To what extent do you think innate linguistic knowledge is involved in second language learning?

. . . all human languages do share the same structure. More explicitly: they have essentially the same primitive elements and rules of composition . . . , although of course there may be variations, such as the obvious ones derived from the arbitrary association between sounds and meanings . . .

(Moro, 2016, p. 15)

3.1 Introduction

In this chapter, we start to consider individual theoretical perspectives on L2 learning in greater detail. Our first topic is the Universal Grammar (UG) approach (the generative linguistics approach), developed by the American linguist Noam Chomsky and numerous followers over the last few decades. This has been the most influential linguistic theory in the field, and has inspired a wealth of publications (for full-length treatments, see Hawkins, 2001; Herschensohn, 2000; Lardiere, 2007; Leung, 2009; Slabakova, 2016; Snape & Kupisch, 2016; Thomas, 2004; White, 2003; Whong, Gil, & Mars- den, 2013).

The main aim of linguistic theory is twofold: firstly, to characterize what human languages are like (descriptive adequacy), and, secondly, to explain why they are that way (explanatory adequacy). In terms of L2 acquisition, a linguistic approach sets out to describe the evolving language produced by L2 learners, and to explain its characteristics. UG is therefore a prop- erty theory (as defined in Chapter 1); that is, it attempts to characterize the underlying linguistic knowledge in L2 learners’ minds. In contrast, a detailed examination of the learning process itself (transition theory) will be the main concern of the cognitive approaches which we describe in Chapters 4 and 5.

First in this chapter, we will give a broad definition of the aims of the Chomskyan tradition in linguistic research, in order to identify the aspects of second language acquisition (SLA) to which this tradition is most relevant. Secondly, we will examine the concept of UG itself in some detail, and finally we will consider its application in L2 learning research.

3 Linguistics and Language Learning The Universal Grammar Approach

C o p y r i g h t 2 0 1 9 . R o u t l e d g e .

A l l r i g h t s r e s e r v e d . M a y n o t b e r e p r o d u c e d i n a n y f o r m w i t h o u t p e r m i s s i o n f r o m t h e p u b l i s h e r , e x c e p t f a i r u s e s p e r m i t t e d u n d e r U . S . o r a p p l i c a b l e c o p y r i g h t l a w .

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX AN: 2006630 ; Mitchell, Rosamond, Myles, Florence, Marsden, Emma.; Second Language Learning Theories : Fourth Edition Account: s9814295.main.ehost

82 Linguistics and Language Learning

3.2 Why a Universal Grammar?

3.2.1 Aims of Linguistic Research

The main goals of linguistic theory, as defined by Chomsky (1986a), are to answer three basic questions about human language:

1. What constitutes knowledge of language? 2. How is knowledge of language acquired? 3. How is knowledge of language put to use?

(“Knowledge of language” is an ambiguous term. Here, it means the subcon- scious mental representation of language which underpins all language use.)

All three questions are also of concern to SLA researchers. They can be briefly developed as follows:

1. What Constitutes Knowledge of Language?

From a generative perspective, linguistic theory aims to describe the rep- resentations of language which are stored in the human mind. It aims to define what all human languages have in common, as well as the distinc- tive characteristics which make human language different from other sys- tems of communication. It also needs to specify in what ways individual human languages can differ from one another, as they clearly do. However, Chomsky (e.g. 2000) argues that to a Martian landing on Earth, the dif- ferences between human languages would seem like variations on a single theme.

The UG approach views language as a genetic endowment. In the 1980s, this was conceptualized by claiming that all human beings inherit a universal set of abstract principles and parameters which constrain (limit) the shape human languages can take, and which make human languages similar to one another. In his Government and Binding theory, Chomsky (1981, 1986a, 1986b) argues that the core of human language must comprise these two components. His proposed principles are unvarying and apply to all natural languages (see Section 3.3.1). In contrast, parameters possess a limited num- ber of open values which characterize differences between languages (para- metric variation; see Section 3.3.2). In recent years, the notion of parameter has undergone major reconceptualization following Chomsky’s Minimalist Program (1995, 2000, 2005, 2007a, 2007b), in which he argues that the core of human language is the lexicon (the morpheme and word store). The lexi- con itself consists of two elements:

lexical categories (dog; give) Lexicon

functional categories (the; which)

 

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 83

We will define these elements in more detail later (in Section 3.3.1). Basi- cally, lexical categories include “content” words such as verbs and nouns, and functional categories include “grammatical” words such as determin- ers or auxiliaries, as well as abstract grammatical features such as tense or agreement.

In the Minimalist Program, parametric variation is located within the lexicon, primarily within functional categories. UG makes available a set of features (phonological; semantic; formal morphosyntactic), and languages select and combine these features differently, causing the various surface differences in word order, morphology and so on which we are familiar with. UG in this view is thought to consist of a universal computational system which specifies how words can combine to form syntactic units, and a universal inventory of features. These features are then assembled into lexical and functional items (Domínguez, 2013; Eguren, Fernandez- Soriano, & Mendikoetxea, 2016; Lardiere, 2012; Roberts, 2017b; Slabakova, 2016; Sorace, 2011; White, 2009b). Parametric variation in this view is due to different languages selecting different features from the universal set avail- able, and also bundling them differently within functional categories, giving rise to different form-function mappings. Within SLA theorizing, the earlier principles and parameters approach is gradually being replaced by this newer conceptualization of parametric variation being localized in the different ways in which languages map functional features onto functional categories. SLA involves remapping of these features as appropriate to the L2, which can be the source of considerable difficulty for the learner.

UG is a general theory of language, which should therefore apply to learner language, as one possible variety of language alongside other lan- guage varieties. In practice, one of the main interests of the UG approach for SLA research is that it provides a detailed descriptive framework which enables researchers to formulate well-defined hypotheses about the task facing the learner, and to analyse learner language in a more principled manner.

2. How Is Knowledge of Language Acquired?

Evidently each language is the result of the interplay of two factors: the initial state and the course of experience. We can think of the initial state as a “language acquisition device” that takes experience as “input” and gives the language as an “output”—an “output” that is internally represented in the mind/brain.

(Chomsky, 2000, p. 4)

How does the child create the mental construct that is language? Chomsky first resorted to the concept of UG because he believes that children could not learn their first language so quickly and effortlessly without the help of an innate language faculty to guide them. The so-called logical problem of language learning is that on the basis of highly complex and often messy

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

84 Linguistics and Language Learning

input (spoken language is full of false starts, unfinished sentences, slips of the tongue and so on), children create a mental representation of language which not only goes beyond the input they are exposed to (children use language creatively), but is also strikingly similar to that of other speakers of the same language variety. Children achieve this at an age when they have difficulty grasping abstract concepts, yet language is among the most abstract pieces of knowledge they will ever possess. If there is a biologically endowed UG, this would make the task facing children much easier, by providing a genetic blueprint which predetermines much of the shape which language will take.

If we now turn to L2 learning, learners are faced with the same logical problem of having to construct abstract mental representations of the new language on the basis of limited samples of language. But this does not mean to say that L2 learners necessarily set about the learning challenge in the same way as children. After all, their needs are very different, if only because they are already successful communicators in one language, and because they already have a mental representation of language, with the parameters set to the values of their L1. Moreover, L2 learners are cognitively mature and therefore presumably much better equipped to solve problems and to deal with abstract concepts. From a theoretical point of view, therefore, different possible scenarios are open to consideration:

• L2 grammars are fully constrained by UG (Full Functional Representation). The L2 is one example of a natural language, and it is constrained by UG in the same way as native grammars. Within this view, there is a range of different possibilities. For example, some researchers believe that L2 learners start off with the parameter settings of their L1, and reset them on the basis of input. Others believe that L2 learners have the full range of UG parameters available to them from the beginning, like L1 children, and do not resort to L1 parameter settings in the first instance. Others again believe that L2s gradually draw on UG, and that, e.g., functional categories are not available to learners at the beginning of the learning process but are acquired gradually. Within the Minimal- ist view of parameters as the various options available for selecting and mapping features onto lexical items, the task facing learners is seen as having to remap features and morphophonological forms appropriately onto the L2. All these approaches believe that the L2 grammar is con- strained by UG and can (but does not necessarily) become native-like (Hawkins, 2008a, 2008b; Lardiere, 2000, 2007, 2009, 2012; Rothman & Slabakova, 2018; Schwartz & Sprouse, 1996; Slabakova, 2016; Vainikka & Young-Scholten, 1996a; White, 2003, 2009a, 2009b).

• UG does not constrain L2 grammars OR UG is impaired (Rep- resentational Deficit). Some researchers believe that L2 grammars are fundamentally different from L1 grammars because they are not

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 85

constrained any longer by UG, and learners have to resort to general learning mechanisms (Bley-Vroman, 1989), potentially giving rise to “wild” grammars, i.e. grammars which do not necessarily conform to the general rules underlying natural human languages (e.g. Mei- sel, 1997), although that view has not received much support recently (Schwartz & Sprouse, 2017). Other researchers believe that only the parameters instantiated (activated) in the learners’ L1 will be available, and that parameter resetting is impossible; for example, features which have not been selected in the L1 are no longer available (grammati- cal gender in genderless languages for example). Within this view, the L2 grammar does not violate UG principles and parameters (it is not “wild”), but it can never become identical with that of L1 speakers of the same language (Hawkins, 2001, 2009; Hawkins & Chan, 1997; Hawkins & Franceschina, 2004; Tsimpli & Dimitrakopoulou, 2007).

There is considerable controversy around all these issues, and we revisit them in Section 3.5.

3. How Is Knowledge of Language Put to Use?

The UG approach is concerned with the abstract mental representation of language and the computational mechanisms associated with it which all human beings possess. This mental representation is commonly called competence. It is not about performance, i.e. about how language is used in real life. Performance requires a theory of language use, in which lin- guistic competence is only one aspect, and psycholinguistic factors such as the brain’s information processing systems also come into play, as well as domain-general cognitive variables and sociocultural factors. Although Chomsky acknowledges that this is an important area for research, he has been concerned almost exclusively with exploring linguistic competence, and this has also been true for UG-inspired research in SLA. Recently, how- ever, generative linguists have been increasingly concerned with how dif- ferent language modules (syntax, phonology, semantics) interface with one another (Domínguez, 2013; Serratrice, Sorace, & Paoli, 2004; Sorace & Ser- ratrice, 2009), and with non-linguistic modules such as the sensory-motor system (SM—involved in perception and production of language) and the conceptual-intentional system (CI—involved in thought and information structure), as well as the processing system (Sharwood Smith, 2017: see dis- cussion in Chapter 10). In turn, these new foci, as well as the new methodol- ogies available to researchers to investigate language processing in real time, have led to somewhat increased attention to the domain of performance and what it can tell us about competence (Chomsky, 2005, 2007b; de Villiers & Roeper, 2011; Di Sciullo et al., 2010; Lardiere, 2012; Rothman, Bayram, Cunnings, & González Alonso, in press; Sharwood Smith, 2017; Tsimpli & Sorace, 2006; White, 2009a). Additionally, the role of the input in providing

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

86 Linguistics and Language Learning

sufficient evidence and clear cues about certain properties is increasingly seen as crucial within generative SLA (Rankin & Unsworth, 2016; Roth- man & Slabakova, 2018).

3.2.2 Arguments for UG from L1 Acquisition

In this section, we will review in more detail the arguments which sup- port the existence of an innate language faculty in children. We will start from the characteristics of L1 acquisition outlined in Chapter 2, and sum- marized here:

1. Children go through developmental stages; 2. These stages are very similar among children learning the same lan-

guage, although individual children’s rate of progress is variable; 3. These stages are similar across languages; 4. Child language is rule-governed and systematic, though the rules cre-

ated by the child often do not correspond to adult ones; 5. Children are resistant to correction; 6. Children’s processing capacity limits the number of rules they can apply

at any one time, and they will revert to earlier hypotheses when two or more rules compete.

Universalists could not conclude from such evidence alone that there must be a specific language module in the brain. These regularities, although very striking, could be due to the more general cognitive makeup of human beings. After all, children learning maths or learning to play the piano also go through fairly well-defined stages, although not at such a young age and not necessarily so successfully, and they need to be taught these skills, unlike language. Indeed, many non-UG researchers believe that language acquisi- tion uses the same kind of information processing as other kinds of learning, as we will see in Chapters 4 and 5.

However, another striking feature of child language is that it does not seem to be linked in any clear way to intelligence. Children do vary in the age and speed at which they go through each developmental step. By age 4 or 5, though, individual differences have largely disappeared, and the late starter has usually caught up with the precocious child.

Not only is language development not directly linked to intelligence, but it also involves mastering a highly complex and abstract kind of knowledge. To give an example of the complexities which children have to disentangle, consider the following English reflexive sentences, some of them grammati- cal and others ungrammatical:

a. John saw himself. b. *Himself saw John. c. Looking after himself bores John.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 87

d. John said that Fred liked himself. e. *John said that Fred liked himself. f. John told Bill to wash himself. g. *John told Bill to wash himself. h. John promised Bill to wash himself. i. John believes himself to be intelligent. j. *John believes that himself is intelligent. k. John showed Bill a picture of himself.

(Examples are taken from White, 1989, cited in Lightbown & Spada, 2013, pp. 20–21. In all these sentences, italics are used to indicate the noun and the pronoun which refer to the same person; asterisks indicate ungrammaticality— i.e. the two words in italics cannot co-refer.)

Now imagine you are the child trying to work out the relationship between the reflexive pronoun and its antecedent. You might conclude from (a) and (b) that the reflexive pronoun must follow the noun it refers to, but (c) disproves this. Sentences (d), (e), (f) and (g) might lead you to believe that the closest noun is the antecedent, but (h) shows that this cannot be right either. It is also evident from (h) that the reflexive and its antecedent do not have to be in the same clause. Furthermore, the reflexive can be in subject position in (i), an untensed clause, but not in (j), a tensed clause. Moreover, the reflexive can sometimes have two possible antecedents, as in (k), where himself can refer to either John or Bill.

These few sentences illustrate the magnitude of the task facing children; how can they make sense of data like this, and invariably arrive at the correct rule?

In support of the view that language is not linked to intelligence, there is also relevant evidence from children with cognitive deficits who nonethe- less develop language normally (Bartke & Siegmuller, 2004; Bishop, 2001; Bishop & Mogford, 1993). For example, Bellugi, Van Hoek, Lillo-Martin, and O’Grady (1993) studied children suffering from Williams syndrome, a rare metabolic disorder which causes mental impairment, as well as other physical symptoms. They demonstrated that individuals whose overall mental develop- ment is otherwise very slow and remains below that of a 7-year-old can make sophisticated use of language with complex syntax and adult-like vocabulary.

Smith and Tsimpli (1995; Smith, Tsimpli, Morgan, & Woll, 2011) have studied in detail the exceptional case of a brain-damaged man, Christopher, who is institutionalized because he is unable to look after himself, but who can read, write and communicate in many languages:

The most salient feature is a striking mismatch between his verbal and non-verbal abilities, supported by test results over a prolonged period and with recent documentation across a wide range of different tests. The basic generalization is that he combines a relatively low perfor- mance IQ with an average or above average verbal IQ.

(Smith & Tsimpli, 1995, p. 4)

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

88 Linguistics and Language Learning

Evidence of the opposite is also found: children who are cognitively “nor- mal”, but whose language is impaired, sometimes severely. This condition, known as Specific Language Impairment (SLI), is characterized by “(persis- tent) difficulties with the productive rules of word-formation, the morpho- syntactic prerequisites of feature agreement and the construction of complex phonological units” (Lorenzo & Longa, 2003, p. 645; see also van der Lely, 1998; J. Paradis, Rice, Crago, & Marquis, 2008). One English-speaking fam- ily has been studied, in which 16 out of 30 members in the last three gen- erations suffer from SLI, suggesting that it is an inherited disorder, and that some aspects of language at least might be genetically controlled (Gopnik & Crago, 1991). Recently, the gene FOXP2 has been discovered, whose muta- tion apparently leads to SLI (Lai, Fisher, Hurst, Vargha-Khadem, & Faraneh, 2001; Pinel et al., 2012; Reuter et al., 2017).

The picture we have just outlined of the relationship between brain and language is necessarily very oversimplified. (For more detailed accounts, see, for example, Friederici, 2017; Jenkins, 2000; Lorenzo & Longa, 2003.) From such evidence, it has been concluded that specific areas of the brain deal with specific aspects of language. Recent advances in brain-imaging tech- niques have also supported this view, although the picture is becoming more complex as techniques become more sophisticated (Carter, 1998; Dörnyei, 2009; M. Paradis, 2004).

Evidence such as this has been used by universalists to propose that there must be some kind of innate language faculty which is biologically trig- gered, in order to explain why language in children just seems to “grow”. An influential book by Lenneberg (1967) called The Biological Foundations of Language outlined the characteristics which are typical of biologically trig- gered behaviour (such as learning to walk), and argued that language con- forms to these.

Aitchison (2008, p. 71) presents Lenneberg’s criteria as a list of six features:

1. “The behaviour emerges before it is necessary”. Children start talking long before they need to: they are still being fed and looked after, and therefore do not need language for survival.

2. “Its appearance is not the result of a conscious decision”. It is quite obvious that children do not get up one morning and decide to start talking, whereas they might consciously decide to learn to ride a bike or play the piano.

3. “Its emergence is not triggered by external events (though the sur- rounding environment must be sufficiently ‘rich’ for it to develop ade- quately)”. Although children need language around them in order to learn it, there is no single event which will suddenly trigger language development.

4. “Direct teaching and intensive practice have relatively little effect”. We have seen in Chapter 2 how oblivious children seem to be to correction.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 89

5. “There is a regular sequence of ‘milestones’ as the behaviour develops, and these can usually be correlated with age and other aspects of devel- opment”. Just as a baby will sit up before standing up before walking before running, we have seen how children go through well-defined stages in their language development, which tend to run parallel to physical development. The onset of the first words usually roughly cor- responds to the onset of walking for example.

6. “There may be a ‘critical period’ for the acquisition of the behaviour”. It is often argued that for normal development, human beings have to be exposed to language before puberty. This is a controversial issue; the evi- dence from children who have been deprived of language in their early years is difficult to interpret, as it is not usually known whether they might have been born with other disabilities (Curtiss, 1977; Eubank & Gregg, 1999; Smith, 1999). We examine in Section 3.5.3 the evidence that adult L2 learners bring to this ongoing debate.

After having reviewed the kind of argumentation used by universalists in order to propose the existence of a language-specific module in the brain, we now turn to the question of what this so-called language faculty or UG might be like.

3.3 What Does UG Consist Of?

The aim of this section is to illustrate UG sufficiently to understand how it has been applied to the study of language acquisition. Generative linguistics is very complex and has changed considerably in the last 60 years or so, from the early phase of phrase structure rules to the more recent Minimalist Program. How- ever, its primary goal has remained the same, that is, to characterize the innate language faculty: “The problem that has virtually defined the serious study of language since its ancient origins, if only implicitly, is to identify the specific nature of this distinctive human possession” (Chomsky, 2007a, p. 1). The vary- ing emphases over the years have essentially resulted from tension between “descriptive” and “explanatory” adequacy. In the case of UG, the search for descriptive adequacy has attempted to account for the details of increasing numbers of typologically unrelated languages, while the search for explanatory adequacy has aimed to make effective cross-language generalizations:

A theory of language must show how each particular language can be derived from a uniform initial state under the “boundary conditions” set by experience . . . The search for descriptive adequacy seems to lead to ever greater complexity and variety of rule systems, while the search for explanatory adequacy requires that language structure must be invariant, except at the margins.

(Chomsky, 2000, p. 7)

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

90 Linguistics and Language Learning

In recent accounts, three factors are argued to contribute to language design (Berwick & Chomsky, 2016; Chomsky, 2005; Huang & Roberts, 2017; Yang, Crain, Berwick, Chomsky, & Bolhuis, 2017):

• Genetic endowment (UG); • Experience (PLD—Primary Linguistic Data); • Principles not specific to the faculty of language (located in general

cognitive systems such as logical reasoning and memory, computational efficiency, minimality and general laws of nature).

Whatever exact description of UG is entertained, the general concept remains the same: “UG will be the smallest amount of innate structure needed to explain outcome competence that ultimately cannot be reduced to deduction from the input and domain-general cognition alone” (Roth- man et al., in press, p. 5).

Next, we will examine more concretely what UG consists of, that is, what might be invariant in human language (the principles of the Principles and Parameters model; the inventory of features and the computational system in the Minimalist Program), and what varies across languages (the parameters of Principles and Parameters and, in the Minimalist Program, the parametric variation linked to feature selection and feature mapping onto functional categories). We consider both the earlier Principles and Parameters model and the later Minimalist Program, as both remain active points of reference in L2 research.

3.3.1 What Is Invariant in Human Language?

We have seen earlier that, according to early generative views, the L1 learner’s initial state consists of a set of universal language principles. Individual lan- guages vary in limited ways, expressible in terms of innate parameters which need to be selected on the basis of exposure to language evidence (Lardiere, 2012; Slabakova, 2016; White, 2009a). In current Minimalist thinking, the genetic endowment for languages is thought to consist only of a closed inventory of universal features which languages select from and assemble in different ways, and the unique structure-building operation Merge, which we will exemplify (Eguren et al., 2016; Roberts, 2017a).

From the point of view of acquisition, the general idea is that language learning is highly constrained in advance, thus making the task for the child much more manageable. In this section, we will first of all examine one concrete example of a principle common to all languages. Next, we will consider the array of lexical and functional categories which UG also makes available, and how the Minimalist idea of universal features is relevant to language learning (de Villiers & Roeper, 2011; Meisel, 2011).

Our first example is the well-known principle of structure dependency, which states that language organization depends on the structural relation- ships between elements in a sentence, which are combined in a binary

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 91

fashion through a single operation, Merge. That is to say, individual words and morphemes are regrouped into higher level abstract structures which are the building blocks of language. Intuitively, we know that this is the case. In the following sentences,

a. She bought a new car yesterday b. My friend bought a new car yesterday c. The friend that I met in Australia last year bought a new car yesterday d. The friend I am closest to and who was so supportive when I lost my job two

years ago bought a new car yesterday.

we know that she, my friend, the friend that I met in Australia last year, and the friend I am closest to and who was so supportive when I lost my job two years ago are groupings which play the same role in the sentence, and in fact might refer to one single individual. Moreover, we also know that we could carry on adding details about this friend more or less indefinitely by using devices such as and, that, which etc., running the risk of boring our listener to tears! We also know that the crucial word in these groupings is friend, or she if we have already referred to this person earlier in the conversation. This kind of grouping is called a Phrase; in the examples above, we are dealing with a Noun Phrase, as the main or central element (the head) of this phrase is a noun (or pronoun). It turns out that all languages in the world are structured in this way, and are made up of sentences which consist of at least a Noun Phrase (NP) and a Verb Phrase (VP), as in [

NP Paul][

VP sings], which in turn may optionally con-

tain other phrases or even whole sentences, as in (d) above. This knowledge—that languages are structure-dependent—is a crucial

aspect of all human languages which has many implications; it is a principle or invariant feature of UG which explains many of the operations we rou- tinely perform on language. For example, when we ask a question in English, we change the basic order of the sentence:

Your cat is friendly Subject Verb Complement

Is your cat friendly?

Verb Subject Complement

This movement is not based on the linear order of the sentence, but is structure-dependent. We do not move the first verb we encounter, or, say, the third word in the sentence, rules which would work in the above example, but would generate ungrammatical sentences in the following example:

The cat who is friendly is ginger *Is the cat who friendly is ginger? *Who the cat is friendly is ginger?



EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

92 Linguistics and Language Learning

The correct question here is of course Is the cat who is friendly ginger?, where the second is is moved to the beginning of the sentence. Note that there is no immediately obvious reason why this should be the case; computers would have no problems dealing with either of the two artificial rules above. In fact, computers find it considerably more difficult to apply a rule which is based on a hierarchical structure, as is the case in this natural language example.

To take another example, the same restrictions apply to passive sentences. The sentence The car hit the girl can be made into a passive by raising the object NP to the subject position: The girl was hit by the car. Notice that it is the whole NP which is moved to the front; it could just as well have been Lisa, or The girl with the blue trousers, or The girl who won first prize in the crea- tive writing competition. French passive constructions work in exactly the same way: L’enfant chatouille le nounours (= the child tickles the teddy) becoming Le nounours est chatouillé par l’enfant (= the teddy is tickled by the child). The general claim can be made that:

structure-dependency can therefore be put forward as a universal prin- ciple of language: whenever elements of the sentence are moved to form passives, questions, or whatever, such movement takes account of the structural relationships of the sentence rather than the linear order of words.

(Cook & Newson, 1996, p. 11)

The movement we have just described, underlying the formation of inter- rogative and passive sentences, is a very common feature of many languages, called Move α.

Finally, according to this theory, the lexical and functional categories used in language also form part of our UG endowment, and do not have to be learned. UG includes a universal inventory of categories and features which the child selects from on the basis of the input, as not all languages will nec- essarily make use of all categories or their features.

The task for L2 learners, then, is to identify which syntactic categories are required in the L2, where the selection might be different from that of their L1. For example, some syntacticians argue that Japanese lacks the functional category Det (Determiner) (Fukui & Speas, 1986). The features associated with syntactic categories might also be different; for example, the French determiner phrase has a grammatical gender feature, whereas the English determiner phrase does not, as in la (feminine) chaise, “the chair”; le (mascu- line) fauteuil, “the armchair”. The way in which features are “assembled” or packaged to form functional and lexical categories can vary from language to language. Thus, the French determiner phrase has gender and number features, while the German determiner phrase has gender, number and case features. Feature values can also vary: for instance, Infl (Inflection) is said to be “strong” in French and “weak” in English, which leads to different word

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 93

orders in these two languages in the context of negative structures, interrog- atives and adverb placement (White, 2003, p. 10). The task of the L2 learner is to “identify, select, and redistribute the required features among the lexical items of the L2” (Lardiere, 2012, p. 110).

We will come back to these sources of variation shortly. Before doing so, let us define in more detail what is meant by these functional categories.

What we call lexical categories are groups such as nouns, verbs or adjec- tives, that is, so-called content words that carry a specific meaning. The kind of items we are now turning our attention to are grammatical words or “function” words, such as determiners (the, my) and complementizers (whether), or grammatical morphemes such as plural -s, past tense -ed and so on. Another way of conceptualizing the difference between lexical and functional categories is in terms of an open class of language items, and a closed class. An open class (a lexical category) is one to which you can add new items quite freely; for example, in the lexical categories Noun or Verb, words such as selfie or fracking are being added all the time. A closed class (a functional category) is one to which items cannot easily be added. For example, you cannot add new determiners or new past tense morphemes to a language, in the straightforward way in which you can add new nouns, verbs or adjectives.

In itself, this distinction between content words and functional items is not new to linguistics. However, generative theory claims that these “functional” items, whether words or morphemes, also have phrases attached to them in the same way as “lexical” words do. These functional phrases are organized in the same way as any other phrase, with the function word or morpheme as head of that phrase. We will therefore have Determiner Phrases (DP) and Complementizer Phrases (CP), with determiners such as the or complemen- tizers such as whether as their heads, and also Inflection Phrases (IP) made up of Tense Phrases (TP) and Agreement Phrases (AgrP), which carry tense and agreement markers such as past tense -ed or 3rd person singular -s in English. The structure of these functional phrases is basically the same as that of lexical phrases, such as NPs, and they can be represented in the same way.

3.3.2 Parameters and Language Variation

The structure dependency principle which we discussed earlier is common to all languages, so they are all organized hierarchically in terms of phrases (Noun Phrases, Verb Phrases, Prepositional Phrases and so on). From a UG perspective, such a principle would form part of the innate computational module and will therefore not have to be learned. However, we also know that not all languages have identical structural properties. This is where parameters come in. Parametric variation across languages is due to the fact that languages select different sets of features out of the universal inventory provided by UG, and also bundle them differently onto functional phrases. These features can reflect grammatical meanings (for example tense, case,

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

94 Linguistics and Language Learning

finiteness or grammatical gender) or conceptual meanings (for example evi- dentiality, habituality, definiteness or animacy). The features that contribute to the meaning of the sentence, for example tense or definiteness, are called “interpretable features”; those that do not contribute to meaning but have a grammatical role only, such as grammatical gender or case, are termed “unin- terpretable features” (Rothman & Slabakova, 2018).

Universal Grammar provides a common store of atomic elements of meaning (such as semantic and grammatical features); languages select and assemble these features into lexical items (LIs). Now, one language may choose to assemble one array of these features into one L1, while another language may assemble a slightly different array into what superficially looks like an equivalent lexical item. A lexical item is noth- ing but a bundle of features, including phonological, grammatical and semantic features.

(Slabakova, 2016, p. 43)

In the Minimalist Program (Berwick & Chomsky, 2016; Chomsky, 1995, 2000, 2002, 2005, 2007a, 2007b, 2008, 2013), Chomsky suggests that the lan- guage faculty consists of a computational procedure, sometimes called “nar- row syntax”, which contains operations such as Merge (the operation by which two syntactic objects are combined to form a new syntactic unit, for example a verb and its complement combining to form a Verb Phrase), Move α (described above) or Agree (when two items have to match for features such as number or gender). This narrow syntax is virtually invariant across languages, and the lexicon is the site of parametric variation. The functional lexicon in particular will contain features dictating, for example, whether movement or agreement is required. In this view, languages are different from one another only because their lexicons are different (Chomsky, 2005; Domínguez, 2013; Eguren et al., 2016; Lardiere, 2012; Roberts, 2017b).The task facing children (or L2 learners) is therefore to learn the lexicon of the language around them, including the settings of the parameters applying to that language. This is the “lexical parameterization hypothesis”, which sug- gests that parameters are contained in the lexicon, primarily attaching to the functional categories. For example, the functional category Agr, which governs agreement phenomena, contains a gender feature in languages such as French or Italian, but not in others such as English.

Let us now turn our attention to an often-discussed example of a param- eter, which was proposed to account for one of the more obvious ways in which languages can vary: the so-called head parameter (Chomsky, 1981). (For more detailed analyses, see, for example, Cook & Newson, 2007; Eguren et al., 2016; Hawkins, 2001; Herschensohn, 2000; Huang & Roberts, 2017; Slabakova, 2016).

The head parameter deals with the way in which phrases themselves are structured. It applies to phrases headed by both lexical and functional

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 95

categories. Each phrase has a central element, called a head; in the case of a Noun Phrase, the head is the noun; in the case of a Verb Phrase, it is the verb; in the case of a Determiner Phrase, it is the determiner; and so on. One dimension along which languages vary is the position of the head in relation to other elements inside the phrase, called complements. For example, in the Noun Phrase (the) girl with blue trousers, the head noun girl appears to the left of the complement with blue trousers; in the Verb Phrase hit the girl, the head hit appears to the left of its complement the girl; similarly, in the Prepositional Phrase with blue trousers, the head with is on the left of its complement blue trousers; in the Complementizer Phrase whether he is too old, the complement he is too old follows the head whether. In fact, English is a head-first language, because the head of the phrase normally appears before its complements.

Japanese, on the other hand, is a head-last language, because the comple- ments precede the head within the phrase:

E wa kabe ni kakatte imasu (picture wall on is hanging) “The picture is hanging on the wall”.

The head verb kakatte imasu occurs on the right of the verb complement kabe ni, and the postposition ni (on) comes on the right of the PP comple- ment kabe.

(Cook & Newson, 2007, pp. 42–43)

Because of its head-last setting, all Japanese phrases will be ordered in that way. So, the head parameter tells us how the head and its complements are ordered in relation to one another in a given language, and it has two pos- sible settings: head-first (like English) or head-last (like Japanese). (Some languages are not unidirectional, but these are rare: Eguren et al., 2016.)

From an acquisitional point of view, what this means is that children do not need to discover that language is structured hierarchically into phrases, as this principle forms part of the UG blueprint for language in their mind (Guasti, 2017). They also assume that all phrases in the language they are learning are going to behave in the same way, a phenomenon which has been called the Uniformity Principle (Boeckx, 2011), Harmonic Systems (Baker, 2008) or Generalization of the Input (Roberts, 2012, 2016; Rob- erts & Holmberg, 2010). These terms allude in this case to learners’ expecta- tions that a value assigned to one head will apply to all heads, unless there is contrary evidence in the input (Eguren et al., 2016, p. 21). The only task remaining is to learn which parameter setting actually applies in the lan- guage which the child is learning: in this case, is it head-first or head-last? In theory, the only input the child needs in order to set the head parameter to the correct value is one example of one phrase, and they will then automati- cally “know” the internal structure of all other phrases.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

96 Linguistics and Language Learning

In this view, the task facing children is considerably simpler than if they had to discover the order of constituents within each type of phrase. Moreo- ver, they only need minimal exposure in order to make wide-ranging gen- eralizations which affect different parts of the syntax of the language they are learning. For example, Radford claims:

young children acquiring English as their native language seem to set the head parameter at its appropriate head-first setting from the very earliest multiword utterances they produce (at around age 18 months), and seem to know (tacitly, not explicitly, of course) that English is a head-first language.

(1997, p. 22)

Remember also the puzzle which we posed in Section 3.2.2, when we asked how children could work out the precise relationships which apply between reflexives such as himself and their NP antecedents, in English? The answer offered by UG theory to this problem is another classic UG parameter, the “Governing Category” parameter. This parameter determines which “bind- ing domains” are possible in a given language, that is to say, what elements in the sentence can co-refer across which structural boundaries. For example, in English, reflexive pronouns must be bound within a local domain (such as a clause), which means that in the sentence Mark wanted Tom to treat him- self, “himself ” can only refer to Tom and not to Mark. In other languages which allow long distance binding (across clause boundaries), such as Chi- nese, “himself ” could refer to either Tom or Mark. Once again, the assump- tion is that the relevant Binding principles, and the associated Governing Category parameter, are already pre-existing in the child’s language module. So, acquiring a highly complex area of grammar is reduced to the simple matter of picking the appropriate binding domain out of a restricted set of possibilities (for further details, see, for example, Hawkins, 2001; Herschen- sohn, 2000).

Let us now illustrate parametric variation for a functional category, Inflec- tion (Infl, or I). Inflection is the functional category which contains the tense and agreement features of verbs (tense, person, number: whence its name, as these features are often realized through an inflectional paradigm). Just as nouns and verbs can head NPs and VPs, Infl can also head an Inflectional Phrase (IP). Features associated with functional categories can have either weak or strong values, with implications for syntactic properties of that lan- guage. For example, Infl in English is weak, whereas in French it is strong. This parametric variation (+/- strong) has consequences for word order. In languages like French, finite verbs have to move to the I position for feature checking (i.e. to “collect” their tense, number and agreement features), result- ing in the verb preceding the adverb. In English, however, where Infl is weak, the verb remains in the VP. This is illustrated in the tree diagrams that follow:

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 97

This parametric variation in feature strength has important consequences for other areas of grammar, and explains a number of word order differences between French and English, which otherwise have very similar structures. These differences are summarized here:

English French

Declaratives Patrick reads the newspaper S V O

Patrick lit le journal S V O

Adverb placement

Patrick often reads the newspaper S A V O

Patrick lit souvent le journal S V A O

Negation Patrick doesn’t read the newspaper S neg V O

Patrick (ne) lit pas le journal S V neg O

Questions Does he read the newspaper? S V O

Lit-il le journal? V S O (pronominal subjects only)

Within this view of learning, all learners should have to do is set the parameter (called the Verb Movement parameter for obvious reasons) to either weak or strong, on the basis of the input (French or English), and all these properties should be in place. Empirical evidence, however, shows that L2 learners do not acquire them all at the same time (White, 1991; Trahey & White, 1993). (For fuller treatments, see, e.g., Hawkins, 2001; Herschensohn, 2000; White, 2003.)

But let us now turn specifically to the way in which UG explains lan- guage acquisition data.

English French IP IP

Spec I’ Spec I’

David David I VP I VP

joue Adv V’ Adv V’

always toujours V NP V PP

plays football ti P NP

au football

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

98 Linguistics and Language Learning

It is claimed in generative syntax that the same underlying structural con- figuration of head, complement and specifier usually applies to all phrases in a given language. The following examples show how this works in English for the Verb Phrase (Diagram B), the Adjectival Phrase (Diagram C) and the Prepositional Phrase (Diagram D):

Diagram B: [ NP

My brother] [ VP

regularly wins the first prize]

3.4 UG and L1 Acquisition

So, what is the evidence in the child acquisition literature for the UG view- point? Do children indeed build grammars guided by UG, as described above?

Before we can deal with this question, we need to first examine in more detail the structure of phrases (see Hawkins, 2001, pp. 13–16). We have seen already that all languages are made up of phrases consisting of a head cat- egory (the core element of the phrase) and of complements which option- ally modify the head. Another type of modifier—also optional—is called a specifier, as shown in the example of an English Noun Phrase (Diagram A). Here the head noun holiday is modified by its complement in the Caribbean Islands, and the grouping holiday in the Caribbean Islands is itself modified by the specifier my mother’s.

Diagram A: [ NP

my mother’s holiday in the Caribbean Islands] [ VP

was fantastic]

NP

Specifier N’

my mother's N0 Complement

holiday in the Caribbean Islands

VP

Specifier V’

regularly V0 Complement

wins the first prize

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 99

Diagram D: [ VP

He did this] [ PP

quite without reason]

Diagram C: [ VP

She became] [ AP

incredibly clever at making excuses]

AP

Specifier A’

incredibly A0 Complement

clever at making excuses

PP

Specifier P’

quite P0 Complement

without reason

XP

Specifier X’

X0 Complement

All phrases are organized in this hierarchical manner, with an optional speci- fier modifying an X′, itself consisting of an X0 (the head) modified by an optional complement, where X can be any of the head categories: N0 (noun), V0 (verb), A0 (adjective), P0 (preposition), D0 (determiner), Infl0 (inflection). (The notation X′, X0 is used to indicate the different levels in the hierarchi- cal structure of phrases, with X0 representing the head element on its own, X′ representing the unit “head element + complement” and so on.) The only possible variant is the situation of head, specifier and complement in relation to one another. Thus, in a language such as English, the general con- figuration illustrated in Diagrams A–D above can be summed up as follows:

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

100 Linguistics and Language Learning

Following this pattern, a literal translation of the examples given above would be my mother’s in the Caribbean Islands holiday, incredibly at making excuses clever and quite reason without.

A third possible ordering also found in natural languages comprises the head followed by both complement and specifier:

That is, in all types of phrase in English, the specifier typically precedes the head element, and the complement follows it. However, in languages such as Japanese, Turkish and Burmese, both specifier and complement precede the head (Hawkins, 2001, p. 15):

XP

Specifier X’

Complement X0

XP

X’ Specifier

X0 Complement

This would give rise to the following reordering of our examples: holiday in the Caribbean Islands my mother’s, clever at making excuses incredibly and with- out reason quite. This configuration is found in languages such as Malagasy, Gilbertese and Fijian (Hawkins, 2001, p. 15).

In terms of L1 acquisition, what are the implications? We have noted that the structure of phrases is an invariant principle of UG, part of the narrow syntax, the universal computational system containing operations such as Merge. Children would therefore “know” that sentences are made of phrases which consist of (specifier)—head—(complement), and would not have to work this out. However, they would not know the precise ordering of these elements which is found in their own language; that is, they would have to “set” the head parameter on the basis of language input. Notice, though, that the number of possibilities is very limited; specifiers either precede or fol- low X′ categories, and complements either follow or precede X0 categories (Hawkins, 2001, p. 16).

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 101

There is evidence from L1 acquisition research that children have set the head parameter as early as the two-word stage (Radford, 1997, p. 22), and that they “know how to project productively X0 categories into X′ categories, and X′ categories into XP categories” (Towell & Hawkins, 1994, p. 65), at least as far as lexical categories are concerned. This is shown in the examples that follow (from Radford, 1990, cited in Towell & Hawkins, 1994, p. 66):

X0 Complement cup tea (N′) “a cup of tea” ball wool (N′) “a ball of wool” open box (V′) “open the box” get toys (V′) “get my toys”

(put) in there (P′) “put it in there” (get) out cot (P′) “I want to get out of the cot”

Specifier X′ Mummy car (NP) “Mummy’s car” Hayley dress (NP) “Hayley’s dress” Dolly hat (NP) “Dolly’s hat” Daddy gone (VP) “Daddy has gone” Hayley draw (boat) (VP) “Hayley is drawing (a boat)” Paula play (with ball) (VP) “Paula is playing (with a ball)”

UG theory would predict these findings, as the result of the general prin- ciple underlying phrase structure. However, it would also predict that chil- dren have to learn the particular parameter settings for the language they are exposed to on the basis of language input; these are not “inbuilt” (Guasti, 2017; Lidz & Gagliardi, 2015).

There is also evidence that the kind of parametric variation found in functional features which we have illustrated earlier (strong or weak Infl) is acquired by children in a cluster-like fashion. That is to say, when L1 French children start to use inflected verbs (or in formal terms, when they project the Inflection Phrase), all properties linked to the “strong Infl” parameter setting fall into place. Notably, the verb rises to I, past adverbs, negators and so on. We find such children producing sentences such as Pas aller dodo (“no go bed”), in which IP has not yet been projected and the verb is there- fore non-finite and has not moved from its VP-internal position. They may almost simultaneously produce sentences in which IP has been projected and the verb is therefore finite and has risen past elements like negators, such as bébé va pas dodo (“baby goes not bed”). What we do not find, however, are sentences in which non-finite verbs have risen past adverbs or negators, e.g. *bébé aller pas dodo (“baby go not bed”), or in which the finite verb does not rise, e.g. *bébé pas va dodo (“baby not goes bed”) (Pierce, 1992).

Guasti (2016, p. 189), reviewing the available evidence, claims that “from the earliest syntactic productions there is evidence that children have assigned

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

102 Linguistics and Language Learning

the correct value to the parameters that govern clausal structure (head direc- tion parameter, verb movement parameter, V2 parameter)”. (In languages like German or Dutch, the “V2 parameter” specifies that the finite verb must appear in second position.) Overall, the notion of parameters remains an important tool for generativist L1 acquisition researchers.

This simplified account has illustrated the kinds of predictions a UG approach can offer, in the context of children acquiring their L1. Controver- sies remain, not least regarding how soon functional categories are acquired. For a discussion of these issues, and for wider accounts of the UG approach to L1 acquisition, see Guasti (2009, 2016), Snyder (2007), Ambridge and Lieven (2011), Meisel (2011) and de Villiers and Roeper (2011). We next consider how far the approach can also be applied to L2 acquisition.

3.5 UG and L2 Acquisition

3.5.1 Theoretical Relevance of UG to L2 Learning

To address the potential of the UG model for L2 acquisition, we need to go back to the 1970s developments outlined in Chapter 2. For example, in that chapter we outlined similarities in the development of a number of English morphemes and of English negative and interrogative structures, in both L1 and L2 acquisition. In L1 acquisition, the UG explanation was that there was some kind of language blueprint in the brain constraining the hypotheses that children can entertain. This is the work we have summarized so far in this chapter.

If L2 learners also go through fairly rigid stages when acquiring certain constructions in the L2, producing interim forms which are unlike both their L1 and the L2 input they are receiving, and which are not unlike the stages children go through, then a similar explanation is surely worth inves- tigating. However, there are several complicating factors:

• L2 learners are cognitively mature; • L2 learners already know at least one other language; • L2 learners have different learning motivations (language learning does

not take place in order to answer the basic human need to communicate).

These points have important implications, and even if the UG hypothesis is correct for L1 learning, there are still a number of logical possibilities con- cerning its role in L2 learning:

1. L2s Are Not UG-Constrained

L2s are not constrained by UG principles and parameters, and they do not behave like natural languages.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 103

2. L2s Are UG-Constrained

FULL ACCESS

The whole of UG is available to L2 learners, as it is to L1 learners. Within this view, there are different hypotheses about the initial grammars of L2 learners, which we will review shortly.

PARTIAL ACCESS

Some parts of UG are not available any longer. For example, functional fea- tures which are not activated in the L1, particularly uninterpretable features such as grammatical gender, cannot be acquired. Within this view, L2 gram- mars do not violate principles and parameters, but learners might not be able to reset parameters, and therefore continue to operate with L1 settings for some parts of the new language.

3.5.2 Principles and Parameters in SLA

3.5.2.1 The Head Parameter

To begin to address these possibilities, let us return to the first examples which we used to illustrate L1 acquisition, namely the structure dependency principle and the head parameter.

Firstly, there seems to be no evidence in L2 grammars that learners ever violate the structure dependency principle. From the onset of L2 develop- ment, learners seem to know that the L2 will be hierarchically structured in terms of phrases.

Secondly, we saw that there are two possible settings for the head param- eter, head-first and head-last. Both French and English are head-first lan- guages; i.e. the head precedes its complements. However, in French, although all phrases normally exhibit this order, there is one exception (Hawkins, 2001, pp. 11–12): the case of unstressed object pronouns, as exemplified here:

1. Le chat [ VP

mange [ NP

la souris]] (“the cat eats the mouse”) 2. Le chat [

VP [ NP

la] mange] (the cat it eats = “the cat eats it”)

In VPs in French where the complement is a full NP (1), the head verb precedes its complement as normal; however, when the complement is an unstressed pronoun (2), the head verb follows it.1 From an acquisitional point of view, we have seen that children quickly set the head-direction param- eter, as all phrases in a given language normally follow the same order. For French children, there is ample evidence in the language around them that French is head-first. We would therefore expect French children to set the

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

104 Linguistics and Language Learning

parameter early on, and place the head before its complement. This is in fact the case, and children produce utterances such as *Le chat mange la, before going through a stage of omitting the pronoun altogether *Le chat mange ⊘, only later still inserting it in its target position Le chat la mange (Clark, 1985; Hamann, Rizzi, & Frauenfelder, 1996).

If French children follow this developmental sequence for ordering of unstressed object pronouns, then we should expect the same to happen for L2 French learners, if they are also guided by UG. There is indeed evidence to support this:

In fact, the stages of development that L1 English speakers go through in acquiring this pattern in L2 French are very similar to the stages that child L1 learners of French go through in acquiring it. Following an ini- tial stage where learners leave object pronouns postverbally in the posi- tion occupied by full noun phrases, e.g. Le chien a mangé les, “The dog has eaten them” (Zobl, 1980; Clark, 1985), they go to a stage of omission of the pronoun: Le chien a mangé ⊘ (Adiv, 1984; Schlyter, 1986; Véro- nique, 1986) before eventually acquiring preverbal object pronouns: Le chien les a mangés.

(Towell & Hawkins, 1994, p. 69)

On the other hand, L1 French learners of L2 English do not have problems in acquiring object pronouns in English, and do not go through a stage of preposing the pronoun (*the cat it eats), which would be expected if they transferred the French order, nor through a stage of omitting the pronoun (Zobl, 1980). This is to be expected if we assume that, on the basis of ample evidence in English that it is head-first, L2 learners set the head-direction parameter early on and apply it consistently. In the absence of evidence in the input that pronouns can precede the verb, they do not alter their initial hypothesis.

Because both French and English are head-first languages, we cannot say whether these observations are due to the fact that L2 learners reset the parameter to its correct value, or simply transfer their L1 parameter value. What is interesting, however, is that French learners do not transfer the idi- osyncratic property of French for pronoun placement.

In order to know whether the head parameter can be reset, it is neces- sary to investigate the acquisition of, say, a head-first language by learners whose L1 is head-last. Flynn has studied this question with Japanese learn- ers of English. (We have already seen that Japanese is a head-last language.) She concludes “that, from the earliest stages of acquisition, Japanese speakers learning English as a Second Language (ESL), are able to acquire the English value of the head-direction parameter” (Flynn, 1996, p. 135).

The evidence presented here therefore seems to suggest that, in the case of the head parameter at least, L2 learners have access to UG in the same way that children do. We cannot draw hasty conclusions on the basis of evidence

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 105

relating to one structure only, however, and other non-UG explanations have indeed been put forward.

3.5.2.2 Strong/Weak Infl (I)

We have explained earlier that the functional lexicon is now seen as the main site for parametric variation, as different languages select and bundle together functional features from the universal inventory differently, giving rise to the variations in word order and morphology evident in the world’s languages. Here we return to the example of strong/weak Infl in French and English, introduced in Section 3.3.1.

Remember that in French, Infl is strong, so that the verb rises past adverbs and negators, unlike in English, where Infl is weak and the verb remains within the Verb Phrase (except for auxiliaries and modals, which do rise to SpecIP, i.e. the specifier position in the Inflection Phrase). French learners of English therefore have to reset the verb movement parameter located in Infl to [-strong], and English learners of French have to reset it to [+strong]. Several studies have investigated this property (see White, 2003, for a review). Yuan (2001) studied the acquisition of L2 Chinese (weak Infl) by French (strong Infl) and English (weak Infl) learners. He found that all learners, regardless of their L1 or their proficiency level, realized the ungrammati- cality of verb-raising in Chinese, suggesting that they were able to reset this parameter. Rogers (2009), in her study of English learners of French, shows similar findings. Another study by White (1992), however, found somewhat different results. She studied the acquisition of verb-raising in questions, negatives and adverb placement, by French learners of L2 English. Her learners (beginners) seemed to have realized that English has weak Infl in the context of questions and negatives, but not in the context of adverbs. Learners rejected sentences such as:

• Like you pepperoni pizza? • The boys like not the girls.

with a high degree of accuracy. However, they commonly accepted:

• Linda takes always the metro.

White argues this might be because we are dealing with two different parameters underlying these properties (2003, pp. 129–132), though results to date are somewhat inconclusive.

3.5.3 Current Debates about the Role of UG in SLA

It should be clear by now that the general question which has generated so much research over several decades, namely whether UG is available to L2

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

106 Linguistics and Language Learning

learners or not, is being replaced by more focused questions and debates among generativist researchers, some of which we now introduce.

The Initial State

One area of active debate concerns what is termed the Initial State (the subconscious linguistic representations L2 learners have at the start of L2 learning). By its very nature, the Initial State of L2 acquisition differs from that of L1 acquisition by the very presence of another fully fledged language (as well as important cognitive and maturational differences). As in L1 acquisition (see Section 3.4), some generativist researchers have argued that functional categories are absent in the very early stages of adult L2 acquisition (Vainikka & Young-Scholten, 1996b, 1998, 2007; White, 2003; Hawkins, 2001, 2009; Myles, 2005), evidenced by a lack of morphological markings and of syntactic movement. Others, how- ever, have argued that functional categories are indeed present in the early stages in child L2 (Grondin & White, 1996; Haznedar, 2001) and also in adult L2 (Prévost & J. Paradis, 2004; Prévost & White, 2000; Renaud, 2014; Schwartz & Sprouse, 1996) and that the lack of morpho- logical markings is not a syntactic issue, but a processing difficulty. Some accounts argue that L2 learners gradually “build” syntactic structure, starting in local domains (the Structure-Building approach: Hawkins, 2001; Herschensohn, 2000). Yet others argue that the Initial State is the L1 with all its parameter values fixed (Schwartz & Sprouse, 1996). This exchange of views is complicated by the fact that the exact nature of functional features and categories themselves is a subject of debate in UG theory, and that L1 influences have to be taken into account (Lar- diere, 2012).

Ultimate Attainment/Steady State

Interest in L2 learners’ “ultimate attainment” arises from the well-docu- mented fact that the vast majority of them do not become native-like, even when the context is optimal—i.e. input and interaction are plentiful—as might be the case for immigrants who operate in their L2 on a daily basis. This is of course in stark contrast to children, who always become native-like in their L1. Another difference concerning L2 attainment is that it is less stable than L1 grammars, with learners sometimes applying a rule, other times not. This phenomenon is referred to as optionality (Sorace, 2005, 2011; Sorace & Serratrice, 2009). For example, Lardiere (1998a, 1998b, 2007) has worked in detail over time with a very proficient L2 user of English named Patty, an L1 Chinese speaker who has lived and worked in the USA for 18 years. Yet Patty’s L2 morphological system is still unstable and deviates from the target in specific ways, in spite of plentiful input; for example, she regularly

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 107

omits the 3rd person singular -s or past tense -ed. Researchers within the generative paradigm have argued that it is important to study learners—like Patty—whose development has reached its end point, in order to be able to assert that certain linguistic features are or are not acquirable by L2 learners. Such research has generally found systematic, if subtle, differences between end-state L2 learners and L1 speakers (Abrahamsson & Hyltenstam, 2009; Birdsong, 2009; DeKeyser, 2012; Herschensohn, 2007, 2013), especially when learning has begun post-puberty. These findings have been used to inform debates around the Critical Period Hypothesis, which claims that UG is no longer available to adult L2 learners, although no consensus has yet been reached, as we will discuss in Section 3.5.4.

The Role of Interfaces

Another departure from previous work focusing primarily on the availability of UG in SLA has been the recent interest in interfaces. This shift occurred as a result of research findings showing that some areas of syntax which require parameter resetting are acquired by L2 learners without undue dif- ficulty, yet learners have persistent problems in using those same structures in appropriate discourse contexts. For example, so-called pro-drop languages such as Italian or Spanish allow the subject of a sentence to be omitted, giv- ing rise to sentences such as:

(egli) mangia la mela (he) eats the apple (“he eats the apple”)

piove rains (“it’s raining”)

Selection of the + pro-drop value on this Null Subject parameter is usually linked in those languages to other properties such as much freer word order and richer inflectional morphology:

la mangia il ragazzo it eats the boy (“the boy eats it”)

Languages like English and French, on the other hand, do not allow subjects to be dropped in this way, and their word order is much more rigid. This property does not seem to pose many problems for L2 learners of pro-drop languages. For example, English or French learners do not have difficulties with dropping the subject or inverting the verb and its subject in L2 Italian or Spanish. However, even at very advanced levels, they have continuing problems with knowing when to do it. L1 speakers obey discourse constraints when choosing whether or not to omit the pronoun: they prefer to omit

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

108 Linguistics and Language Learning

the subject pronoun when there is no change in topic, and to have an overt subject pronoun when there is a change of topic. To offer an Italian example:

Mia madre ha comprato un dolce perche *lei/Ø aveva fame “My mother bought a cake because she was hungry”

In this case native speakers would omit the pronoun lei if “mother” is the subject of “was hungry”, as the topic remains constant (unless wishing to stress or contrast lei in some way). If a different (female) person was the subject of “was hungry”, however, they would produce the pronoun. L2 users have been shown to produce the overt pronoun optionally, in both cases, suggesting that they are not sensitive to the relevant discourse constraints.

The Interface Hypothesis has been proposed to explain findings like this. The suggestion is that “language structures involving an interface between syntax and other cognitive domains are less likely to be acquired completely than structures that do not involve this interface” (Sorace, 2011, p. 1). (For detailed discussion of this issue, see, for example, Domínguez, 2013; Roth- man & Slabakova, 2011; Sorace & Filiaci, 2006; Sorace & Serratrice, 2009; White, 2009a.) The Interface Hypothesis has received much attention, but mixed support. For example, Mai and Yuan (2016, p. 247), in their study of L2 Chinese by L1 English learners, conclude that the reassembly of features involving cross-domain operations (e.g. from prosody to syntax) is more difficult than reassembly within the same linguistic domain. Conversely, in a study of L2 Chinese by L1 Russian learners, Dugarova (2014, p. 431) concludes that wh-topicalization, a syntax-discourse phenomenon, can be acquired by very advanced learners, suggesting that not all interface phe- nomena are beyond L2 learners. Özçelik’s (2018) study of English learners of Turkish, and Turkish learners of English, found that learners had more difficulties with internal than external interfaces. In a study of L2 Japanese and Korean, Laleko and Polinsky (2016) suggest that interfaces might play a part in explaining difficulty, but they argue that structural complexity as well as memory demands also play a role. What the Interface Hypothesis has enabled researchers to do, however, is to generate testable (and increasingly sophisticated) hypotheses about how different modules of language might interact with one another during the acquisition process, and with other aspects of cognition.

The Relationship between Linguistic Knowledge and Language Processing

A longstanding debate in SLA research centres around L2 learners’ ultimate attainment, which is typically not fully native-like. Reasons put forward by generativist researchers include the (non)availability of UG reviewed above, and/or difficulties in processing linguistic knowledge in real time. Because

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 109

L2 learners’ mental representations of language can only be inferred indi- rectly, whether from their productions or their intuitions about language, teasing these two explanations apart has proved difficult, and researchers have engaged in lively debates about which explanation can best account for the same empirical findings. For example, the well-documented failure, even at advanced stages, to supply verb morphology consistently (3rd person -s or past tense -ed), in spite of its focus in instruction and its abundance in the input, has been explained as either a representational deficit (the syntax of these learners has not fully developed) or a processing problem (the learner has acquired and stored the morphological forms, but cannot retrieve them reliably).

The advent of new psycholinguistic and neurolinguistics methodologies has enabled new insights into how learners process specific language struc- tures in real time. For example, self-paced reading and eye-tracking have been widely used by SLA researchers to test whether L2 learners are sensi- tive to the same syntactic constraints as L1 speakers. (In self-paced reading, the participants are presented with a text word by word on a computer screen and have to press a button for the next word to appear—they are slower at pressing the button when encountering something unexpected or ungrammatical, showing processing difficulty. In eye-tracking studies, the movement of the eyes while reading a text is recorded—again here, processing difficulties lead to backtracking while reading the text, and the landing site of this backtracking, and the length of pausing, are indicative of where exactly processing is hindered.) Many phenomena of interest to UG researchers have been tested in this way, for example the constraints underpinning wh-movement (Juffs & Harrington, 1995, 1996), or anaphora resolution, i.e. what the pronouns him and himself can take as antecedents; see Section 3.2.2 above (Felser & Cunnings, 2012). These empirical studies are producing a range of fresh proposals from generativists about the source of differences in L2 vs L1 processing, and what these tell us about the under- pinning linguistic representations (Rothman et al., in press).

Neurolinguistic tools are also now used in SLA research as a means to explore the linguistic representations underpinning L2 use. The main tech- niques used are electroencephalography (EEG), which measures Event- Related Potentials (the brain’s electrophysiological response to a stimulus), or functional Magnetic Resonance Imaging (fMRI), which measures brain activity by detecting changes associated with blood flow. These tools are increasingly used by both UG and non-UG researchers to investigate and compare L1 and L2 speakers’ processing of linguistic stimuli; for a recent review of this work, see Benati and Rastelli (2018). Recent longitudinal research has shown, for example, that brain processes change with increased grammatical knowledge, becoming increasingly native-like (L. Roberts, González Alonso, Pliatsikas, & Rothman, 2018). These methodologies are further discussed in Chapter 5. And in Chapter 10, we discuss the MOGUL framework, as a recent principled approach to integrate a UG view of how

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

110 Linguistics and Language Learning

language is represented in the mind, with a transition theory grounded in language processing.

What Attrition and Heritage Language Speakers Can Tell Us about SLA

Increasingly, L2 acquisition is conceptualized as the development and interaction of multiple languages in the mind of an individual. Research has convincingly shown that when we learn a second or third language, our L1 and any other previously acquired language change as a result. The interaction between L1 and L2 is not unidirectional, with L1 properties transferring onto the L2; it is bidirectional, with changes also taking place in our L1 (see Cook’s Multi-competence Model, 2016, or Dynamic Sys- tems Theory, discussed here in Chapter 10). As Schmid and Köpke (2018, p. 638) put it:

Research on second language acquisition and bilingual development strongly suggests that when a previously monolingual speaker becomes multilingual, the different languages do not exist in isolation: they are closely linked, dependent on each other, and there is constant interac- tion between these different knowledge systems.

There are many long-term transnational migrants today who use one or more L2s for most functions of daily life, and make only infrequent use of their L1. Under such circumstances, the L1 can become affected through competition with the L2. This phenomenon is described as language attri- tion (Schmid, 2011); typical problems include forgetting some words, becoming hesitant and disfluent, using odd expressions and developing a “foreign” accent. Current UG-inspired research on language attrition is pro- viding useful information about the nature and resilience of the underpin- ning linguistic system established in early childhood, the extent to which feature reassembly may take place within the L1 over extended periods of time, and the role of age in language acquisition, maintenance and change (Schmid, 2014; Domínguez, 2018).

Heritage language speakers are another multilingual population of cur- rent interest to UG researchers. Heritage speakers are L1 speakers of a minority language they learn at home, who live in a community where a different language is spoken. In such cases, it is usual for the heritage language not to fully develop (Montrul, 2015). If these speakers are acquir- ing their home language from birth, any representational deficit cannot be explained through lack of access to UG, and other factors have to be explored. Which elements of the linguistic system are vulnerable, and why? And how far do the representational deficits of heritage speakers resemble those of L2 learners, whether children or adults? All these questions have been investigated recently, with mixed results. For example, Silva-Corvalán

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 111

(2014) and Montrul (2018) found that heritage speakers have difficulties with the discourse-pragmatic rules underlying the use of subject pronouns in Spanish, just like adult L2 learners. Polinsky (2011) showed that Russian adult heritage speakers performed differently from monolingual Russian adults and from Russian child heritage speakers in their comprehension of relative clauses, suggesting that their grammar has actually attrited. Kaltsa, Tsimpli, and Rothman (2015) compared monolingual speakers of Greek with groups of attriters and heritage speakers who were bilingual in Greek and Swedish. They again focused on the use of null versus overt pronouns, and found the two bilingual groups to be different from the monolingual group. However, they suggest that the source of divergence is different in the two groups:

. . . while L1 attrition is clearly loss at some abstract level, heritage speaker divergence can be viewed as the outcome of a developmental path that is destined to be distinct from monolinguals as a result of a general byproduct of bilingualism, qualitative input differences among other factors.

(p. 283)

The inclusion of these different populations where multiple languages are learned and used in different contexts at different times of life brings a new dimension to UG-inspired attempts to understand how different languages are acquired, are represented and coexist/compete in the human mind, with important implications for our understanding of the nature of human lan- guage and its genetic endowment.

3.5.4 Empirical Evidence

After having illustrated the UG perspective and its relevance to L2 learn- ing, taking the example of one principle (structure dependency) and two parameters (head-direction and strength of Infl), and having introduced some more recent theoretical debates and concepts, we can now reassess the theoretical positions we outlined in 3.5.1. These various positions attempt to reconcile somewhat contradictory facts about L2 acquisition:

• Learners do not seem to produce “wild” grammars, i.e. grammars which fall outside the constraints of UG (Schwartz & Sprouse, 2017). Does that suggest that UG is available to them? Additionally, learners have grammatical knowledge that does not appear to derive from experience, known as Poverty of the Stimulus (POS) phenomena (Hawkins, 2004, 2008a; Rothman & Slabakova, 2018; Schwartz & Sprouse, 2013);

• Learners produce grammars which are not necessarily like either their L1 or their L2. Does this suggest that parameter settings other than those of their L1 or L2 are available to them?

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

112 Linguistics and Language Learning

• Some parameters seem to be easy to reset (like the head parameter), oth- ers more difficult, or even impossible. Why?

• Learners sometimes fail to fully acquire properties of the L2 in spite of abundant evidence in the input and after extended exposure to the L2 (as for Lardiere’s participant Patty: Lardiere, 2007).

3.5.4.1 Hypothesis 1: UG Does Not Play a Role in SLA

UG theorists who believe that UG is no longer available to adult L2 learn- ers argue that there is a “critical period” for language acquisition during children’s early development, and that adults have to resort to other learning mechanisms. There are several reasons for adopting such a position (Slaba- kova, 2009; Bley-Vroman, 1989, 2009), but perhaps the most convincing one is the common-sense observation that immigrant children generally become native-like speakers of their L2, whereas their parents rarely do. For example, an influential study (Johnson & Newport, 1989) found a correla- tion between age of arrival in the USA and native-like judgements on a number of grammatical properties of English. Immigrants who had arrived before age 7 performed in a native-like way, and the older learners were on arrival, the more errors they made in the test. More recently, Abrahamsson and Hyltenstam (2009) have conducted very detailed analyses of the L2 of advanced adult learners who pass as L1 speakers, and suggest that they still differ from those speakers in subtle ways. This study failed to find a clear cut- off point, however, which would be the best indicator of a critical period for the availability of UG. (Some UG researchers have challenged the Johnson and Newport study, arguing their evidence is compatible with continuing availability of UG: Hawkins, 2001; White, 2003. See Chapter 5 for further discussion of age effects in SLA, from another theoretical perspective.)

In an extensive study of the acquisition of negation in French and Ger- man by L1 and L2 learners, Meisel (1997, p. 258) concluded: “I would like to hypothesize that second language learners, rather than using structure- dependent operations constrained by UG, resort to linear sequencing strat- egies which apply to surface strings”. Here, Meisel was claiming that one of the most fundamental principles of UG (structure dependency) is not available to L2 learners any more. It must be said, however, that this view has not received much support, and Meisel himself (2011) reviewed his posi- tion, claiming that “L2 learners’ hybrid systems are indeed natural grammars” (p. 251), and that “L2 knowledge and acquisition processes are in large part domain-specific and share with native grammars the crucial property of structure dependency” (p. 252). Evidence for wild grammars which violate UG has not been forthcoming in L2 research (Schwartz & Sprouse, 2017).

Researchers from outside the UG framework, and in particular propo- nents of the “emergentist” position, of course also compare more broadly whether a usage-based account better explains some of the facts of SLA, and conclude that UG is not involved (see Chapter 4).

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 113

3.5.4.2 Hypothesis 2: UG Is Fully Available to Second Language Learners

FULL ACCESS/NO TRANSFER

Some early UG work adopted this position. For example, Flynn (1996) argued that UG continues to underpin L2 learning, for adults as well as children, and that there is no such thing as a critical period after which UG ceases to operate. If it can be shown that learners can acquire the principles and/or parameter settings of the L2, which differ from those of their L1, she claimed, the best interpretation is the continuing operation of UG, a view supported by her review of a range of empirical work with L2 learners moving from Japanese to English (pp. 134–148). We have already met her claim that adult Japanese learners of L2 English can successfully reset the head-direction parameter (i.e. from head-last to head-first). She also claimed that similar learners can activate principles which do not operate in Japanese, such as the Subjacency principle (which controls wh-movement in English; i.e. the way in which we move the wh- phrase to the beginning of the sen- tence, which is subject to certain constraints), and can acquire functional categories which are supposedly non-existent in Japanese. Flynn concluded her review thus:

It appears that L2 learners do construct grammars of the new TLs [tar- get languages] under the constraints imposed by UG; those principles of UG carefully investigated thus far indicate that those not instantiated or applying vacuously in the L1 but operative in the L2, are in fact acquir- able by the L2 learner.

We are thus forced to the conclusion that UG constrains L2 acquisi- tion; the essential language faculty involved in L1 acquisition is also involved in adult L2 acquisition.

(Flynn, 1996, pp. 150–151)

Other researchers who have argued that UG is still available to L2 learners include Thomas (1991), on the basis of work on the acquisition of reflexive binding, and White, Travis, and MacLachlan (1992), on the basis of work on wh-movement. However, more recent research which argues for full access to UG usually acknowledges that the L1 is involved in shaping initial hypotheses at least.

FULL TRANSFER/FULL ACCESS

This position agrees that L2 learners have full access to UG and its param- eters, whether they are present in the learners’ L1 or not (Ionin, Zubi- zarreta, & Maldonado, 2008; Schwartz & Sprouse, 1994, 1996). But in this view, L2 learners are thought to begin by transferring all the parameter set- tings from their L1, and to revise their hypotheses afterwards, when the L2

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

114 Linguistics and Language Learning

fails to conform to these L1 settings. Learners then develop new hypotheses which are UG-constrained. For example, a study of the acquisition of Eng- lish articles by Spanish and Russian learners (Ionin et al., 2008) found that Spanish learners (whose article system encodes definiteness and specificity in the same way as English) transfer their L1 settings. Russian learners, how- ever, whose L1 does not have articles, access the semantic universals of defi- niteness and specificity, and initially fluctuate between them before adopting the correct setting for English. This suggests that they have full access to uni- versal features not realized in their L1. In a recent study of the L2 acquisition of the count/mass distinction in English by Korean and Mandarin speakers, Choi, Ionin, and Zhu (2017) found that learners had access to the semantic universal of atomicity. For a review of studies supporting the full transfer/full access hypothesis, see White (2003).

FULL ACCESS/IMPAIRED EARLY REPRESENTATIONS

Some researchers also believe that learners can reset parameters to the L2 values, but that initially they are lacking functional categories altogether. The Minimal Trees approach, later renamed as Organic Grammar (Vainikka & Young-Scholten, 1996b, 1998, 2007, 2011), forms the starting point for a number of accounts of the development of syntax: only lexical categories are projected initially, via transfer from the L1. Functional categories develop later, but are not transferred from the L1.

This view has much in common with the approaches we will review next, but crucially believes that all parameters can be reset.

3.5.4.3 Hypothesis 3: L2 Grammars Are Partially Constrained by UG

NO PARAMETER RESETTING

Proponents of this position claim that learners only have access to UG via their L1. They have already accessed the range of principles applying to their L1, and set parameters to the L1 values, and this is the basis for their L2 development. Other parameter settings are not available to them, and if the L2 possesses parameter settings which are different from those of their L1, they will have to resort to other mechanisms in order to make the L2 data fit their internal representations. These mechanisms will be rooted in general problem-solving strategies, rather than being UG-based.

For example, Schachter studied Korean L1 learners of English as L2, who performed randomly in grammaticality judgement tests of wh- movement (Schachter, 1990, cited in Schachter, 1996). In English, wh-movement is allowed, but is restricted by Subjacency (the extracted wh-word can move only across certain structural boundaries). In Korean, there is no wh-movement, so Subjacency is presumably not operative. If UG is still available to the learner, the absence of Subjacency from their L1 should

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 115

not matter, and it should still be acquirable in English L2. Schachter claims that the Korean subjects’ failure to recognize wh-movement constraints reflects the non-availability of aspects of UG which were not already operative in their L1. However, other studies have claimed that adult L2 learners whose L1 lacks overt wh-movement do eventually observe Sub- jacency in the L2 (Belikova & White, 2009).

Schachter does accept that UG may be available for child L2 learners, but believes that a critical period (or periods) applies for the successful acquisi- tion of L2 principles and/or parameter settings which have not been opera- tive in the learner’s L1 (1996). In support, she cites a study by Lee (1992) which tested Korean-English bilinguals on a particular parameter, the Gov- erning Category parameter, which is set differently in the two languages (as described in Section 3.3.2).

In Lee’s (1992) study, the Korean learners of English were of different ages; the youngest and oldest participants had not acquired the English set- ting for the Governing Category parameter, while the older children had apparently succeeded in doing so. Schachter concludes that these findings show a Window of Opportunity not yet operative for the youngest learners, but available to the older children. However, she concludes that the only parameter settings easily available to the adult L2 learner are those already activated in the L1.

IMPAIRED FUNCTIONAL FEATURES

Finally, we will review influential approaches which believe that L2 gram- mars are UG-constrained, but that not all parameter settings are available to learners. Different terms have been used to describe this approach: the Failed Functional Features Hypothesis (Hawkins, 2001; Hawkins & Chan, 1997), the Representational Deficit Hypothesis (Hawkins & Liszka, 2003) and the Interpretability Hypothesis (Hawkins & Hattori, 2006; Tsimpli & Dimitrakopoulou, 2007). This approach argues that uninterpretable func- tional features (explained above in Section 3.3.2) cannot be reset in the L2. For example, Cantonese learners of English studied by Hawkins and Chan (1997) failed to acquire properties linked with wh-movement, which does not exist in Cantonese. Similarly, Franceschina (2005) showed that the grammatical gender feature is not available to L2 learners whose L1 does not activate it.

3.5.4.4 Access to UG Revisited

With the advent of Minimalism, the role of UG in L2 acquisition has been reconceptualized somewhat. With language variation now understood as being located in the different selection and bundling of universal features, the questions being asked in generativist SLA research have shifted; UG researchers are now making close comparisons of how the feature inventory

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

116 Linguistics and Language Learning

is grammaticalized (or not) in different languages. This has led to the pro- posal by Lardiere (1998a, 1998b, 2000, 2007, 2009, 2012) of the Feature Reassembly Hypothesis, which argues that L2 learners’ problems with mor- phology are due to mapping problems: functional categories in different languages not only select different functional features from the universal set provided by UG, but also bundle features in different ways when mapping onto specific morphosyntactic forms. The task facing L2 learners is there- fore not only to realize which features are operational in the L2, but also to discover how they map onto specific forms in the L2. For Lardiere, this mapping or reassembling of features in new clusters is the main source of persistent difficulties.

Other proposals such as Slabakova’s Bottleneck Hypothesis also argue that functional morphemes and their features are the main challenge in L2 acquisition; acquisition of universal syntax, semantics and pragmatics flows smoothly (Slabakova, 2006, 2008, 2013).

The accounts reviewed in this section all share the belief that L2 acquisi- tion is UG-constrained, but that the acquisition of parametric options might be problematic and unlike L1 acquisition. However, they have different views on the Initial State, on the role of the L1, on the possibility of param- eter resetting, on the Steady State and on the role of non-UG-constrained mechanisms.

To round off this section, it is fair to say that the argument concerning access to UG in L2 learning is not concluded, and that defenders of all these positions can still be found. Often, they seem to be arguing about the best technical interpretation of admittedly indirect and tantalizing evidence. Research in this area has shifted from the initial question of the availability versus non-availability of UG, towards a more modular view of language and the language faculty, while understandings of UG itself have also been evolving in fundamental ways.

3.6 Evaluation of UG-Based Approaches to L2 Acquisition

3.6.1 The Scope and Achievements of the UG Approach

UG is a well-established theory which aims to describe and explain all human language. In evaluating the relevance of UG theory to L2 learn- ing, however, we must remember that it is a linguistic theory, and not a learning or transition theory. Although one of Chomsky’s stated objectives mentioned earlier on in this chapter is to understand how knowledge of language is acquired and put to use, most UG work to date has focused on his first question: what constitutes knowledge of language? These questions are related, though, and both L1 and L2 acquisition data have increasingly been used to refine and test hypotheses about the nature of human language. Additionally, the UG descriptive framework has been used by researchers to

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 117

draw up sophisticated hypotheses about a range of issues which are central to our understanding of SLA, such as the exact nature of the language system at different stages of acquisition (the learner system as well as the L1 and target L2 systems), the interplay between L1 and L2, and the linguistic knowledge learners bring to the task of L2 acquisition.

As a general theory of language, the scope of UG is potentially very broad. It would be fair to say, however, that UG research has been primarily con- cerned with the description and explanation of the formal system underly- ing language, with a main focus on morphosyntax. The UG contribution to our understanding of the acquisition of morphosyntactic properties in L2 acquisition has indeed been outstanding. However, its scope does not include a theory of processing, nor a theory of learning, although it has increasingly been concerned with what processing can tell us about linguistic knowledge. It has very little to say about what triggers development in either L1 or L2 acquirers, or about the role of the input (Rankin & Unsworth, 2016). To repeat, it is a property theory and not a transition theory, and must therefore be evaluated as such.

3.6.2 The UG View of Language

The UG approach is concerned to describe the mental framework under- lying all human languages. Until very recently as we have seen, syntax was the privileged object of study. UG is primarily concerned with the sentence and its internal structure, rather than any larger unit of language. Work at the level of smaller units (words, morphemes, phonemes) has also been con- cerned with structure and how different elements relate to one another. This is one of the major criticisms of work in this tradition; it studies language somewhat clinically, in a vacuum, as a mental object rather than a social or psychological one. Moreover, it separates rigidly language knowledge and language use, and many linguists disagree with this dichotomy, as we see elsewhere in this book.

Following from this, the methodologies used by UG theorists have sometimes been criticized for lack of ecological validity. The theory is pre- occupied with the modelling of linguistic competence, and the study of naturalistic performance has not been accepted as a suitable window into mental representations of language (Towell & Hawkins, 2004). We have seen (Chapter 1) that grammaticality judgement tests used to be seen as the most appropriate methodology to access native speakers’ intuitions about their native language, and that native speakers usually agree about what is gram- matical or ungrammatical in their language. L2 learners’ intuitions, however, are much more likely to be unstable, and therefore less reliable; for example, their responses to grammaticality judgement tests may also be influenced by explicit knowledge. We have seen in earlier sections how data on L2 compe- tence deriving from grammaticality judgement tests are often disputed and reinterpreted (for discussion see Chaudron, 2003; Sorace, 1996; Ionin, 2012).

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

118 Linguistics and Language Learning

UG theorists have taken criticisms about the lack of reliability of L2 judgements seriously, and more recent work in this tradition has used a much wider range of elicitation techniques, ranging from analysis of large naturalistic corpora to experimental methodologies such as eye-tracking, measurement of Event-Related Potentials and brain-imaging. The problem of drawing inferences about mental representations from such data remains. However, in spite of these criticisms, UG remains probably the most sophis- ticated tool available for linguistic analysis.

3.6.3 The UG View of Language Acquisition

When applied specifically to the analysis of interlanguage and the problem of L2 acquisition, how successful can UG theory claim to be?

UG-based approaches to SLA have been criticized for exactly the same reasons as the theory itself. Firstly, linguistically, this approach has tradition- ally been almost exclusively concerned with morphosyntax, though recent interest in phonology (e.g. Archibald, 1998, 2004, 2009, 2018; Özçelik & Sprouse, 2017; Roncaglia-Denissen, Schmidt-Kassow, Heine, & Kotz, 2015; Schmid, Gilbers, & Nota, 2014), morphology (as seen in previous pages) and the lexicon (e.g. Hopp, 2017; Juffs, 2009; van Hout, Hulk, Kuiken, & Towell, 2003) has redressed the balance somewhat. Additionally, semantics, pragmat- ics and discourse are now of concern in work on interfaces. Secondly, the UG approach has been exclusively concerned with documenting and explain- ing the nature of the L2 linguistic system. Social and psychological variables which affect the learning process are beyond its remit and therefore ignored.

Overall, there is little doubt that the UG approach has greatly enhanced our understanding of L2 morphosyntactic development. It has enabled researchers to formulate well-defined hypotheses which could then be tested in empirical work. This powerful linguistic tool has been useful in describ- ing not only the language produced by learners, but also the language to be acquired as well as the L1. In addition to establishing some of the facts about L2 acquisition, the UG approach has also proposed explanations for such facts. For example, it has enabled L2 researchers to theorize cross-linguistic influence in a new way, in terms of parametric variation, and this has led to a productive programme of empirical research.

3.6.4 The UG View of the Language Learner

The UG approach is only interested in the learner as the possessor of a mind which contains language(s); the assumption is that all human beings are endowed with such a mind, and variations between individuals are of little concern to UG theorists. The emphasis is very much again here on language as the object of study, rather than on the speaker or learner as a social being, and the focus is on what is universal within this mind.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 119

Overall, there is little doubt that the UG approach to L2 research meets the criteria for a good theory as defined in Chapter 1, by making clear and explicit statements of the ground it aims to cover and the claims it makes, by having systematic procedures for theory evaluation, by attempting to explain as well as describe at least some L2 phenomena and, finally, by engaging increasingly with other theories in the field. As one of the most active and developing theories, it can be expected to continue to make highly valuable contributions to the field, within its self-determined limitations.

Note

1 In fact, this is not a violation of this parameter setting but occurs because unstressed pronouns in French cliticize onto the verb (i.e. attach themselves to the verb). In other words, object pronouns originate after the verb as expected given that French is head- first, and subsequently move to a preverbal position.

References

Abrahamsson, N., & Hyltenstam, K. (2009). Age of onset and nativelikeness in a second language: Listener perception versus linguistic scrutiny. Language Learning, 59, 249–306.

Adiv, E. (1984). Language learning strategies: The relationship between L1 operating principles and language transfer in L2 development. In R. W. Andersen (Ed.), Second languages: A cross-linguistic perspective (pp. 125–142). Rowley, MA: Newbury House.

Aitchison, J. (2008). The articulate mammal (5th ed.). Abingdon/New York: Routledge. Ambridge, B., & Lieven, E. V. M. (2011). Child language acquisition: Contrasting theoretical

approaches. Cambridge: Cambridge University Press. Archibald, J. (1998). Second language phonology. Amsterdam: John Benjamins. Archibald, J. (2004). Interfaces in the prosodic hierarchy: New structure and the phono-

logical parser. The International Journal of Bilingualism, 8, 29–50. Archibald, J. (2009). Phonological feature reassembly and the importance of phonetic

cues. Second Language Research, 25, 231–233. Archibald, J. (2018). Transfer, contrastive analysis and interlanguage phonology. In O.

Kang, R. Thomson, & J. Murphy (Eds.), The Routledge handbook of contemporary English pronunciation (pp. 9–24). Abingdon/New York: Routledge.

Baker, M. (2008). The macroparameter in a microparametric world. In T. Biberauer (Ed.), The limits of syntactic variation (pp. 351–374). Amsterdam: John Benjamins.

Bartke, S., & Siegmuller, J. (Eds.). (2004). Williams syndrome across languages. Amsterdam: John Benjamins.

Belikova, A., & White, L. (2009). Evidence for the fundamental difference hypothesis or not? Island constraints revisited. Studies in Second Language Acquisition, 31, 199–223.

Bellugi, U., Van Hoek, K., Lillo-Martin, D., & O’Grady, L. (1993). The acquisition of syntax in young deaf singers. In D. Bishop & K. Mogford (Eds.), Language development in exceptional circumstances (pp. 132–150). Hove: Lawrence Erlbaum.

Benati, A., & Rastelli, S. (Eds.). (2018). Neurolinguistics and the language classroom [Spe- cial issue]. Second Language Research, 34.

Berwick, R. C., & Chomsky, N. (2016). Why only us: Language and evolution. Cambridge, MA: MIT Press.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

120 Linguistics and Language Learning

Birdsong, D. (2009). Age and the end state of second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), The new handbook of second language acquisition (pp. 401– 424). Bingley: Emerald Group.

Bishop, D. (Ed.). (2001). Language and cognitive processes in developmental disorders. Hove: Psychology Press.

Bishop, D., & Mogford, K. (Eds.). (1993). Language development in exceptional circumstances. Hove: Lawrence Erlbaum.

Bley-Vroman, R. (1989). What is the logical problem of foreign language learning? In S. Gass & J. Schachter (Eds.), Linguistic perspectives on second language acquisition (pp. 41–68). Cambridge: Cambridge University Press.

Bley-Vroman, R. (2009). The evolving context of the fundamental difference hypothesis. Studies in Second Language Acquisition, 31, 175–198.

Boeckx, C. (2011). Approaching parameters from below. In A.-M. Di Sciullo & C. Boeckx (Eds.), Biolinguistic approaches to language evolution and variation (pp. 205–221). Oxford: Oxford University Press.

Carter, R. (1998). Mapping the mind. London: Weidenfeld & Nicolson. Chaudron, C. (2003). Data collection in SLA research. In C. Doughty & M. H. Long

(Eds.), The handbook of second language acquisition (pp. 762–828). Oxford: Wiley Blackwell. Choi, S. H., Ionin, T., & Zhu, Y. (2017). L1 Korean and L1 Mandarin L2 English learn-

ers’ acquisition of the count/mass distinction in English. Second Language Research, 34, 147–177.

Chomsky, N. (1981). Lectures on government and binding. Dordrecht: Foris. Chomsky, N. (1986a). Knowledge of language, its nature, origin and use. New York: Praeger. Chomsky, N. (1986b). Barriers. Cambridge, MA: MIT Press. Chomsky, N. (1995). The minimalist program. Cambridge, MA: MIT Press. Chomsky, N. (2000). New horizons in the study of language and mind. Cambridge: Cam-

bridge University Press. Chomsky, N. (2002). On nature and language. Cambridge: Cambridge University Press. Chomsky, N. (2005). Three factors in language design. Linguistic Inquiry, 36, 1–22. Chomsky, N. (2007a). Approaching UG from below. In U. Sauerland & H. Gärtner (Eds.),

Interfaces + recursion = language? (pp. 1–31). Berlin: Mouton de Gruyter. Chomsky, N. (2007b). Of minds and language. Biolinguistics, 1, 9–27. Chomsky, N. (2008). On phases. In R. Freidin, C. Otero, & M.-L. Zubizaretta (Eds.),

Foundational issues in linguistic theory: Essays in honor of Jean-Roger Vergnaud. Cambridge, MA: MIT Press.

Chomsky, N. (2013). Problems of projection. Lingua, 130, 33–49. Clark, E. (1985). The acquisition of Romance, with special reference to French. In D.

Slobin (Ed.), The crosslinguistic study of language acquisition (pp. 687–782). Hillsdale, NJ: Lawrence Erlbaum.

Cook, V. (2016). Premises of multi-competence. In V. Cook & Li Wei (Eds.), The Cam- bridge handbook of linguistic multi-competence (pp. 1–25). Cambridge: Cambridge Univer- sity Press.

Cook, V., & Newson, M. (1996). Chomsky’s Universal Grammar: An introduction (2nd ed.). Oxford: Wiley Blackwell.

Cook, V., & Newson, M. (2007). Chomsky’s Universal Grammar: An introduction (3rd ed.). Oxford: Wiley Blackwell.

Curtiss, S. (1977). Genie: A psycholinguistic study of a modern-day “wild child”. New York: Academic Press.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 121

de Villiers, J., & Roeper, T. (Eds.). (2011). Handbook of generative approaches to language acquisition. Dordrecht: Springer.

DeKeyser, R. (2012). Age effects in second language learning. In S. M. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition (pp. 442–460). Abingdon/ New York: Routledge.

Di Sciullo, A.-M., Piattelli-Palmarini, M., Wexler, K., Berwick, R. C., Boeckx, C., Jenkins, L., . . . Bever, T. (2010). The biological nature of human language. Biolinguistics, 4, 4–34.

Domínguez, L. (2013). Understanding interfaces: Second language acquisition and first language attrition of Spanish subject realization and word order variation. Amsterdam: John Benjamins.

Domínguez, L. (2018). Bridging the gap between selective and non-selective L1 attrition. Linguistic Approaches to Bilingualism, 7, 686–690.

Dornyei, Z. (2009). The psychology of second language acquisition. Oxford: Oxford University Press.

Dugarova, E. (2014). Russian speakers’ L2 Chinese acquisition of wh-topicalization at the syntax—discourse interface. Second Language Research, 30, 411–437.

Eguren, L., Fernandez-Soriano, O., & Mendikoetxea, A. (Eds.). (2016). Rethinking param- eters. Oxford: Oxford University Press.

Eubank, L., & Gregg, K. (1999). Critical periods and (second) language acquisition: Divide et impera. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis (pp. 65–99). Mahwah, NJ: Lawrence Erlbaum.

Felser, C., & Cunnings, I. (2012). Processing reflexives in a second language: The timing of structural and discourse-level information. Applied Psycholinguistics, 33, 571–603.

Flynn, S. (1996). A parameter-setting approach to second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 121–158). San Diego: Academic Press.

Franceschina, F. (2005). Fossilized second language grammars: The acquisition of grammatical gender. Amsterdam: John Benjamins.

Friederici, A. (2017). Language in our brain: The origins of a uniquely human capacity. Cam- bridge, MA: MIT Press.

Fukui, N., & Speas, M. (1986). Specifiers and projection. MIT Working Papers in Linguistics, 8, 128–172.

Gopnik, M., & Crago, M. B. (1991). Familial aggregation of a developmental language disorder. Cognition, 39, 1–50.

Grondin, N., & White, L. (1996). Functional categories in child L2 acquisition of French. Language Acquisition, 5, 1–34.

Guasti, M. (2009). Universal Grammar approaches to language acquisition. In S. Foster- Cohen (Ed.), Language acquisition (pp. 87–108). Basingstoke: Palgrave Macmillan.

Guasti, M. (2016). Language acquisition: The growth of grammar (2nd ed.). Cambridge, MA: MIT Press.

Hamann, C., Rizzi, L., & Frauenfelder, U. (1996). On the acquisition of subject and object clitics in French. In H. Clahsen (Ed.), Generative perspectives on language acquisition (pp. 309–334). Amsterdam: John Benjamins.

Hawkins, R. (2001). Second language syntax: A generative introduction. Oxford: Wiley Blackwell.

Hawkins, R. (2008a). Commentary: Can innate linguistic knowledge be eliminated from theories of SLA? Lingua, 118, 613–619.

Hawkins, R. (2008b). The nativist perspective on second language acquisition. Lingua, 118, 465–477.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

122 Linguistics and Language Learning

Hawkins, R. (2009). Second language acquisition of morphosyntax. In W. C. Ritchie & T. K. Bhatia (Eds.), The new handbook of second language acquisition (pp. 211–236). Bing- ley: Emerald Group.

Hawkins, R., & Chan, C. (1997). The partial availability of Universal Grammar in sec- ond language acquisition: The “failed functional features hypothesis”. Second Language Research, 13, 187–226.

Hawkins, R., & Franceschina, F. (2004). Explaining the acquisition and non-acquisition of determiner-noun gender concord in French and Spanish. In P. Prévost & J. Paradis (Eds.), The acquisition of French in different contexts: Focus on functional categories (pp. 175– 205). Amsterdam: John Benjamins.

Hawkins, R., & Hattori, H. (2006). Interpretation of English multiple wh-questions by Japanese speakers: A missing uninterpretable feature account. Second Language Research, 22, 269–301.

Hawkins, R., & Liszka, S. (2003). Locating the source of defective past tense marking in advanced L2 English speakers. In R. van Hout, A. Hulk, F. Kuiken, & R. Towell (Eds.), The lexicon-syntax interface in second language acquisition (pp. 21–44). Amsterdam: John Benjamins.

Haznedar, B. (2001). The acquisition of the IP system in child L2 English. Studies in Second Language Acquisition, 23, 1–39.

Herschensohn, J. (2000). The second time around: Minimalism and second language acquisition. Amsterdam: John Benjamins.

Herschensohn, J. (2007). Language development and age. Cambridge: Cambridge University Press.

Herschensohn, J. (2013). Age-related effects. In J. Herschensohn & M. Young-Scholten (Eds.), The Cambridge handbook of second language acquisition (pp. 317–337). Cambridge: Cambridge University Press.

Hopp, H. (2017). Cross-linguistic lexical and syntactic co-activation in L2 sentence pro- cessing. Linguistic Approaches to Bilingualism, 7, 96–130.

Huang, C.-T. J., & Roberts, I. (2017). Principles and parameters of Universal Grammar. In I. Roberts (Ed.), The Oxford handbook of universal grammar (pp. 307–354). Oxford: Oxford University Press.

Ionin, T. (2012). Formal theory-based methodologies. In A. Mackey & S. M. Gass (Eds.), Research methods in second language acquisition: A practical guide (pp. 30–52). Oxford: Wiley Blackwell.

Ionin, T., Zubizarreta, M. L., & Maldonado, S. B. (2008). Sources of linguistic knowledge in the second language acquisition of English articles. Lingua, 118, 554–576.

Jenkins, L. (2000). Biolinguistics. Cambridge: Cambridge University Press. Johnson, J., & Newport, E. (1989). Critical period effects in second language learning:

The influence of maturational state on the acquisition of ESL. Cognitive Psychology, 21, 60–99.

Juffs, A. (2009). Second language acquisition of the lexicon. In W. C. Ritchie & T. K. Bhatia (Eds.), The new handbook of second language acquisition (pp. 181–210). Bingley: Emerald Group.

Juffs, A., & Harrington, M. (1995). Parsing effects in second language sentence processing: Subject and object asymmetries in wh-extraction. Studies in Second Language Acquisition, 17, 483–516.

Juffs, A., & Harrington, M. (1996). Garden path sentences and error data in second lan- guage sentence processing. Language Learning, 46, 283–323.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 123

Kaltsa, M., Tsimpli, I. M., & Rothman, J. (2015). Exploring the source of differences and similarities in L1 attrition and heritage speaker competence: Evidence from pronomi- nal resolution. Lingua, 164, 266–288.

Lai, C., Fisher, S., Hurst, J., Vargha-Khadem, F., & Faraneh, M. (2001). A fork-head-domain gene is mutated in a severe speech and language disorder. Nature, 413, 519–523.

Laleko, O., & Polinsky, M. (2016). Topic and case marking in heritage speakers and L2 learners of Japanese and Korean. Linguistic Approaches to Bilingualism, 6, 396–439.

Lardiere, D. (1998a). Case and tense in the “fossilized” steady state. Second Language Research, 14, 1–26.

Lardiere, D. (1998b). Dissociating syntax from morphology in a divergent L2 end-state grammar. Second Language Research, 14, 359–375.

Lardiere, D. (2000). Mapping features to forms in second language acquisition. In J. Archibald (Ed.), Second language acquisition and linguistic theory (pp. 102–129). Chiches- ter: Wiley Blackwell.

Lardiere, D. (2007). Ultimate attainment in second language acquisition: A case study. Mahwah, NJ: Lawrence Erlbaum.

Lardiere, D. (2009). Some thoughts on the contrastive analysis of features in second lan- guage acquisition. Second Language Research, 25, 173–227.

Lardiere, D. (2012). Linguistic approaches to second language morphosyntax. In S. M. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition (pp. 106– 126). Abingdon/New York: Routledge.

Lee, D. (1992). Universal Grammar, learnability, and the acquisition of English reflexive bindings by L1 Korean speakers. Doctoral dissertation, University of Southern California, Los Angeles.

Lenneberg, E. (1967). The biological foundations of language. New York: Wiley. Leung, Y. I. (Ed.). (2009). Third language acquisition and Universal Grammar. Clevedon: Mul-

tilingual Matters. Lidz, J., & Gagliardi, A. (2015). How nature meets nurture: Universal grammar and statis-

tical learning. Annual Review of Linguistics, 1, 333–353. Lightbown, P. M., & Spada, N. (2013). How languages are learned (4th ed.). Oxford: Oxford

University Press. Lorenzo, G., & Longa, V. (2003). Minimizing the genes for grammar: The minimalist

program as a biological framework for the study of language. Lingua, 113, 643–657. Mai, Z., & Yuan, B. (2016). Uneven reassembly of tense, telicity and discourse features in

L2 acquisition of the Chinese shì . . . de cleft construction by adult English speakers. Second Language Research, 32, 247–276.

Meisel, J. (1997). The acquisition of the syntax of negation in French and German: Con- trasting first and second language development. Second Language Research, 13, 227–263.

Meisel, J. (2011). First and second language acquisition. Cambridge: Cambridge University Press.

Montrul, S. (2015). The acquisition of heritage languages. Cambridge: Cambridge University Press.

Montrul, S. (2018). Heritage language development: Connecting the dots. International Journal of Bilingualism, 22, 530-546.

Moro, A. (2016). Impossible languages. Cambridge, MA: MIT Press. Myles, F. (2005). The emergence of morpho-syntactic structure in French L2. In

J.-M. Dewaele (Ed.), Focus on French as a foreign language: Multidisciplinary approaches (pp. 88–113). Clevedon: Multilingual Matters.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

124 Linguistics and Language Learning

Özçelik, Ö. (2018). Interface hypothesis and the L2 acquisition of quantificational scope at the syntax-semantics-pragmatics interface. Language Acquisition, 25, 213–223.

Özçelik, Ö., & Sprouse, R. A. (2017). Emergent knowledge of a universal phonological principle in the L2 acquisition of vowel harmony in Turkish: A “four”-fold poverty of the stimulus in L2 acquisition. Second Language Research, 33, 179–206.

Paradis, M. (2004). A neurolinguistic theory of bilingualism. Amsterdam: John Benjamins. Paradis, J., Rice, M., Crago, M., & Marquis, J. (2008). The acquisition of tense in English:

Distinguishing child second language from first language and specific language impair- ment. Applied Psycholinguistics, 29, 689–722.

Pierce, A. (1992). Language acquisition and syntactic theory: A comparative analysis of French and English child grammars. London: Kluwer.

Pinel, P., Fauchereau, F., Moreno, A., Barbot, A., Lathrop, M., Zelenika, D., . . . Dehaene, S. (2012). Genetic Variants of FOXP2 and KIAA0319/TTRAP/THEM2 locus are associated with altered brain activation in distinct language-related regions. Journal of Neuroscience, 32, 817–825.

Polinsky, M. (2011). Reanalysis in adult heritage language: New evidence in support of attrition. Studies in Second Language Acquisition, 33, 305–328.

Prévost, P., & Paradis, J. (Eds.). (2004). The acquisition of French in different contexts: Focus on functional categories. Amsterdam: John Benjamins.

Prévost, P., &White, L. (2000). Missing surface inflection or impairment in second lan- guage acquisition? Evidence from tense and agreement. Second Language Research, 16, 103–133.

Radford, A. (1990). Syntactic theory and the acquisition of English syntax. Oxford: Basil Blackwell.

Radford, A. (1997). Syntax: A minimalist introduction. Cambridge: Cambridge University Press.

Rankin, T., & Unsworth, S. (2016). Beyond poverty: Engaging with input in generative SLA. Second Language Research, 34, 563–572.

Renaud, C. (2014). A processing investigation of the accessibility of the uninterpretable gender feature in L2 French and L2 Spanish adjective agreement. Linguistic Approaches to Bilingualism, 4, 222–255.

Reuter, M. S., Riess, A., Moog, U., Briggs, T. A., Chandler, K. E., Rauch, A., . . . Zweier, C. (2017). FOXP2 variants in 14 individuals with developmental speech and language disor- ders broaden the mutational and clinical spectrum. Journal of Medical Genetics, 54, 64–72.

Roberts, I. (2012). Macroparameters and minimalism: A programme for comparative research. In C. Galves, S. Cyrino, R. Lopez, F. Sândalo, & J. Avelar (Eds.), Parameter theory and linguistic change (pp. 320–335). Oxford: Oxford University Press.

Roberts, I. (2016). Object clitics. In A. Ledgeway & M. Maiden (Eds.), The Oxford guide to the romance languages (pp. 786–801). Oxford: Oxford University Press.

Roberts, I. (2017a). Introduction. In I. Roberts (Ed.), The Oxford handbook of Universal Grammar (pp. 1–34). Oxford: Oxford University Press.

Roberts, I. (Ed.). (2017b). The Oxford handbook of Universal Grammar. Oxford: Oxford University Press.

Roberts, I., & Holmberg, A. (2010). Introduction: Parameters in minimalist theory. In T. Beiberauer, A. Holmberg, I. Roberts, & M. Sheehan (Eds.), Parametric variation: Null subjects in Minimalist theory (pp. 1–57). Cambridge: Cambridge University Press.

Roberts, L., González Alonso, J., Pliatsikas, C., & Rothman, J. (2018). Evidence from neurolinguistic methodologies: Can it actually inform linguistic/language acquisition

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 125

theories and translate to evidence-based applications? Second Language Research, 34, 125–143.

Rogers, V. (2009). Syntactic development in the second language acquisition of French by instructed English learners. PhD dissertation, Newcastle University, Newcastle upon Tyne, UK.

Roncaglia-Denissen, M. P., Schmidt-Kassow, M., Heine, A., & Kotz, S. A. (2015). On the impact of L2 speech rhythm on syntactic ambiguity resolution. Second Language Research, 31, 157–178.

Rothman, J., Bayram, F., Cunnings, I., & González Alonso, J. (in press). Formal linguistics approaches to adult second language acquisition and processing. In J. W. Schwieter & A. Benati (Eds.), The Cambridge handbook of language learning. Cambridge: Cambridge University Press.

Rothman, J., & Slabakova, R. (Eds.). (2011). Acquisition at the linguistic interfaces [Spe- cial Issue]. Lingua, 121.

Rothman, J., & Slabakova, R. (2018). The generative approach to SLA and its place in modern second language studies. Studies in Second Language Acquisition, 40, 417–444.

Schachter, J. (1996). Maturation and the issue of Universal Grammar in second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 159–193). San Diego: Academic Press.

Schlyter, S. (1986). Surextension et sous-extension dans l’acquisition des verbes de mou- vements/déplacement. In A Giacomi & D. Véronique (Eds.), Acquisition d’une langue étrangère (pp. 475–491). Aix-en-Provence: Université de Provence.

Schmid, M. S. (2011). Language attrition. Cambridge: Cambridge University Press. Schmid, M. S. (2014). The debate on maturational constraints in bilingual development:

A perspective from first-language attrition. Language Acquisition, 21, 386–410. Schmid, M. S., Gilbers, S., & Nota, A. (2014). Ultimate attainment in late second language

acquisition: Phonetic and grammatical challenges in advanced Dutch—English bilin- gualism. Second Language Research, 30, 129–157.

Schmid, M., & Köpke, B. (2018). The relevance of first language attrition to theories of bilingual development. Linguistic Approaches to Bilingualism, 7, 637–667.

Schwartz, B., & Sprouse, R. (1994). Word order and nominative case in nonnative lan- guage acquisition: A longitudinal study of (L1 Turkish) German interlanguage. In T. Hoekstra & B. Schwartz (Eds.), Language acquisition studies in generative grammar (pp. 317–368). Amsterdam: John Benjamins.

Schwartz, B., & Sprouse, R. (1996). L2 cognitive states and the full transfer/full access model. Second Language Research, 12, 40–72.

Schwartz, B., & Sprouse, R. (2013). Generative approaches and the poverty of the stimu- lus. In J. Herschensohn & M. Young-Scholten (Eds.), The Cambridge handbook of second language acquisition (pp. 137–158). Cambridge: Cambridge University Press.

Schwartz, B., & Sprouse, R. (2017). The role of Universal Grammar in nonnative lan- guage acquisition. In I. Roberts (Ed.), The Oxford handbook of Universal Grammar (pp. 289–304). Oxford: Oxford University Press.

Silva-Corvalán, C. (2014). Bilingual language acquisition: Spanish and English in the first six years. Cambridge: Cambridge University Press.

Slabakova, R. (2006). Is there a critical period for semantics? Second Language Research, 22, 302–338.

Slabakova, R. (2008). Meaning in the second language. Berlin: Mouton de Gruyter. Slabakova, R. (Ed.). (2009). The fundamental difference hypothesis twenty years later

[Special Issue]. Studies in Second Language Acquisition, 31.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

126 Linguistics and Language Learning

Slabakova, R. (2013). What is easy and what is hard in second language acquisition: A gen- erative perspective. In M. del Pilar García Mayo, M. Junkal Gutierrez-Mangado, & M. Martínez Adrián (Eds.), contemporary approaches to second language acquisition (pp. 5–28). Amsterdam: John Benjamins.

Slabakova, R. (2016). Second language acquisition. Oxford: Oxford University Press. Serratrice, L., Sorace, A., & Paoli, S. (2004). Crosslinguistic influence at the syntax-prag-

matic interface: Subjects and objects in English-Italian bilingual and monolingual acquisition. Bilingualism: Language and Cognition, 7, 183–205.

Sharwood Smith, M. (2017). Introducing language and cognition: A map of the mind. Cam- bridge: Cambridge University Press.

Smith, N. (1999). Chomsky: Ideas and ideals. Cambridge: Cambridge University Press. Smith, N., & Tsimpli, I. (1995). The mind of a savant: Language learning and modularity.

Oxford: Wiley Blackwell. Smith, N., Tsimpli, I., Morgan, G., & Woll, B. (2011). The signs of a savant: Language against

the odds. Cambridge: Cambridge University Press. Snape, N., & Kupisch, T. (2016). Second language acquisition: Second language systems. Lon-

don: Palgrave. Snyder, W. (2007). Child language: The parametric approach. Oxford: Oxford University Press. Sorace, A. (1996). The use of acceptability judgements in second language acquisition

research. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 375–409). San Diego: Academic Press.

Sorace, A. (2005). Selective optionality in language development. In L. Cornips & K. Corrigan (Eds.), Syntax and variation: Reconciling the biological and the social (pp. 55–80). Amsterdam: John Benjamins.

Sorace, A. (2011). Pinning down the concept of “interface” in bilingualism. Linguistic Approaches to Bilingualism, 1, 1–33.

Sorace, A., & Filiaci, F. (2006). Anaphora resolution in near-native speakers of Italian. Second Language Research, 22, 339–368.

Sorace, A., & Serratrice, L. (2009). Internal and external interfaces in bilingual language development: Beyond structural overlap. International Journal of Bilingualism, 13, 1–16.

Thomas, M. (1991). Universal Grammar and knowledge of reflexives in a second language. PhD dissertation, Harvard University, Cambridge, MA.

Thomas, M. (2004). Universal Grammar in second language acquisition: A history. Abingdon/ New York: Routledge.

Towell, R., & Hawkins, R. (1994). Approaches to second language acquisition. Clevedon: Multilingual Matters.

Towell, R., & Hawkins, R. (Eds.) (2004). Empirical evidence and theories of representa- tion in current research in second language acquisition. [Special issue]. Transactions of the Philological Society, 102.

Trahey, M., & White, L. (1993). Positive evidence and pre-emption in the second lan- guage classroom. Studies in Second Language Acquisition, 15, 181–204.

Tsimpli, I., & Dimitrakopoulou, M. (2007). The interpretability hypothesis: Evidence from wh-interrogatives in second language acquisition. Second Language Research, 23, 215–242.

Tsimpli, I., & Sorace, A. (2006). Differentiating interfaces: L2 performance in syntax- semantics and syntax-discourse phenomena. In D. Bamman, T. Magnitskaia, & C. Zaller (Eds.), Proceedings of the 30th annual Boston University conference on language development (BUCLD) (pp. 653–664). Somerville, MA: Cascadilla Press.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use

Linguistics and Language Learning 127

Vainikka, A., & Young-Scholten, M. (1996a). The early stages in adult L2 syntax: Addi- tional evidence from Romance speakers. Second Language Research, 12, 140–176.

Vainikka, A., & Young-Scholten, M. (1996b). Gradual development of L2 phrase structure. Second Language Research, 12, 7–39.

Vainikka, A., & Young-Scholten, M. (1998). The initial state in the L2 acquisition of phrase structure. In S. Flynn, G. Martohardjono, & W. O’Neil (Eds.), The generative study of second language acquisition (pp. 17–34). Mahwah, NJ: Lawrence Erlbaum.

Vainikka, A., & Young-Scholten, M. (2007). Minimalism vs. organic syntax. In S. Karimi, V. Samiian, & W. Wilkins (Eds.), Phrasal and clausal architecture: Syntactic derivation and interpretation. In honor of Joseph Emonds (pp. 319–338). Amsterdam: John Benjamins.

Vainikka, A., & Young-Scholten, M. (2011). The acquisition of German: Introducing organic grammar. Berlin: de Gruyter Mouton.

van der Lely, H. K. J. (Ed.). (1998). Specific Language Impairment in children [Special Issue]. Language Acquisition, 7.

van Hout, R., Hulk, A., Kuiken, F., & Towell, R. (Eds.). (2003). The lexicon-syntax interface in second language acquisition. Amsterdam: John Benjamins.

Véronique, D. (1986). L’apprentissage du français par des travailleurs marocains et les pro- cessus de pidginisation et de créolisation. In A. Giacome & D. Véronique (Eds.), Acqui- sition d’une langue étrangère (pp. 559–584). Aix-en-Provence: Université de Provence.

White, L. (1991). Adverb placement in second language acquisition: Some effects of posi- tive and negative evidence in the classroom. Second Language Research, 7, 133–161.

White, L. (1992). Long and short verb movement in second language acquisition. Cana- dian Journal of Linguistics, 37, 273–286.

White, L. (2003). Second language acquisition and Universal Grammar. Cambridge: Cam- bridge University Press.

White, L. (2009a). Grammatical theory: Interfaces and L2 knowledge. In W. C. Ritchie & T. K. Bhatia (Eds.), The new handbook of second language acquisition (pp. 49–68). Bingley: Emerald Group.

White, L. (2009b). Some questions about feature reassembly. Second Language Research, 25, 343–348.

White, L., Travis, L., & MacLachlan, A. (1992). The acquisition of wh-question formation by Malagasy learners of English: Evidence for universal grammar. Canadian Journal of Linguistics, 37, 341–368.

Whong, M., Gil, K.-H., & Marsden, H. (Eds.). (2013). Universal Grammar and the second language classroom. Dordrecht: Springer.

Yang, C., Crain, S., Berwick, R. C., Chomsky, N., & Bolhuis, J. J. (2017). The growth of language: Universal Grammar, experience, and principles of computation. Neurosci- ence & Biobehavioral Reviews, 81, 103–119.

Yuan, B. (2001). The status of thematic verbs in the second language acquisition of Chi- nese. Second Language Research, 14, 248–272.

Zobl, H. (1980). The formal and developmental selectivity of L1 influence on L2 acquisi- tion. Language Learning, 30, 43–57.

EBSCOhost - printed on 4/6/2020 6:13 AM via UNIVERSITY OF ESSEX. All use subject to https://www.ebsco.com/terms-of-use