquisition can shed a unique light on questions of significant importance
to generative linguistic and acquisition theories in general and ponder future
research directions.
Keywords: generative theory, third language (L3) acquisition, initial state
models, Typological Proximity Model, Cumulative-Enhancement Model, L2
status factor, Universal Grammar (UG), L1 transfer, interlanguage, morphology/
morphological, syntax/syntactic, mental representation

* Authors appear alphabetically. We would like to thank the Obermann Center for Advanced
Research for the L3 Acquisition: Building a Research Base grant that funded the writing of this
article and the workshop from which it derives. María del Pilar García Mayo acknowledges the
support of funding from University of the Basque Country (UPV/EHU) grant UFI11/06, Basque
Government grant IT311–10 and Spanish Ministry of Education grants FF12009–10264 and
CSD2007–00012. Jason Rothman acknowledges the support of funding from the National Science Foundation grant BCS#1132289, as well as funding from the University of Florida College
of Liberal Arts and Sciences, Departments of Linguistics and Spanish and Portuguese Studies.
We are grateful for the detailed comments of the reviewers and the editors, both of the volume,
Jennifer Cabrelli Amaro and Suzanne Flynn, as well as of the series, Dalila Ayoun and Robert
DeKeyser. Additionally, we thank the many colleagues at the University of Iowa workshop for
helpful discussion and comments as well as our RA, Anne Lingwall, for her editorial expertise
and care. Any and all errors are inadvertent and solely our own.


Free ebooks ==> www.ebook777.com


María del Pilar García Mayo and Jason Rothman

1. Generative theory and acquisition: A concise overview of relevant issues
Generative linguistics is a theory about the mental representation of the subcomponents of language (e.g. syntax, semantics, phonology, etc.), their interfaces as
well as their acquisition.1 In fact, all hypotheses regarding the mental constitution
of language are at the same time proposals of how linguistic properties are acquired. Generative theory maintains that language acquisition is biologically determined, exhibiting the telltale properties that pertain to any instance of natural
growth. Generative theory proposes that children are born with innate linguistic
knowledge of a specific design or Universal Grammar (UG). UG provides a set of
linguistic primitives in the way of universal principles as well as a superset of formal features, functional and lexical categories whose language-specific combinations give rise to cross-linguistic differences (i.e. parameterization). These two
types of properties provided by UG (principles and parameters) filter the input
children are exposed to. At the same time, UG provides the general recipe through
which particular grammars take shape. As such, UG constrains the form that any
human language can take as well as the search space of linguistic possibilities
learners entertain.
Occam’s razor requires that the simplest explanation for any phenomenon be
accepted, provided that such explanation is descriptively and explanatorily adequate. With this in mind, the burden of proof for the indispensability of hypothesizing UG rests in the hands of the generative linguist who must demonstrate why
acquisition in the absence of innate linguistic knowledge would prove an insurmountable task. Alternatively, if the full range of mature linguistic knowledge can
be explained via an interaction of experience (i.e. input) and complex learning
guided by principles of general cognition, then UG should be abandoned in favor
of theories that require less abstraction and domain-specificity. Generative linguists often refer to the logical problem of language acquisition as prima facie evidence in support of UG. The logical problem refers to the complexity, universality,
specificity and subtlety of linguistic knowledge that goes far beyond what is available in the input. That is, as speakers of a language we know much more than we
could have possibly deduced or recovered from input alone. Of course, such a
scenario does not pertain to many properties of specific grammars that are clearly
learned, at least in part, on the basis of input availability and frequency. For some
properties known as poverty-of-the-stimulus, however, there is no positive
1. The section is by design not an exhaustive treatment. We refer the reader to Chomsky
(2007) for a nice overview of the genesis and development of generative linguistic theory over
the past five decades. We further refer the reader to Snyder (2007) and White (2003) for how this
theory has been applied to child and adult language acquisition over the past few decades.

www.ebook777.com


Free ebooks ==> www.ebook777.com


L3 morphosyntax in the generative tradition

evidence in the input from which the ensuing linguistic knowledge could have
derived. Such properties are thus good candidates for the type of knowledge that
is provided innately by UG and thus needs not be learned.
Hypothesizing UG does more than account for the logical problem of acquisition. That the world’s languages only take particular, predictable forms, that
children cross-linguistically achieve the same acquisition milestones at roughly
the same ages and that language acquisition is universally successful for children
to a degree that no other learned experience is are all evidence in favor of a
theory that envisions language acquisition as modular and encapsulated. In light
of this supportive evidence, the generative theory of language is maximally
parsimonious even though it theorizes a higher degree of abstraction when
compared to theories that assume less domain-specificity, such as emergentism/
connectionism (e.g. O’Grady 2005; Tomasello 2003), precisely because generative theory accounts for how the full range of complex syntax and semantics
is acquired.
Within this paradigm, UG-accessibility is never in question as it pertains to
the case of child acquisition, be it L1 or L2/Ln. Alternatively, the past three decades have presented much debate regarding UG-continuity for adult language
learners. For some, the critical period viewed in the generative paradigm is the
post-pubescent loss of accessibility to innate linguistic mechanisms (e.g. Meisel
1997). Adult L2 acquisition, therefore, makes use of non-linguistic learning to
construct grammars and results in L2 grammars that are predicted to not always
conform to UG. Others maintain that L2 interlanguage grammars are entirely
UG-constrained at the level of representation, not because adults have direct accessibility to new L2 features from UG, but rather parasitically through their
UG-sanctioned L1 grammar (e.g. Franceschina 2001; Hawkins & Chan 1997).
More recently, so-called partial access approaches have claimed that adults have
access to some new L2 features, but crucially not to purely syntactic ones
(uninterpretable) needed for new movement operations in the L2 (e.g. Hawkins &
Hattori 2006). Whether some or no new L2 features are hypothesized to be accessible, partial accessibility theories claim that the mental representation of L2
morphosyntax is maturationally conditioned to be different. On the opposite end
of the spectrum, other researchers claim that UG is fully accessible to adults who
have normally acquired their L1, accessed either immediately at the initial state
(e.g. Epstein, Flynn & Martohardjono 1996) or after an initial state of parsing L2
input via partial (e.g. Vainikka & Young-Scholten 1996) or full transfer (e.g. Schwartz
& Sprouse 1996) of the L1.
Many of the same issues and questions that apply to generative L2 theorizing
are equally relevant to L3/Ln acquisition. From a particular point of view, it might
seem reasonable to claim that studying L3 acquisition is in no better (or worse) a




Free ebooks ==> www.ebook777.com


María del Pilar García Mayo and Jason Rothman

position to specifically inform debates on adult UG-accessibility than the amply
studied case of L2 acquisition (see Cabrelli Amaro, Iverson & Judy 2009; Iverson
2009, 2010; Rothman, Iverson & Judy 2011). Nevertheless, it is immediately clear
that the case of L3 acquisition is significantly different than L2 as it pertains to the
possibility of transfer at the initial state and its effects for development and ultimate attainment. These inherent differences provide a locus for new insights into
language acquisition and generative linguistics more generally, upon which we
elaborate further in the next section as we make the case for why researchers must
differentiate empirically between L2 and L3 acquisition.

2. Why L3 as opposed to adult L2 acquisition?
Pointing out that many researchers in second language acquisition (SLA) from all
paradigms have conducted studies on L3 acquisition without, knowingly or
unwittingly, labeling it as such should come as no surprise. Even a random review
of a limited sampling of L2 studies done over the past decades reveals that many
so-called L2 subjects are in fact L3/Ln learners, at least in a chronological sense.
....the field of SLA lacks a clear working distinction between those who are learning a second language and those who are learning third or additional languages....
it is usually up to the researcher to decide whether learners’ prior knowledge has
the potential to bias the result of a study or not. Such freedom of choice, needless
to say, conflicts with the most basic principles of methodological rigor in language
acquisition research. While it may seem obvious to many that the prior knowledge of a non-native language is a variable that needs to be properly controlled,
the reality is that the control for this specific variable is often poor, inadequate, if
not lacking altogether.... (De Angelis 2007: 5–6)

It might be the case that some such learners are child bilinguals (simultaneous bilinguals, child L2ers or heritage speaker bilingual adults) learning their first adultacquired language. Still, it is not uncommon to find adult learners who at the time
of so-called L2 acquisition are in fact learning a third or subsequent language in
adulthood. This fact is not kept secret but becomes immediately clear by reading
the participant section of many studies, particularly those conducted in Canada
and Europe where earlier language learning in schools and the general expectation
that one studies more than one foreign language is more common than in the
United States, for example. Is this imprecision always problematic? Certainly from
the point of view of the researchers who have contributed to this observation, at
least as it pertains to the questions relevant for their particular studies, coupling L2
and L3/Ln learners together as one group is unproblematic. It would appear that

www.ebook777.com


Free ebooks ==> www.ebook777.com


L3 morphosyntax in the generative tradition

the label L2 is, therefore, sometimes used to refer to the acquisition of any language
in adulthood after normal L1 acquisition irrespective of its chronological order.2
In Section 1, we alluded to the fact that this practice might not be overly questionable depending on the research questions one is attempting to address. Let us
provide an example of what we mean by this. As is well known, one of the perennial questions of generative SLA involves determining the role UG plays, if any, in
adult language acquisition. It is not clear that examining L3 acquisition at any L3
stage could be more informative for this specific query than studying L2 acquisition. Thus, for addressing UG-accessibility, separating L2 and L3/Ln learners
might not prove crucial. This is not to suggest that L3 learners specifically cannot
enrich the collective evidence for issues related to adult UG-continuity. For
example, Cabrelli Amaro et al. (2009) and Iverson (2009, 2010) maintain that
examining transfer at the L3 initial state can provide a unique epistemological litmus test for comparative adult L2 theories of UG accessibility. They point out that
if evidence is found in early L3 grammars of L2 functional categories/features
(not present in the L1), then it can be concluded that UG must be accessible in
adulthood, in order to explain how these features came to be present in L2 adulthood and also to be available for L3 initial transfer.3 Another example of how L3
research specifically adds new insights to this question is the growing body of
studies that examine L3 developmental sequence (see Section 4). Such studies,
among other things, are able to comment on UG-accessibility indirectly by providing unique ways of verifying hypotheses that relate to divergent performance as
opposed to competence (e.g. the tenability of the Missing Surface Inflection
Hypothesis, Prévost & White 2000).
Despite the fact that not properly differentiating L2 from L3/Ln learners might
prove to have a negligible effect for certain questions such as UG accessibility, it
certainly can have a detrimental impact for others. In fact, this practice can prove
to be the most deterministic variable explaining some unexpected results in
2. While this description might be true for some, it certainly cannot be generalized to all research that claims that there can be multiple L2s. We refer the reader to Hammarberg (2010) for
a more detailed discussion.
3. Here we must highlight that not showing such evidence at initial stages of L3 acquisition
would not necessarily provide evidence against adult UG-accessibility, unless the possible confound of proficiency level in the L2 is controlled for. That is, if the L3 learners tested have not
reached a proficiency level in the L2 at which a given L2 target feature would have been acquired,
but could be at a later stage of proficiency, then one could hardly expect such transfer to obtain.
As was done by Cabrelli Amaro et al. (2009) and Iverson (2009, 2010), it is best to use very advanced L2 learners as subjects for studies of this nature and, if possible, also test their L2 for the
same properties being tested at the L3 initial stages. In general, proficiency level in the L2 is an
important factor for L3 acquisition studies and the minimum levels for inclusion depend on the
L3 questions at stake in any given study.




Free ebooks ==> www.ebook777.com


María del Pilar García Mayo and Jason Rothman

empirical L2 and L3 investigations (see De Angelis 2007: Chapter 1). At a minimum, the typical adult L3 learner is quite different from the typical L2 learner on
many fronts. Cenoz (2003) reviews what she calls the additive effect of bilingualism on L3 acquisition, arguing that the heightened metalinguistic skill/knowledge
of bilinguals is likely advantageous to L3 learning. From a more formal linguistic
perspective and assuming adult UG-continuity as we do, it follows that L3 learners
have access to more grammatical options (the underlying representations of two
languages) when it comes to initial hypotheses that feed into L3 interlanguage
development. Unless we can justify an absolute L1 factor effect for transfer, which
we will see in the next section cannot be done, irrespective of whether having
more options for initial state transfer proves to be facilitative, non-facilitative or
both, L3 acquisition is de facto different than L2 acquisition from its start, which
of course has implications for development and ultimate attainment (i.e. the possibility of novel learnability problems that arise).
Related to transfer, examining L3/Ln transfer specifically provides an unprecedented window into the dynamic nature of something that cannot be examined
in L1 acquisition to any degree and not deeply enough in L2 acquisition. In the
case of L2 acquisition, transfer has only one potential source and, therefore, examining it cannot tell us much, if anything, about the economic nature of transfer and
how it might inform debates on the mental constitution of grammar more generally. Following Flynn, Foley and Vinnitskaya (2004) and Rothman and Cabrelli
Amaro (2010), we contend that studying L3 acquisition can and we will highlight
how this has been done thus far in the next section, which examines the current
models of the L3 initial state.
There are many more reasons why we should always differentiate properly between true adult L2 learners and multilingual learners, ranging from issues of scientific methodological prudence to the range of obscuring effects that not doing so
might entail. However, space does not permit us to explore them in detail. With
this general proviso in mind, the remainder of this chapter reviews the exemplary
work within generative acquisition that has heralded these concerns and seen the
value that studying L3 has for contributing to old questions and the creation of
new ones.
3. The initial state and different proposals for L3/Ln
Essentially, the process of language acquisition and its examination can be divided into three parts: the initial state, interlanguage development and ultimate
attainment. In this section, we will focus on the first stage of L3/Ln acquisition,
that is, the L3/Ln initial state. The initial state refers to the very beginning stages

www.ebook777.com


Free ebooks ==> www.ebook777.com


L3 morphosyntax in the generative tradition

of language acquisition and is defined as the set of linguistic hypotheses with
which the learner begins the acquisition process. For child L1, the initial state is
assumed to be UG. Because transfer, at least partially, is ubiquitously observed for
L2 learners (for children and adults alike, see Schwartz 2003) and because by
some accounts adults lose access to UG after puberty, one cannot assume that the
initial states of primary and non-primary acquisition are the same (but see
e.g. Epstein et al. 1996).
Within adult generative L2 morphosyntactic acquisition, the initial state has
been the focus of much research, especially in the 1990s (see White 2003:
Chapter 3). Understanding the initial state is of great importance beyond determining the preliminary representations used to parse the target linguistic data in
non-primary acquisition. Accurately describing the initial state is a necessary first
step to explaining the behavior attested to by adult learners in interlanguage development and figuring out the unique learnability problems they face as it relates to
both developmental sequence and ultimate attainment. Just as it is the case that
one cannot assume that the initial states of L1 and L2 are the same, one cannot
make such an assumption for L2 and L3/Ln precisely because L3/Ln learners have
more transfer sources for initial hypotheses. Notwithstanding, it is equally true for
the case of L3/Ln acquisition that the accurate description of the initial state has
entailed implications for studying and understanding developmental sequence
and ultimate attainment potential for multilingualism specifically. Thus, it is not
surprising that a considerable amount of current generative L3 research focuses on
initial state transfer models, which we review in this section.
Without further discussion and before detailing the current L3 initial state
models, we put aside some of the previously articulated L2 models of the initial
state that do not assume full transfer such as Minimal Trees (Vainikka &
Young-Scholten 1996), The Valueless Features Hypothesis (Eubank 1993/1994)
and the Initial Hypothesis of Syntax (Platzack 1996). The latter two hypotheses are
based on the notion of feature strength and checking as described in Chomsky
(1995) and are thus rejected on the basis that, as syntactic theory has developed
and abandoned the notion of feature strength, they can no longer be deduced from
UG. Minimal Trees, which is amendable to current theoretical assumptions,
proposes that adults transfer lexical categories only (and thus parameter settings
therein, such as head directionality). Minimal Trees would, therefore, predict that
only lexical categories could be transferred into the initial state of L3 as well. While
Minimal Trees could be tested in L3, we reject it, however, on the basis of much
work in L2 acquisition that demonstrates robust counter evidence to its predictions. The current models of the L3 initial state also eliminate a priori proposals of
inaccessibility or partial access to UG, showing instead that new L2 morphosyntactic properties are acquired by successful L2ers insofar as they are available for




Free ebooks ==> www.ebook777.com


María del Pilar García Mayo and Jason Rothman

initial state transfer in L3/Ln acquisition. As such, all existing L3 models that we
cover herein can be viewed as corollaries at some level to the Full Transfer/Full
Access initial state model for L2 acquisition (Schwartz & Sprouse 1996).4
Because the field of generative L3/Ln is in its early stages, there are logical models of the L3/Ln initial state that have not been proposed (yet) or investigated. Given
space limitations, we will focus exclusively on the models that have appeared in the
recent literature, which include the L2 status factor, the Cumulative-Enhancement
Model and the Typological Primacy Model. One exception to this will be a brief
discussion of absolute L1 transfer and its very clear, straightforward predictions for
an L3/Ln initial state. Because this chapter is concerned primarily with generative
morphosyntax, we will not review other models that make predictions for transfer
relating to the lexicon and other domains, which have been explored in other
paradigms (e.g. de Bot 2004; Williams & Hammarberg 1998).
3.1

Absolute L1 transfer

Without question, absolute transfer of an L1 is certainly a possibility for the L3/Ln
initial state despite the fact that such a position has never been systematically advanced within the generative paradigm (but see Na Ranong & Leung 2009). De
Bot (2004), from a psycholinguistic perspective and referring specifically to lexemes, suggested that the L1 should have some type of privileged status in that
domain, although his claim is far from uncontroversial (see Lindqvist 2009).
Absolute L1 transfer would suggest that the L1 acts as a filter of sorts, impeding
access to acquired L2 properties. Interestingly, such a position would also follow
directly from theories claiming that adult L2ers do not have access to UG under
which target L2 mental representations that differ from the L1 are hypothesized
not to be attainable. Logically, if the underlying representation of properties in the
L2 that diverge from the L1 cannot be acquired, then this is tantamount to saying
that the L1 grammatical system is the sole source of transfer in all instances of
adult language acquisition irrespective of chronological order.
An absolute L1 transfer position is a strong hypothesis inasmuch as it makes
very clear predictions. Such a position anticipates no difference in the initial states
of L2/L3 acquisition, at least any that can be clearly likened to the L2 grammatical
system. Absolute L1 transfer could be supported by studies comparing L2 and L3
learners of the same adult target language at the initial state where transfer effects
4. As intended by Schwartz and Sprouse (1996), the initial state is more precisely what the
learner brings to the acquisition processes at the very first moments of new target language exposure. The L3 models covered herein take a more liberal definition whereby they are referring
to the initial stages of L3 acquisition. We note that this is not entirely imprecise since this more
liberal definition includes the actual initial state as intended by Schwartz and Sprouse (1996).

www.ebook777.com


Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Tải bản đầy đủ ngay

×