Ultimate Attainment in Second Language Acquisition

A Case Study

Donna Lardiere

• Publisher: Routledge
• Year: 2017
• Print ISBN:
  9780805834567
• eBook ISBN:
  9781351538237
• Topic: Education
• Subtopic: Education Theory & Practice

The first book-length treatment of its type, Ultimate Attainment in Second Language Acquisition is a case study with a solid theoretical grounding that examines the language of an immigrant learner of English, and thereby presents a much needed understanding of the linguistic competence of second language speakers. Based on longitudinal data collected over a period of 16 years, this clear and accessible presentation is well-grounded in linguistic theory and in second language acquisition research issues.

Author Donna Lardiere presents the narrative of Patty, an adult Chinese immigrant learner of English, who achieves native-like proficiency in some areas of her English idiolect, although reaches a plateau in her language acquisition, known as the concept of fossilization. By addressing this concept, a central idea in second language acquisition research, Lardiere fills a void in existing literature. Individual chapters focus on Patty's end state knowledge of grammatical areas of finiteness, past-tense marking, word order, wh-movement and relativization, passivization, number marking, and use of determiners. Important topics discussed throughout the book include:
*learner variability in production;
*case study methodology;
*the roles of motivation and prior language (L1) knowledge; and
*sensitivity to input in circumscribing ultimate attainment in adult second language acquisition.

Ultimate Attainment in Second Language Acquisition is intended for anyone whose research is in the areas of second language acquisition, language acquisition, theoretical, applied, or developmental linguistics. It is also appropriate for graduate level students of TESOL and teachers who work with more advanced learners of foreign languages.

1
Some Preliminary Issues in Adult L2 Ultimate Attainment

The impetus for this study arose from a long-standing acquaintance with Patty, a Chinese-American naturalized immigrant whose typical everyday exposure to the English language is, in many respects, very similar to mine and has been so for well over two decades. It is true that our daily exposure to technical jargon is different—hers in corporate accounting and mine in academic linguistics—but otherwise we each live, work, and play with native English speakers in the northeastern United States in an English language environment which is by and large urban, educated, and professional. We see many of the same films (although she has probably seen more of them), read the same newspapers, and watch the same newscasts and commercials on TV. We listen to the same radio stations while driving. We read English-language fiction and nonfiction and occasionally exchange with each other those books that one or the other of us has particularly enjoyed. We browse the Internet and read and send e-mail. We subscribe to magazines. We each socialize at dinner parties and backyard barbecues, sometimes the same ones. We are both avid bargain hunters and not shy about negotiating prices with store clerks and managers and yard-sale hosts (although she is much better at it than I am). We both have undergraduate and graduate degrees from American institutions in the same northeastern U.S. metropolitan area. We read cookbook recipes, telephone bills, prescription labels, bank and credit card statements, menus, road maps, advertising and travel brochures, birthday cards, income tax forms, food and product labels, traffic signs, billboards, concert schedules, wedding invitations, and refund and cancellation policies, all in English. We each read a lot of memoranda at work and attend a lot of meetings to discuss the memoranda.

Despite these similarities, the language we both speak in this environment—English—in some ways sounds quite different from each other. Unlike me, Patty is a non-native speaker of English, acquiring it largely after her arrival in the United States more than 25 years ago at the age of 22. As with many second or foreign language (L2) learners whose language learning has commenced after childhood, Patty’s English contains phonetic, morphosyntactic, and lexical forms that in some ways sharply diverge from those produced rather more uniformly by the particular community of native English speakers with whom she interacts daily.¹ The nature and extent of this divergence, as well as the extent of overlap and similarity, are the subjects of this study.

Sorting Through Some Terminology

The apparent persistence of such divergent forms in Patty’s grammar, in the face of ongoing rich target-language input (presumably the same input that would suffice to bring a young child’s grammar into eventual conformity with target norms), is known as language fossilization (Selinker, 1972; Selinker & Lamendella, 1979). A brief example of a “fossilized” speech sample from Patty after she had already lived in the United States for about 10 years, taken from her conversation with two native English speakers, is shown in (1).

(1) ^[1]PAT: You know, I call B. this morning and nobody answer. And I start to worry.

L2]NS1: Do you have his phone number where he is?

^[3]PAT: Well, he either stay in Eliotville, because he said he call me last night, and he never did.

^[4]NS2: He said he’d call you?

^[5]NS1: He said he would?

^[6]PAT: Yeah, he would. In the morning, I talk to him from Eliotville, and then even if he stay over Eliotville, he will give me a call, you know?

^[7]NS1: Yeah.

[8]PAT: And then I call her house and nobody answer.

This sample highlights the variability with which certain obligatory grammatical elements are produced. Note, for example, the frequent omissions of verbal inflection and the error in pronominal gender, as illustrated earlier, as well as the attempt by both native speakers at lines [4] and [5] to clarify the intended meaning of Patty’s preceding utterance, which is missing a modal auxiliary.

The term fossilization is often distinguished from stabilization, the latter term generally used to indicate a more permeable, temporary knowledge state (in more common parlance, something like a “plateau”). For example, Selinker and Lakshmanan (1993) suggest that certain stabilized states can lead to either fossilization or further development. (I return to this suggestion in chap. 7.) Long (2003) points out several difficulties with the terminology and favors a research focus on stabilization in contrast to fossilization. He argues that fossilization as a theoretical construct cannot simultaneously encompass both synchronic and diachronic variability because such variability renders the claim of permanent divergence from nativelike norms essentially unfalsifiable. Moreover, he argues, the meaning of fossilization includes that of stabilization (with the difference being that fossilization implies permanence), and yet the dictionary definition of stabilization includes the descriptor “unfluctuating” (p. 489), interpreted by Long to mean disallowing variability.

I agree with Long that diachronic variation over a certain developmental span (i.e., the learner produces Form A at some developmental stage X but Form B at some later developmental stage Y) is inadmissible evidence of fossilization. However, synchronic variability, a hallmark of adult second language performance, is well known to persist at advanced proficiency levels and indeed, even at the steady, or end state, of acquisition.² To the degree that such variability diverges significantly from that exhibited by native speakers, it does not necessarily “bleed the construct [of fossilization] of any remaining meaning” (p. 510), as Long (2003) writes. Indeed, Long cites one view of variation as “fundamental to any postulated definition of fossilization” and that fossilization would have to include a kind of “head-banging variation” within a general context of cessation of development (p. 526). Putting aside the unflattering imagery, I agree that the status of (synchronic) variability in a so-called fossilized grammar is, as Long points out, a major unresolved issue. I focus in detail on the variability of a particular morphological inflection—namely, past-tense marking—in Patty’s production data in chapter 4, and I return to a more conceptual discussion of variability in chapter 7.

Regardless of the terminological choice between fossilization and stabilization to describe an end-state grammar such as Patty’s, what is even more important for my purposes here is to recognize that both terms implicitly entail a comparison with a target norm, itself a highly problematic notion. A target norm can only be a fairly crude shorthand way of referring to a set of hypothesized grammatical constructs of a language from the researcher’s perspective, not the learner’s. The notion of target is somewhat misleading in that it implies a teleological endpoint to acquisition, that is, that the learner will somehow (tacitly) “realize” when this goal has been attained such that further grammatical learning can now cease. In other words, the learner must know in advance precisely what has not yet been acquired and thus still needs to be learned.

This strikes me as implausible. It is certainly possible for a learner to have some metalinguistic understanding of some learning problem—I realize in some vague way, for example, that I have difficulties with the correct choice and placement of clitics in Italian and with tone sandhi in Chinese. But if I knew (exactly) what these should be in all relevant contexts, then I would, in fact, know them even if I could not always accurately control them in production. More worrisome, however, is the near-certain likelihood that there are many aspects of Italian and Chinese that I do not even know that I don’t know! Thus, it is impossible to speak in any meaningful sense of an ultimate nativelike target from the learner’s perspective, because that target is always moving, so to speak.

Similarly, the term interlanguage, implying some globally vague intermediate knowledge state “on the way” somewhere between knowledge of the L1 and ultimate mastery of the L2, shares some of the problems associated with the notion of target. In describing Patty’s end-state knowledge of English, which clearly differs in some respects from that of a native English speaker, I prefer the term idiolect. Although I occasionally use the conventional IL acronym in referring to her L2 grammar, I have something closer in meaning to idiolect in mind—her individual knowledge state (or I-language, in Chomsky’s terms) of English.³

Therefore, in this and the following chapters, although I use the terms (non-)targetlike and (non-)nativelike in describing Patty’s English production, it should nonetheless be kept in mind that such usage necessarily implies something along the lines of: ‘From the researcher’s perspective this is what appears to be the case in comparison to what we have observed or hypothesized about native speaker grammars.’ (In chap. 4, I discuss some possibly troublesome implications of researcher perspective with respect to lexical aspect in Patty’s data; see also Lardiere, 2003a.) Similarly, I continue to rather loosely use the term fossilization (pace Long, 2003) to describe those aspects of Patty’s grammar which appear to differ from the observed or hypothesized properties of native-speaker grammars (including in terms of variability or non-categoricalness), and for which there is no indication of further diachronic development over the observation period; however, Long’s terminological caveats should also be kept in mind.

Finally, a word about the term ultimate attainment, as contained in the title of this work. I use this phrase to refer to the state of knowledge actually attained at the end or steady state of grammatical development, which may or may not be nativelike. Over the course of the following chapters, it will become clear that Patty’s English idiolect—her state of ultimate attainment in that language—diverges from nativelike competence in some respects and completely converges in others.⁴

What Can a "Fossilized" End-State Grammar Tell Us About the Language Faculty?

The divergence in ultimate attainment from native speaker grammars we observe among adult language learners is theoretically interesting because it is (inversely) related to a well-known question concerning what type of linguistic information is extractable from the environment. Referring to this question as Plato’s Problem, Chomsky (1986, citing Bertrand Russell) asked: How can we know so much, given such limited evidence? Specifically, it has been amply demonstrated by now that (native) speakers of a language appear to end up with a much richer, more abstract system of knowledge than could possibly be induced from only the observable evidence in the environment—a problem known as the poverty of the stimulus.

To give a more concrete example, native speakers of English know that the sentence in (2), involving the displacement of a wh-element to sentence-initial position from its interpreted position as the subject of a relative clause, is ungrammatical, whereas the sentence in (3) involving similar wh-movement of the same phrase from the subject position of a subcategorized subordinate clause is perfectly fine:

• (2) *Which professor did Mary read the book on functional categories had written?
• (3) Which professor did Mary think had written the book on functional categories?

Given that sentences such as (3) occur in the input, what prevents an English learner from (over)generalizing the possibility of movement of the wh-phrase which professor to a sentence such as the one in (2)? How does an English speaker learn that the sentence in (2) is absolutely impossible? The answer given by Chomsky was essentially that such constraints on possible sentences are not learned; rather, they are a reflection of the way human brains are biologically predisposed to organize and represent linguistic knowledge. Such constraints are thought to collectively make up Universal Grammar (UG), and their precise formulation can be viewed as an attempt by generative linguists to model the human capacity for acquiring certain formal properties of possible natural language grammars. The principles of UG have been argued to make up (at least part of) a Language Acquisition Device (LAD), severely restricting the learner’s hypothesis space in the course of grammatical development.

In the case of fossilization, we might ask whether maturational changes in the LAD affect either the extractability of information from the input or the hypothesized constraints on grammatical acquisition, and whether such changes can account for the apparent inability to detect, compute, or otherwise internalize certain aspects of linguistic input that are (ultimately) perfectly accessible to very young children. In fact, assuming optimal exposure conditions of sufficient duration—a point to which I return below—non-native speakers immersed in the second language environment are exposed to the same linguistic input, and in this sense face the same poverty-of-the-stimulus problem as native speakers. One might well imagine that adult learners’ grammatical representations of the target language fail to be constrained by UG, thus leading to fossilization. However, in my view, this seems unlikely.

The UG-constrained characteristics of natural languages, such as hierarchical structure dependence, displacement (such as wh-movement), the autonomy of linguistic modules (e.g., phonetics, phonology, morphology, syntax, semantics), the presence of functional categories and features, and the subjecting of grammatical relations and operations to highly restrictive structural and locality conditions (e.g., c-command, subjacency, etc.)—in other words, the essential formats of linguistic representations within UG— appear to be largely intact in second language knowledge. This has been amply demonstrated (to the extent it can be) by a large number of studies investigating a wide range of formal aspects of language (e.g., Dekydtspotter, Sprouse, & Anderson, 1997; duPlessis, Solin, Travis, & White, 1987; Kanno, 1997; Martohardjono, 1993; Schwartz & Sprouse, 1994, 2000; Schwartz & Tomaselli, 1990; Thomas, 1991a, 1991b; Tomaselli & Schwartz, 1990; Vainikka & Young-Scholten, 1994; White & Genesee, 1996; and many others).

To take one example, Martohardjono (1993) tested native speakers of Chinese, Indonesian, or Italian learning English on their knowledge of the ungrammaticality of sentences such as the one shown earlier in (2) and repeated below as (4):

• (4) *Which professor did Mary read the book on functional categories had written?

The results indicated that, like native speakers, the non-native speakers in all three groups rejected more strongly ungrammatical sentences such as the one in (4) (i.e., “strong” violations), at a higher rate than somewhat more acceptable sentences involving more weakly constrained wh-movement (i.e., “weak” violations) such as the one in (5):

• (5) *?Which professor did Mary think that had written the book on functional categories?⁵

This distinction between strong and weak violations of constraints on wh-movement was theoretically motivated within a UG framework and thus predicted to occur. Although there were differences in the absolute percentages of rejection among the non-native groups (reflecting some lingering influence of their native L1s), and between these groups and the native speakers, Martohardjono (1993) showed that the relative rate of rejection between the types of sentences in (4) and (5) was statistically significantly different for all the non-native groups as well as for the native speakers, suggesting that all groups had knowledge of the relevant constraints.⁶ In chapter 5, I show that Patty appears to display similar knowledge of constraints on wh-movement despite the fact that such movement is not overtly evident in her native Chinese.

Returning to the nature of fossilized production, such as in the ...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• Preface
• 1. Some Preliminary Issues In Adult L2 Ultimate Attainment
• 2. Introducing Patty
• 3. Knowledge of Finiteness
• 4. The Acquisition of Past Tense
• 5. Clausal Word Order and Movement
• 6. Nominal Phrases
• 7. Conclusions
• Postscript
• References
• Author Index
• Subject Index