EDITORS: J.M. Liceras, H. Zobl & H. Goodluck TITLE: The Role of Formal Features in Second Language Acquisition PUBLISHER: Routledge: Taylor and Francis YEAR: 2008 (paperback 2010)
Ferid Chekili, University of Manouba, Tunisia; University of Bahrain
The collection under review investigates the role of formal features in current analyses of second language acquisition (L2) -- and to a lesser degree of first language acquisition, impaired language development (SLI), and aphasia.
In addition to an ''Introduction'' by the editors, the collection is built around four parts: in Part 1, containing four chapters, learnability issues characterizing normal and pathological language acquisition, and language breakdown, are addressed in relation to the Minimalist feature-based theory. Likewise, Part II consists of four chapters dealing with DP features (such as Definiteness, Case, Person, Number...) in L2 and SLI grammars. Themes discussed include the interpretable vs. uninterpretable contrast (Chomsky 1995), and the relationship between functional morphology and syntax. Part III has two sections: section 1 consists of four chapters which deal with the L2 acquisition of Finiteness, Agreement, and Tense; a recurring theme is a comparison between the Failed Functional Features Hypothesis and the Missing Surface Inflection Hypothesis. The second section also contains four chapters and is concerned with Aspect-related features. Finally, Part IV consists of two chapters dealing with CP features.
In their introduction, the editors present an overview of the theory of features in both linguistic theory and learnability theory. The overview draws both on the contents of the edited collection and the general literature. The gist of the first part of the introduction is to demonstrate the important role features have played within Minimalism in accounts of language variation, which makes them quite relevant for any discussion of learnability issues. It also concludes that despite many unresolved questions, the role played by features in accounts of language variation (e.g. in explaining lack of success) is not to be doubted. The second part of the introduction outlines the contents and organization of the book.
Part I Linguistic Theory and Learnability
In chapter 1 (The Role of Features in Syntactic Theory and Language Variation), Travis begins her contribution by arguing that within Chomskyan syntactic theory, features have been used to capture language variation, and hence, “any study of language acquisition done within this framework is now a study of the acquisition of features” (23). Her second section is devoted to showing the role of features -- in early Minimalism -- as an account of language variation, and hence, language acquisition. She concludes that this early use of features was not without problems and will be replaced by a more explanatorily adequate system in which “the feature that needs to be checked [is separated] from the feature that triggers movement” (28). Using Boskovic (1998) and Cheng (2000) as examples, she then addresses the possibility of moving features by themselves, independently of the lexical item which introduces them, and concludes that “by having feature movement, we have an elegant explanation for the covert movement facts of French and the overt realization of a default wh-Comp in German” (35). In the final section, she shows that the feature-based system may seem to be too flexible but that, in fact, “language variation calls for exactly this sort of flexibility” (36). She illustrates using examples from languages where a V-feature triggers XP movement and a D-feature triggers D-movement. She also notes that the feature system (unlike the previous one) is not well-suited to accounting for certain types of ill-formedness and that the answer to this and other problems may come from acquisition data.
In chapter 2 (Uninterpretable Features and EPP: A Minimalist Account of Language Buildup and Breakdown), Platzack, using Pesetsky and Torrego’s (2001) feature-driven version of Minimalism, assumes that early L1 children, children with SLI, adult L2 learners, and Broca’s aphasics have full grammatical knowledge of the features in C and T in Swedish (in particular, EPP). Such knowledge has to do with movement of the verb to C, occurrence of subject DP, and movement of wh-elements to CP. Platzack endeavours to demonstrate that the production errors of the subjects above, “can be accounted for if we assume that these speakers cannot automatically use their knowledge of the distribution of EPP in the language they are speaking” (50), i.e. their performance is not always adultlike. He further assumes that the errors involve CP and TP and not vP.
Radford, in chapter 3 (Feature Correlations in Nominative Case Marking in L1, L2, and Native English) reconsiders Iatridou’s (1993) parametric view of nominative case-assignment (in Greek) in connection with standard English, and argues that mood (and agreement) is what determines nominative case-assignment in English. He shows that the same account can be extended to L1 learners who use both nominative and accusative subjects in the same (finite) contexts and to L1 and L2 learners who use only nominative subjects in the same contexts. Such a parametric account involves a feature selection/combination of the following: mood/person/EPP/number/tense-features. Finally, he shows that the features are ordered according to whether they are obligatorily or optionally specified, resulting in: mood>EPP/person>number>tense.
In chapter 4 (Feature Assembly in Second Language Acquisition) Lardiere addresses parameter resetting in SLA as conceptualized by the representational deficit approach which “attributes L2 inflectional variability or error to a failure in the selection of parametrised formal features” (109). She argues that “accounting for morphological variability simply by appealing to the parametric (non)selection of features is too simplistic” (111). She believes, instead, that L2 acquisition is affected by the way morphological features are assembled or combined. She illustrates the idea with specific examples showing how the different combinations of features in the L1 and the L2 may yield learning problems, concluding that “there is a kind of morphological competence that must be acquired by the learner”, namely, the knowledge of “which forms go with which features” (111).
Part II Determiner Phrase-Related Features
In chapter 1 (Feature Interpretability in L2 Acquisition and SLI: Greek Clitics and Determiners), using data from L2 acquisition and SLI of Greek clitics and determiners, Tsimpli and Mastropavlou approach a theory of learnability from the point of view of the “interpretability status of formal features” (143). In particular, they argue that uninterpretable formal features are inaccessible to both types of learners, though the reasons may be different: critical period effects for the first; genetically-based deficiencies for the second. In particular, the definite article and the third person accusative clitic (consisting of clusters of uninterpretable features only) are shown to be more problematic for L2 learners than the indefinite article and the first/second person and genitive clitics (assumed to have an interpretable person/-definite feature). Thus, “interpretable features are ... accessible in all processes of language development, whereas uninterpretable features become inaccessible either due to constraints related to the critical period hypothesis or to the incomplete or deficient analysis of L1 input” (154).
Jakubowicz and Roulet examine, in chapter 2 (Narrow Syntax or Interface Deficit? Gender Agreement in French SLI), the question whether variable use of grammatical morphology may be explained as the result of a syntactic deficit or a production (processing) deficit. This question is discussed in the context of “a study on elicited production and perception of Gender agreement conducted with French-speaking children with SLI and a group of normally developing controls” (186). The production task shows that whereas the latter do not seem to have any problems with Gender agreement, the former do. The perception task, however, shows that both populations are equally sensitive to Gender agreement. The conclusion drawn by the authors is that contra Gopnik (1990) and Clahsen (1989), who take agreement errors to reflect a syntactic deficit, “the source of the errors made by SLI children in the production task does not lie in the computational component of the language faculty but in accessing and integrating different types of information at the interfaces that relate language to other cognitive systems” (186).
In chapter 3 (The Role of Semantic Factors in the Acquisition of English Articles by Russian and Korean Speakers) Ionin, Ko, and Wexler are concerned with the semantic features ([+definite] and [+specific]) that dictate article choice in the DP, and with the importance of these features for the acquisition of English articles by adult speakers of Russian and Korean, languages with no articles. The authors show that the [+definite] and [+specific] features belong to a cross-linguistic inventory of article specifications. This suggests that learners have to discover how this feature contrast is organized in the language they are learning. In other words, learners of English by L1 speakers of articleless languages such as Russian and Korean have to find out whether the English article system is parametrised around the [definite] feature or the [specific] one. The rationale behind the choice of this particular population of learners is to see whether these learners would use ‘the’ and ‘a’ in contexts other than the English [+/-definite] ones, given that they ought to have access to all settings of the parameter; i.e. ‘the’ could be used in [-definite, +specific] contexts, and ‘a’ in [+definite, -specific] contexts. Using a number of tests, they conclude that until they have received sufficient input pointing to the [+definite] feature as the correct option for English, L2 learners of English fluctuate between two possibilities when learning articles: articles are distinguished 1. on the basis of definiteness, and 2. on the basis of specificity. Therefore, the “findings provide evidence for direct access to universal semantic features in L2 acquisition, as well as for the reality of the feature [+specific]” (263).
Chapter 4 (Acquisition of the Spanish Plural by French L1 Speakers: The Role of Transfer), by de Garavito, considers the relation between overt morphology and syntax in L2 acquisition: variable inflectional morphology is seen by some as indicative of problems with the morphosyntactic representation of the grammar in adult learners, a competence problem (e.g. Liceras 1997, Tsimpli and Roussou 1991) and by others as simply a problem of mapping of the correct form (e.g. Lardiere 1998, Prevost and White 2000). To test these positions, de Garavito reports on a disassociation between inflection and syntax found in French learners of L2 Spanish: she notes that these speakers show variability in their production of the plural in Spanish; this cannot be predicted by transfer as French, like Spanish, has strong number features. On the other hand, these speakers have no problem with acquisition of the noun-adjective word order (“which is consistent with knowledge of the strong number feature of Spanish” (272)). She argues that “this contradiction can best be accounted for by transfer -- not transfer in the domain of functional categories but rather at the prosodic level” (272). The results of the study argue against inflectional variability as being the result of a representational functional deficit as well as against the No Parameter Resetting Hypothesis (e.g. Liceras 1997) as “the learners were able to acquire the relevant knowledge of Spanish syllable structure” (293).
Part III Inflection Phrase and Aspect Phrase-Related Features
Section 1 Finiteness, Agreement, and Tense. In chapter 1 (Some Puzzling Features of L2 Features) White deals with the question of morphological variability in L2 acquisition -- more specifically, Mandarin and French learners of English (where French shares with English the overt morphological realization of the features considered, and Mandarin does not). She compares three current hypotheses to L2 morphological variability (the Failed Functional Features Hypothesis (FFFH), the Missing Surface Inflection Hypothesis (MSIH), and the Prosodic Transfer Hypothesis (PTH)) and their predictions with respect to English verbal and nominal features. She then tests the predictions in an experiment described in section 4. The three hypotheses are equally incapable of explaining one aspect of the results: the two groups perform better on noun plurals than on 3rd singular verbal agreement. More generally, none of the hypotheses seems to be capable on its own to account for morphological variability, which suggests that “... a combination of theories is necessary in order to account for the performance of L2 speakers in the morphological domain” (321).
In chapter 2 (The Semantic Effects of Verb Raising and Its Consequences in Second Language Grammars) Hawkins, Casillas, Hattori, Hawthorne, Husted, Lozano, Okamoto, Thomas, and Yamada consider the semantic reflexes of verb raising. Whereas raising is associated with an event-in-progress/existential interpretation, absence of raising is associated with a habitual/generic interpretation. The study investigates the ability of highly proficient second language speakers of English (including speakers whose L1s have no verb raising -- Chinese, Japanese -- and speakers whose L1s do -- Arabic, French, German, Spanish) to judge the appropriateness of a number of continuations in contexts that are either clearly progressive or clearly generic. The results show that the L2 groups under investigation can successfully distinguish between appropriate and inappropriate interpretations of the simple present/past tense and the progressive and therefore, they must all have acquired the English feature representations for T and v which are assumed to determine tense and progressive. However, it is also shown that there are significant differences between the two groups. In connection with how they judge the appropriateness of ‘be+ing’ with an event-in-progress/existential reading, speakers of V-raising languages are like the native control groups, while the Chinese/Japanese speakers are different from the native control group. On the other hand, regarding how they judge the inappropriateness of habitual/generic readings with ‘be+ing’, Chinese/Japanese speakers are like natives whereas speakers of V-raising languages are different from natives. “English appears to provide positive evidence for the uninterpretable [uInfl:] feature of v and the [uInfl:*] feature of Progressive” (348) associated with these distinctions, and therefore, if UG is fully available in L2 acquisition, this will mean that all the uninterpretable features will be represented regardless of L1 (contrary to findings). The authors conclude, following Tsimpli, that “although interpretable syntactic features provided by UG are available for use in grammar construction throughout life, uninterpretable features that are not instantiated in primary language acquisition may be subject to a critical period. Where such features are not available, L2 learners use other UG-determined resources to model input” (348).
In chapter 3 (Knowledge of Morphology and Syntax in Early Adult L2 French: Evidence for The Missing Surface Inflection Hypothesis) Philippe Prevost examines the question of whether morphological variability reflects impaired grammatical features in underlying grammars. He addresses this issue using cross-sectional data. The results point to the correctness of the non-impairment view of acquisition, i.e. there is evidence of knowledge of functional categories realizing finiteness, of features, and of feature strength. Similarly, the data show the Impairment Representation Hypothesis to be incorrect (e.g. finite verbs are shown to be placed to the left of negation most of the time). “The results also suggest that L2 acquisition of syntax is independent of the acquisition of morphology” (373): syntactic properties seem to be acquired despite the absence of inflectional morphology.
Chapter 4 (The Verbal Functional Domain in L2A and L3A: Tense and Agreement in Cantonese-English-French Interlanguage) is contributed by Leung, who investigates the question of L3 acquisition and its relation to L1 and L2 acquisition. The author conducts experiments testing the status of tense and agreement features and feature strength in L1 Cantonese, L2 English, and L3 French; more specifically, Leung aims to discover whether Cantonese L1 speakers learning French as an L3 can build a TP with [+/-finite], [+/-past] and agreement features, and a [+strong] value. The idea is that if the L1 were the initial state, the acquisition problem would be, for these learners, to construct a TP projection -- since Cantonese does not include one; if, on the other hand, the initial state were the L2 (English), then the problem would simply be resetting the value for the [+strong] feature -- as English and French differ in this respect. Contra the FFFH, the results show that the formal features are still accessible in L2 and L3, even if they are not available in the L1. Moreover, the data shows that not only does the L2 grammar facilitate the acquisition of the L3, but the L3, in turn, may influence the L2 grammar. She concludes that no impairment is involved in L2 and L3 grammars and that “the more languages one has acquired, the easier it will be for the learner to acquire a new additional language…” (399).
Section 2 Aspect. In Chapter 5 (On the Role of DP in the Acquisition of Finiteness in Child L2 English) Gavruseva investigates the relationship, in child L2 English acquisition, between the acquisition of DP -- in particular, the feature Q embedded in the DP -- and the acquisition of finiteness in clauses (AspP). She argues that finiteness (tense and agreement) will not emerge until the Q and definiteness features have been acquired. In other words, “the underspecified nature of DP in early child L2 contributes to the RI (Root Infinitive) effect, that is, an extensive omission of finiteness morphology with lexical verbs” (405). The data investigated shows high omission rates of determiner ‘the’, which suggests that Q cannot be licensed via D, which, in turn, means that the two features Q and D are underspecified, resulting in the overuse of Root Infinitives.
In chapter 6 (Non-Native Recognition of Iterative and Habitual Meanings of Spanish Preterite and Imperfect Tenses), Perez-Leroux, Cuza, Majzlanova, and Sanchez-Naranjo propose to investigate the L2 acquisition of aspect. In order to do so, they compare two approaches to the acquisition of aspect: “in the featural model, acquisition consists of feature activation...”; “whereas in the selectional approach, acquisition has a lexical nature development, where learners must bootstrap the selectional features of each functional head independently” (443). To this end, the authors use iterative and habitual contexts (as they are similar in both requiring “coercion” and in their lexical semantics): habitual meaning is expressed by the imperfect and iterative (and “coerced” iterative) meanings are expressed by the preterite. The aim is to determine whether the subjects (adult English L1 learners of Spanish) are able to learn the differences between the aspectual value of the preterite and the iterative value “coerced”. They argue that learners are able to acquire the habitual imperfect and the punctual preterite but have difficulties with certain “coercion” contexts. They conclude that these findings -- namely, the acquisition of certain morphemes and not others -- follow naturally from a selectional approach that applies independently to each head, rather than from the featural -- morphosyntactic -- approach (where the aspectual distinctions depend on the interpretable features).
In Chapter 7 (Aspectual Shifts: Grammatical and Pragmatic Knowledge in L2 Acquisition) Slabakova and Montrul investigate the possibility that “competence-performance discrepancies in the L2 acquisition of syntax and semantics” results from the use of pragmatic knowledge. The authors make a distinction between grammatical knowledge and pragmatic knowledge. The former involves features checked in the syntax; the latter requires the satisfaction of certain interface conditions, namely, the interface between discourse and syntax (which has been shown to be problematic for children). Therefore, “by focusing on the L2 acquisition of predicates that shift aspectual class under different pragmatic conditions in Spanish, [the authors] ask whether L2 learners’ non-targetlike linguistic behavior is due to pragmatic processing abilities or to differences in the operation of grammatical feature checking” (456). The question underlying their study is “whether for English-speaking learners of Spanish, aspectual shifts triggered by feature-checking aspectual operators like direct objects, the telicity marker ‘se’, and grammatical aspect inflection are easier to acquire than aspectual shifts triggered by more discourse pragmatic signals such as adverbials” (464). The authors’ conclusion is that only the investigation of the semantics-pragmatics interface will be able to explain the differences between L1 and L2 acquisition.
Chapter 8 (Interpretable and Uninterpretable Features in the Acquisition of Spanish Past Tenses; Diaz, Bel, and Bekiou) “deals with the acquisition of the aspectual contrasts conveyed by Spanish preterite and imperfect in relation to four types of predicates: states, activities, accomplishments, and achievements”. After describing the different aspectual systems in the L1s under investigation, and presenting a Minimalist Program view of aspect (where the distinction between [+/-interpretable] plays an important role), the authors formulate a number of hypotheses to be tested. The analysis of the data shows that Romance and Greek learners (whose L1s are related to Spanish) have an advantage. For the other language types, difficulties arise in connection with activities and accomplishments which “involve an interaction between interpretable and uninterpretable features”. Overall, the authors maintain that learners from all L1 backgrounds will have difficulties with activities and accomplishments precisely because telicity depends on the interaction of interpretable features with uninterpretable ones.
Part IV Complementizer phrase-related features
Chapter 1 (Complementizer Phrase Features in Child L1 and Adult L3 Acquisition) is contributed by Flynn, Vinnitskaya, and Foley. After dealing with the role that CP features play in free relatives and headed relatives, and arguing that experience with free relatives facilitates the acquisition of headed constructions, the authors review results from L1 studies which indicate: “(a) CP features are accessible to children and active in their construction of relative clauses; (b) language-specific information about the feature content of CP is especially accessible in the free relative, and (c) children are able to use language-specific discoveries from the free relative in constructing the lexically headed form” (516). Using a previous study by Flynn (1989) which showed that CP features are available in L2 acquisition, and that, in case the two languages are similar, the L1 can facilitate L2 acquisition, the authors consider further evidence for this claim from L3 acquisition: in particular, from “the acquisition of relative clauses in a head-initial language, English, by adults whose L1, Kazakh, is head-final, but whose L2, Russian, is head-initial” (516). A production study is undertaken where subjects are tested on free and headed relatives. “The findings support continuous accessibility of CP in language acquisition in L1, L2, and beyond. Logically, therefore, they point to continuous accessibility of the formal features that the CP hosts” (517).
Chapter 2 (On CP Positions in L2 Spanish) is by Valenzuela. Here too, the two general theories of acquisition (the impairment vs. nonimpairment views -- specifically, the No Parameter Resetting Hypothesis (NPRH) and the Full Transfer/Full Access Hypothesis (FTFA)) are compared with respect to the acquisition of topic constructions in L2 Spanish by native speakers of English. The idea is to test whether English speakers can learn the properties of Spanish topic constructions which are not found in their L1, namely, recursive topics, clitic projections, and null anaphoric operators (also present in the L1). The results of the experimental tasks consisting of a sentence completion task and an oral grammaticality judgment task, suggest that “both functional projections and some associated properties that are not present in the L1 are acquirable in post-childhood L2 acquisition. This evidence suggests that the L2 end-state grammar can achieve nativelike competence, in principle, thereby providing evidence in favour of the FTFA” (556).
In my opinion, the contribution of this volume is in providing an answer to the question: how has SLA research coped with developments in linguistic theory? Indeed, as mentioned in the preface, “this book is the first contribution devoted to the use of feature-based linguistic theory in L2A research”. Not only have features been shown here to constitute an important device for describing “parametric contrasts” between the L1, L2 (and L3), and for “describing a learner’s interlanguage grammar”, but they have also been used “to account for difficulties or lack of success in L2A” (vii). In order to do so, most chapters argue for one or the other of two competing theories of acquisition: difficulties are accounted for either in terms of a “representational deficit” or in terms of “output limitations”. This is a common thread linking the chapters together and allowing the book to have a clear-cut internal organization.
Other significant and helpful features of the collection include the author and subject indexes provided, but even more importantly, the chapter summaries at the start of each chapter (with the exception of chapters 3.7 and 4.2) containing relevant information about the chapters, such as the research questions and the findings.
The introduction is another strong feature of the book in that the editors have succeeded in linking together points from different articles. They have managed to do so by identifying certain issues and seeing how the different chapters would deal with them. Comparing the various treatments, the introduction looks for generalizations and draws pertinent conclusions.
Some of the weaker points of the collection include the following:
1. As research in SLA is often the product of refinements in linguistic theory, and owing to the growing importance in recent research of the role that CP features (in particular, information structural features) have come to play in bridging the gap between pragmatics, syntax, and prosody, I would have expected (and wished) part IV to be more representative of this tendency, and to include more chapters on topic/focus features. Indeed, Generative SLA, together with typical theoretical syntactic research, has recently focused on the acquisition of grammatical interfaces in discourse-configurational languages (in particular, syntax/information structure, and syntax/prosody interfaces) best understood in connection with the acquisition of topic/focus features.
2. There are a number of (minor) problems (mistakes, ambiguities, speculations). Some examples:
p. 96, example (21) does not match its description; p.98, there is no ‘wants’ in (22a); p.148, example (3c) needs to substitute for (3d); p. 277, English plural allomorphs are given to be [-s], [-z], and [-es]; p. 278, “In the next section [section 6] ... the results ...”: results appear in section 7 and not 6; pp. 281-282, a confusion between (7) and (8); p. 316 (second paragraph), “regular” instead of “irregular”; p. 196, in order to explain the early learning of gender agreement between the determiner and the noun, the authors have to speculate that “the acquisition process proceeds so quickly that little time seems to be left for errors to show up”. Another speculation can be found on p. 319, where the author, showing that the results are not consistent with the predictions of the MSIH, concludes the paragraph by saying that other results (Ionin and Wexler 2002, p.109) are consistent with the MSIH, without elaborating on the significance/implications of this contradiction. At least one chapter (2.2) is concerned primarily with children with SLI, whereas the collection is meant to be primarily a study of L2 acquisition.
Nonetheless, as shown at the beginning of this section, the book’s strengths outweigh its weaknesses.
Boskovic, Z. 1998. LF movement and the minimalist program. Proceedings of the annual meeting of the North East Linguistic Society 28, 43-58.
Cheng, L. 2000. Moving just the feature. In G. Muller, U. Lutz, & A. von Stechow (Eds.), wh-scope marking, 77-99. Amsterdam: John Benjamins.
Chomsky, N. 1995. The minimalist program. Cambridge, MA: MIT Press.
Clahsen, H. 1989. The grammatical characterisation of developmental dysphasia. Linguistics 27, 897-920.
Flynn, S. 1989. The role of the head-initial/head-final parameter in the acquisition of English relative clauses by adult Spanish and Japanese speakers. In S. Gass & J. Shachter (Eds.). Linguistic perspectives on Second Language Acquisition, 89-108. Cambridge, UK: Cambridge University Press.
Gopnik, M. 1990. Feature blindness: A case study. Language Acquisition 1(2), 139-164.
Iatridou, S. 1993. On nominative case assignment and a few related things. MIT Working Papers in Linguistics 19, 175-196.
Ionin, T. & Wexler, K. 2002. Why is 'is' easier than '-s'?: Acquisition of tense/agreement morphology by child second language learners of English. Second Language Research 18, 95-136.
Lardiere, D. 1998. Dissociating syntax from morphology in a divergent end-state grammar. Second Language Research 14, 359-375.
Liceras, J.M. 1997. The now and then of L2 growing pains. In L. Diaz Rodriguez & C. Perez Vidal (Eds.). Views on the acquisition and use of a second language, 65-85. Barcelona, Spain: Universitat Pompeu Fabra.
Pesetsky, D. & Torrego, E. 2001. Tense-to-C movement: Causes and consequences. In M. Kenstowicz (Ed.), Ken Hale: A life in language, 355-426. Cambridge, MA: MIT Press.
Prevost, P. & White, L. 2000. Missing surface inflection or impairment in second language acquisition? Evidence from tense and agreement. Second Language Research 16(2), 103-133.
Tsimpli, I.M. & Roussou, A. 1991. Parameter resetting in L2? University College London Working Papers in Linguistics 3, 149-169.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Ferid Chekili has taught Linguistics and Language Acquisition in different
colleges in several countries. His primary research interests include
syntactic theory, Second Language Acquisition, and the syntax-information